The 44^th Annual Meeting of the European Society for Blood and Marrow Transplantation: Physicians Poster Sessions

Copyright: Modified and published with permission from CME Congresses Ltd © 2018.

http://www.ebmt2018.org/

Sponsorship Statement: Publication of this supplement is sponsored by the European Society for Blood and Marrow Transplantation.

PHYSICIANS – POSTER SESSION

Acute leukaemia

P001

A pilot study of daunorubicin-augmented hyper-CVAD induction chemotherapy followed by allogeneic hematopoietic cell transplantation for adults with acute lymphoblastic leukemia

Sung-Yong Kim, Ji Hyun Park, So Young Yoon, Yo-Han Cho, Mark Hong Lee

Konkuk University Medical Center, Konkuk University School of Medicine, Seoul, Korea, Republic of

Background: Induction of complete remission (CR) is imperative for long-term survival in adult acute lymphoblastic leukemia (ALL) patients who are eligible for transplantation. Hyper-CVAD chemotherapy is one of the widely-used frontline remission induction regimens for these patients. This combination regimen showed a CR rate of 74% and mortality of 6% after one course of therapy (cycle 1). This indicates room for improvement. Prior studies suggest that daunorubicin has a lower frequency of cardiomyopathy and other severe complications in leukemia patients than doxorubicin, and that a high induction dose of daunorubicin could contribute to a favorable treatment outcome in adult ALL patients. Based on these observations, a pilot trial was conducted with adult ALL patients using a daunorubicin-augmented hyper-CVAD (hyper-CVDD) regimen in which the doxorubicin in the standard hyper-CVAD regimen was replaced with an increased dose of daunorubicin to improve the efficacy of the regimen.

Methods: We conducted a pilot trial of frontline remission induction using daunorubicin-augmented hyper-CVAD regimen (hyper-CVDD) in adult ALL patients (n=14). From the standard hyper-CVAD regimen, doxorubicin was replaced with daunorubicin. Daunorubicin was administrated on days 4, 11, and 12 at a dose of 45 mg/m²/day. If the study patients were less than 65 years of age and achieved CR, they proceeded to allo-HCT after one course of consolidation chemotherapy. A total of 14 patients were enrolled in the trial. The median age was 41.5 (range, 18–62). Six patients (42.9%) were positive for BCR/ABL transcript and received tyrosine kinase inhibitor until transplantation. Four patients (28.6%) had mixed phenotype acute leukemia; B + myeloid (n=2), T + myeloid (n=1), and B + T (n=1). Nine patients had B cell and two had T cell ALL.

Results: All patients completed the hyper-CVDD treatment as per protocol. All 14 patients (100%) achieved CR. The median time to neutrophil recovery and platelet recovery was 20 days (range, 13-30) and 25 days (range, 20-40) from treatment, respectively. Two patients died of infection before and after the first consolidation, respectively. There were no deaths related to non-infectious toxicity including cardiotoxicity. Two patients relapsed after the first consolidation; one received salvage chemotherapy before allo-HCT and the other proceeded to transplantation without salvage chemotherapy because the percentage of blasts in bone marrow was 7%. Thus, twelve patients among 14 transplant-eligible patients were able to proceed to allo-HCT. Overall survival (OS) and event-free survival (EFS) of the study patients was 57.3% and 42.9% at three years. OS and relapse-free survival of transplanted patients was 66.8% and 55.0% at three years.

Conclusions: This pilot trial suggests the favorable efficacy of the hyper-CVDD chemotherapy as a frontline remission induction regimen before transplantation. Further clinical trials using this regimen are warranted.

Conflict of interest: There is no conflict of interest to declare.

[[P001 Figure]

Overall survival of all the study patients (A) and the transplanted patients (B)]

P002

Allogeneic hematopoietic cell transplantation for primary refractory acute leukemia patients: a retrospective GITMO score based analysis

Ilaria Cutini, Chiara Nozzoli, Antonella Gozzini, Stefano Guidi, Irene Donnini, Riccardo Saccardi

AOU Careggi, Cellular Therapies and Transfusional Medicine Unit, Florence, Italy

Background: Patients with acute leukemia (AL) who fail to achieve complete remission (CR) have a dismal prognosis. Only few of them can be rescue after allogeneic hematopoietic cell transplantation (HCT). We retrospectively applied the GITMO score for PIF AL patients, that divides patients in 3 different categories; low, intermediate and high risk (Todisco E, BMT 2017).

Methods: The study population included 25 patients with AL presented as primary induction failure (PIF) who had received an allogeneic HCT between 1 March 2014 and 30 September 2017 at our institution. Median age was 51 yo. Disease characteristics and prior history of hematological diseased and gender are summarized in Table 1. Patients received myeloablative (n=16) or reduced-intensity conditioning (RIC) regimens (n=9). Median time from diagnosis to transplant was 7 months, (range 0-9 months).

Results: Among PIF AL population 16 belong to low risk group; 7 to intermediate group and 2 to high risk group respectfully. 18 out of 25 were evaluable on +30 days from transplant: 16 patients achieved complete hematological remission (CHR), 2 patients shown refractory disease. 7 patients died within 2 months (range 0-4 months) from transplant due to multiorgan failure (2 pts), sepsis (3 pts) and engraftment failure (2pts). 12 out of 25 (35%) died due to TRM. Mean OS was 521 days, median OS was 331 days. Only 7 out of 25 patients (28%) are still alive without active disease, mean follow up was 337 days (79-1348 days). Among the three GITMO categories, the mean and median OS was 527 and 364 days, 581 and 341 days, and 262 days for low, intermediate and high risk group respectively (p=0.126). 8 patients experienced aGVHD and 7 patients cGVHD. All patients had received corticosteroids as frontline treatment. Those unresponsive were treated with second-line immunosuppressors. Mean time for developing GVHD was 2 and 4 months for aGVHD and cGVHD respectively. GVHD characteristic was shown in table 1. A trend to a better OS was shown in patients who developed aGVHD (p=0.380) and cGVHD (p=0.219). No impact was shown for CMV sierological status. (p=0,651). No impact was shown for myeloablative versus RIC regimens (p=0.983).

Conclusions: The clinical outcome of PIF AL patients is poor and only a minor proportion of patients is rescued by HSCT. GITMO score can be used to create a risk score that helps to identify patients who most likely benefit from the procedure. The availability of reliable prognostic factor is particularly important in the era of alternative donor, such as haploidentical source. The small sample size may prevent to assess more significant differences across the population.

Conflict of interest: Nothing to disclose

	n°25	%
Age	51 yo (20-66)
Gender M F	10 15	40 60
Diagnosis AML ALL	24 1	96 4
De novo Prior history of: MDS ET IMF Chemo Chemo+RT HSCT	16 2 2 1 1 1 2	64 8 8 4 4 4 8
AML RISK (ELN 2017) LOW RISK INTERMEDIATE RISK HIGH RISK	2 11 12	8 44 48
CMV STATUS NEGATIVE POSITIVE	5 20	20 80
PIF GITMO SCORE LOW INTERMEDIATE HIGH	16 9	64 36
aGVHD	Skin (8) Global grade 1 (5) Global grade 2 (3) Intestine (2) Global grade 1 (0) Global grade 2 (1) Global grade 3 (1) liver (0)	63 37 0 50 50 0
cGVHD	Mild Moderate Severe	2 4 1

[ [P002 Table] Table 1]

P003

Allogeneic hematopoietic cell transplantation in aplasia after Cladribine/Cytarabine salvage chemotherapy for refractory or relapsed AML and MDS - a single center experience

Jan Vydra, Markéta Šťastná Marková, Veronika Válková, Ludmila Nováková, Barbora Čemusová, Petr Soukup, Michal Kolář, Antonín Vítek, Petr Cetkovský

Institute of Hematology and Blood Transfusion, Prague, Czech Republic

Background: Patients with chemorefractory acute myeloid leukemia (AML) and myelodysplastic syndromes (MDS) have poor prognosis. Results of standard HCT in such setting have been unsatisfactory and several sequential treatment regimens have been published consisting of cytoreductive chemotherapy followed by HCT in aplasia, without awaiting the recovery of normal hematopoiesis. The rationale for this approach is to reduce the tumor burden immediately prior to transplantation.

Methods: We performed a retrospective analysis of the outcome of treatment of patients with refractory or relapsing AML or MDS who received allogeneic HCT in aplasia after cladribine/cytarabine salvage chemotherapy at Institute of Hematology and Blood Transfusion in Prague. Patient characteristics and outcomes were retrieved from our database, which contains prospectively collected information on all patients who receive HCT at our institution. All patients signed informed consent with data collection and analysis and the study was reviewed by the institutional ethical committee.

Cladribine (5mg/m2/day IV) and Cytarabine (2000mg/m2/day IV) were administered on days -15 to -11 before HCT, the conditioning regimen started on day -7. Unmanipulated PBSC grafts were infused on day 0. GvHD prophylaxis consisted of post-transplant high dose cyclophosphamide (50mg/kg on days +3 and 4), cyclosporine and mycophenolate mofetil.

Results: 21 patients (17 with AML and 4 with MDS) treated between 2015 and 2017 were included. 5 patients (24%) relapsed after previous HCT, 10(48%) had primary induction failure, 4 (19%) had relapse refractory to salvage chemotherapy and 2 (9%) had early molecular relapse of poor risk AML. Median followup time is 216 (11 - 822) days, median age 53 (35-61) years. HCT-CI score was 0-1 in 12 (56%), 2-3 in 5(24%) and >3 in 4 (20%). The conditioning regimen was Flu/Cy/Mel in 18 (85.7%), Flu/Bu[12.8mg iv] in 2 (9.5%) and Flu/Cy/2GyTBI in 1 (4.8%) patient. Donors were haploidentical in 19 (90%), matched sibling in 1(5%) and mismatched unrelated in 1(5%) patient.

Neutrophil engraftment was prompt, 90.5% of patients engrafted and achieved CR, 2 (9.5%) patients died before engraftment. Median time to engraftment was 19.5 days. 85.7% of patients engrafted platelets, median time to platelet engraftment was 31 days.

Estimated OS at two years is 36,5% (95%CI 19.6 - 68.9%), RFS 29%(95%CI 14-60.4%). Cumulative incidence of NRM and relapse were 25.9% and 45%, respectively. In subgroup of patients with HCT-CI score < = 3, incidence of NRM was 13.3%.

Incidence of acute GVHD of any grade was 53%. Incidence of grade 3 and 4 aGVHD were 19.6% and 9.5%, respectively. Incidence of chronic GVHD was 36.7%.

14 (66%) patients had fever in the first days following haploidentical graft infusion. Oral mucositis was usually mild, only one patient had grade 3 oral mucositis (after Flu/Bu conditioning). Mild or moderate SOS syndrome was seen in 7 (33%) of patients. Two patients died early before engraftment due to multiorgan failure.

Conclusions: Allogeneic HCT in aplasia after cladribine/cytarabine salvage chemotherapy for AML or MDS is a feasible and active treatment regimen, which needs to be studied in a prospective trial.

Conflict of interest: None of the authors has anything to disclose

P004

Allogeneic HSCT in Acute Lymphoblastic Leukemia: long-term outcome of 100 adult patients in 15 years

Silvia Pierini, Fabio Giglio, Daniele Mannina, Massimo Bernardi, Jacopo Peccatori, Consuelo Corti, Andrea Assanelli, Gabriele Casirati, Raffaella Greco, Maria Teresa Lupo Stanghellini, Magda Marcatti, Sarah Marktel, Simona Piemontese, Fabio Ciceri

San Raffaele Scientific Institute IRCCS, Milano, Italy, Hematology and Bone Marrow Transplantation Unit, Milano, Italy

Background: Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is an effective curative option to treat adult patients with high-risk acute lymphoblastic leukemia (ALL). source of HSCT and type of conditioning continue to be debated. The objective of this study is to perform a retrospective survey on allo-HSCT in adult ALL patients, in our Institution, over the past 15 years.

Methods: A total of 116 allo-HSCT on 100 patients affected by ALL have been performed between 2001 and 2017 in our Institution. Our analysis was then restricted to 72 patients at their first allo-HSCT and all data presented refer this cohort.

Results: Patients and HSCT characteristics are shown in table 1.

3 conditioning regimes were performed using high dose of total dose irradiation(10-12 Gy) plus alkylating agent, treosulfan-based plus low dose TBI (2-4Gy) or treosulfan-based without TBI. GvHD prophylaxis was cyclosporine based in 25 patients, rapamycin based in 31 and cyclophosphamide plus rapamycin in 16. With a median follow up of 43 months, the estimated probabilities of overall survival (OS) were 52+/-6% and 34+/-6% respectively at 1 and 5 years; the estimated probabilities of leukemia free survival (LFS) were 42+/-6% and 26+/-6%, respectively at 1 and 5 years and the non-relapse mortality (NRM) probabilities were 8+/-3% at 100 days and 39+/-7% at 3 years. The cumulative incidence of aGvHD grade III-IV at 100 days was 20+/-5% and the cumulative incidence of cGvHD moderate-severe at 2 years was 25+/-7%.

In univariate analysis, factors significantly associated with better overall survival were complete remission at HSCT (p=0.001), and a low dose of TBI (2 or 4Gy) (p=0.005). The same results were confirmed in multivariate analysis (low dose TBI with p=0.022 and CR at HSCT with p=0.001).

No differences in OS and LFS were observed between MMRD and MUD (p=0.143 for OS) (p=0.286 for OS). MRD had a better OS compared to MMRD/MUD (p=0.083).

Conclusions: in this small retrospective series of HSCT we could confirm that the state of disease at transplant is the most significant prognostic factor.

I our cohort, TBI is associated with better OS and LFS but only at low dose, in fact a high dose of TBI increases the NRM and aGvHD with statistical significance, finally impacting on OS. Low dose of TBI appear to have better outcome than high dose TBI and non TBI conditioning. In patients with ALL at their first transplant no significant differences in OS and LFS emerged between MUD and MMRD transplants, confirming that haplo-HSCT is an effective option in patients lacking a matched donor.

Conflict of interest: Nothing to disclose

	116 all HSCT	72 first HSCT		116 all HSCT	72 first HSCT
Median age at HSCT	37.2 (18-68)	35.50 (18-68)	High dose TBI (10-12Gy)	17	12
B-ALL	78	49	Treosulfan-based and low dose TBI (2-4Gy)	32	18
T-ALL	28	23	Treosulfan-based	67	42
CR1	37	32	MUD	23	21
CR≥2	20	11	MMRD	58	30
Rel	59	29	MRD	25	21
			UCB	10	0

[ [P004 Table] Tab 1. Patients and HSCT characteristics]

P005

Allogeneic Transplantation for Acute Myeloid Leukemia with Monosomy 7 or Deletion 7q: a Study from the Acute Leukemia Working Party (ALWP) of the EBMT

Xavier Poiré ¹ , Myriam Labopin ² , Emmanuelle Polge ² , Liisa Volin ³ , Jürgen Finke ⁴ , Arnold Ganser ⁵ , Didier Blaise ⁶ , Ibrahim Yakoub-Agha ⁷ , Dietrich Beelen ⁸ , Noël Milpied ⁹ , Marie-Pierre Ledoux ¹⁰ , Gérard Socié ¹¹ , Dietger Niederwieser ¹² , Mauricette Michallet ¹³ , Johan Maertens ¹⁴ , Jan Cornelissen ¹⁵ , Charles Craddock ¹⁶ , Mohamad Mohty ² , Jordi Esteve ¹⁷ , Arnon Nagler ¹⁸

¹ Cliniques Universitaires St-Luc, Hematology, Brussels, Belgium; ² Hôpital Saint Antoine, INSERM UMR 938, Paris, France; ³ HUCH Comprehensive Cancer Center, Stem Cell Transplantation Unit, Helsinki, Finland; ⁴ University of Freiburg, Freiburg, Germany; ⁵ Hannover Medical School, Department of Haematology, Hemostasis, Oncology and Stem Cell Transplantation, Hannover, Germany; ⁶ Institut Paoli Calmettes, Department of Hematology, Centre de Recherche en Cancérologie de Marseille (CRCM), Marseille, France; ⁷ Hôpital HURIEZ UAM allo-CSH, CHRU, Lille, France; ⁸ Univeristy Hospital Essen, Essen, Germany; ⁹ CHU Bordeaux Hôpital Haut-leveque, Pessac, France; ¹⁰ Nouvel Hopital Civil, Strasbourg, France; ¹¹ Hôpital Saint Louis & Université Paris 7, Denis Diderot, Paris, France; ¹² University Hospital Leipzig, Division of Haematology & Oncology, Leipzig, Germany; ¹³ Hôpital Lyon Sud, Lyon, France; ¹⁴ University Hospital Gasthuisberg, Department of Hematology, Leuven, Belgium; ¹⁵ Erasmus MC Cancer Institute, University Medical Center, Rotterdam, Netherlands; ¹⁶ Queen Elizabeth Hospital, Birmingham, United Kingdom; ¹⁷ Hospital Clinic, Barcelona, Spain; ¹⁸ Chaim Sheba Medical Center, Tel Hashomer, Israel

Background: Monosomy 7 or deletion 7q (-7/7q-) is the most frequent cytogenetic feature reported in acute myeloid leukemia (AML) and is associated with a dismal outcome. Allogeneic stem cell transplantation (SCT) allows to significantly improve their survival. Nevertheless, -7/7q- occurs frequently with other high-risk cytogenetic abnormalities, which may further influence the outcomes after SCT. Our aim was to evaluate the role of SCT in AML with -7/7q- in the context of additional cytogenetic abnormalities.

Methods: We selected adult with AML and -7/7q- reported to the EBMT registry with a complete cytogenetic report. We included only first SCT from a sibling or unrelated donor (UD) performed between 2000 and 2016.

Results: A total of 1109 patients has been allocated. Median age was 52-year-old (range, 18-76) and median follow-up was 34 months (range, 0.7-193). At the time of SCT, 702 patients (63%) were in first remission (CR1), 58 (5%) in subsequent remission and 349 (32%) had active disease. An UD was used in 613 patients (55%). A myeloablative conditioning regimen was administered in 50%.

The 2-year leukemia-free survival (LFS) and overall survival (OS) were 30% and 36%, respectively. The 2-year non-relapse mortality (NRM) was 20% and the 2-year cumulative incidence of relapse (RI) was 50%. The 2-year graft-versus-host disease (GvHD) and relapse-free survival (GRFS) was 21%. In multivariate analysis, active disease was significantly associated with increased NRM and RI, which translated into worse LFS, OS and GRFS.

To evaluate the impact of additional cytogenetic abnormalities, we have been able to sort out 5 different groups. The first group included 523 patients with -7/7q- with or without complex karyotype (CK), but no monosomal karyotype (MK), monosomy 5 or deletion 5q (-5/5q-), 17p abnormalities (abn(17p)) or inversion of chromosome 3 (inv(3)).

Group 2 included 199 patients with -7/7q- within MK but without -5/5q-, abn(17p) or inv(3).

Group 3 included 151 patients with -7/7q- in association with -5/5q- but no abn(17p) or inv(3).

Group 4 included 125 patients with -7/7q- and abn(17p) but without inv(3). Finally, 111 patients with the combination of -7/7q- and inv(3) were assigned to the fifth group. NRM was similar across the groups. The 2-year LFS was 39.7%, 28.5%, 22.5%, 16.4% and 13.7% from group 1 to group 5, respectively (p< 0.001). The 2-year OS in the

group 5 were 47.7%, 34.9%, 28.2%, 15.9% and 19.7%, respectively (p< 0.001). The 2-year GRFS was 29.7%, 16.6%, 14.5%, 8.1% and 8.9% from group 1 to group 5 (p< 0.001). Multivariate analysis confirmed those significant differences across groups, even when the analysis was focused on CR1 patients.

Conclusions: SCT in -7/7q- AML provides durable response in one third of the patients. Disease status at the time of transplantation remains the strongest prognostic factor for worse outcomes. The presence of -7/7q- with or without CK in the absence of MK, -5/5q-, abn(17p) or inv(3) was associated with a better survival after SCT. On the contrary, addition of -5/5q-, abn(17p) or inv(3) identifies a sub-group with a limited benefit from SCT, who may be candidate for post-SCT interventions.

Conflict of interest: Nothing to disclose

[P005 Figure]

$$$$

P006

Abstract previously published

P007

Analysis of three salvage treatment regimens after relapsed/refractory acute myeloi leukemia (AML) (FLAG-Ida, FLAGO-IDA and PLERIFLAG)

Juan Bergua¹, David Martinez-Cuadron², Beatriz Boluda², Pilar Martinez³, Rebecca Rodriguez-Veiga², Jordi Esteve⁴, Susana Vives⁵, Josefina Serrano⁶, Belen Vidriales⁷, Olga Salamero⁸, Carlos Carretero², Ana Jimenez-Ubieto⁹, Julio Prieto¹⁰, Marina Díaz-Beyá⁴, Salut Brunet¹¹, Celina Benavente¹², Jose Antonio Perez-Simon¹³, Sara Suarez-Varela¹, Federico Moscado¹⁴, Miguel Angel Sanz², Pau Montesinos²

¹ Hospital San Pedro de Alcántara, Hematologia, Caceres, Spain; ² Hospital La Fe, Hematologia, Valencia, Spain; ³ Hospital 12 de Octubre, Hematologia, Madrid, Spain; ⁴ Hospital Clinic de Barcelona, Hematologia, Barcelona, Spain; ⁵ Hospital Trias y Pujol, Badalona, Badalona, Spain; ⁶ Hospital Reina Sofía, Hematology, Cordoba, Spain; ⁷ Hospital Clinico de Salamanca, Hematologia, Salamanca, Spain; ⁸ Hospital Vall d’Hebron, Hematology, Barcelona, Spain; ⁹ Hospital 12 de Octubre, Hematology, Madrid, Spain; ¹⁰ Hospital de Don Benito, Hematologia, Don Benito, Spain; ¹¹ Hospital Sant Pau, Hematologia, Barcelona, Spain; ¹² Hospital Clinico San Carlos, Hematologia, Madrid, Spain; ¹³ Hospital Virgen del Rocio, Hematologia, Sevilla, Spain; ¹⁴ Hospital La Fe, Hematologia, Valenica, Spain

Background: We retropectively compare three salvage therapies based in FLAG-IDA scheme: FLAG-IDA (Fludarabine, Idarubicine, Cytarabine G-CSF as priming agent), FLAGO-IDA (Gemtuzumab plus FLAG-IDA, and PLERIFLAG (Plerixafor as priming factor). We use to compared two risk-criteria classifications HOVON (De Breems, 2005) and SALFLAGE (Bergua, 2016).

Methods: Statistical analysis: Differences between groups were tested using Fisher exact test. We compare differences onCR+Cri between treatments adjusting prognostic survival score groups using Mantel-Haenszel test using HOVON criteria (de Breems et al) and SALFLAGE criteria. Homogeneity of variables was tested using Woolf-test .Survival analysis by Kaplan-Maier stratified analysis. To compare PLERIFLAG against the other two types of salvage treatment we performed match-pair analysis adjusting variables as age, previous allogenic transplant, time to relapse (refractory, < 12 months and >12 months), karyotype using MRC, and FLT3 status. Kariotype risk was considered by HOVON criteria (inv16, t(8;21) and all the others Kariotype), and SALFLAGE (inv(16), unfavourable risk by MRC risk plus t(8;21), intermediate risk.

Patients: We analysed 300 patients relapsed or resistant to induction therapy, 251 treated with FLAG-IDA, 41 patientas with FLAGO-Ida, and 42 patients treated with PLERIFLAG. Analysis was conducted in different times; FLAG-Ida was the regiment of treatment of relapsed/resitant patients in PETHEMA studies, FLAGO was used in a short period before withdraw Gemtuzumab use by EMEA agencies, and PLERIFLAG was a phase 2 trial on relapsed/refractory patients. Both treatments were similar in terms of Age, Sex and previous myelodisplasia. The three types of treatment differed in characteristics. There were more unfavourable characteristics in PLERIFLAG vs. FLAG-Ida and FLAGO-IDA treatment: high risk karyotype (55% in PLERIFLAG,vs 41 in FLAG-IDA vs. 33% in FLAGO-Ida), FLT3 (10% in PLERIFLAG, 22% in FLAG-Ida, 20% in FLAGO-ida), time to relapse less than 12 months, or previous allogenic stem-cell transplantation. There were more patients classified as high risk using SALFLAGE criteria in PLERIFLAG 33(80%),vs FLAG-Ida (93, 48%) vs. FLAGO-Ida (12, 36%).

Results: Complete remission (CR) was achieved in 20patients (48.7%) treated with Pleriflag, in 96 (48%) treated with FLAG-IDA, and in 22 (61%) of the patients treated with treated with FLAGO-IDA. Adjusted by Hovon score andSalflage SALFLAGE score does not show statatisitical differences in CR+Cri rate (p=0.23, Odds ratio between PLERIFLAG and FLAG-Ida and FLAGO-Ida (0.27,CI: -1.1-0.5) . The number of patients who received allogenic- stem cell transplantation were more in the case of PLERIFLAG (25(60%), 9 (22%) of them sequential vs FLAG-IDA (85(34%)) vs. FLGO-IDA (7(18%)). There no were differences in overall survival (OS) between PLERIFLAG (median 10.56 monts, CI(6.7-NR), FLAG-IDA (medina 8.45 months, CI(7.2-13) and FLAGO-IDA (median 9.53 months, CI(3.78-25)(p=0.9) and EFS between the treatments, but there were differences when adjusted by SALFLAGE risk categories (p=0.02).

Conclusions: No differences between groups could be detected. These results may can be influenced by the high rate of Allegenic stem cell transplant in PLERIFLAG group; also supports the need of Allogenic stem trasplantation in sequential time schedule even if no CR has been achieved.

Clinical Trial Registry: eudrat 2011-000971-15

Conflict of interest: No conflict of interest

P008

Antibodies to red blood cells antigens in patients of hematological clinic

Elena Butina ¹ , Galina Zaitseva ¹ , Filipp Sherstnev ², Oleg Maximov³, Natalia Minaeva³, Igor Paramonov³

¹ Kirov Research Institute of Hematology and Blood Transfusion of the Federal Medical and Biological Agency of Russia, Immunohematology, Kirov, Russian Federation; ² Kirov Research Institute of Hematology and Blood Transfusion of the Federal Medical and Biological Agency of Russia, Transfusiology, Kirov, Russian Federation; ³ Kirov Research Institute of Hematology and Blood Transfusion of the Federal Medical and Biological Agency of Russia, Hematology, Kirov, Russian Federation

Background: Screening of red blood cell (RBC) alloantibodies is mandatory for every hospitalization of hematological patient. In case of antibodies detection, their specificity is established and, if necessary, individual selection of RBC containing blood components is carried out. Aim. To study frequency of detection of RBC alloantibodies in hematological patients.

Methods: Analysis of results of the immunohematological examination of 3132 patients of hematological clinic was carried out. Screening and identification of antibodies were performed with equipment and reagents from BioRad (USA).

Results: Specific RBC alloantibodies were detected in 2.1% of hematological patients: 2.28% of men, 2.44% of women and 0.48% of children. The most frequently detected antibodies were anti-D (0.32%), -E (0.32%) and -K (0.32%). Antibodies to antigens C (0.19%), c (0.13%), M (0.13%), Le^a (0.13%), DC (0.1%), Fy^a (0.1%) were also detected. The most uncommon were anti-C^W, -S, -Jk^a, -Jk^b, -Le^b, -Lu^a, -Lu^b antibodies, their frequency varied from 0.06% to 0.03%.

The maximum level of alloimmunization was observed in the group of patients with hemophilia A - 12.0% of patients. Relatively more frequently, antibodies are detected in patients with acute leukemia (3.5%), myelodysplastic syndrome (4.7%) and von Willebrand disease (4.9%). In the remaining groups of patients, alloantibodies were detected at a frequency of 0.1% to 2.3%.

Conclusions: Given that patients with blood diseases receive intensive transfusion therapy, it is extremely necessary to select donors that are compatible with recipient both for RBC antigens of ABO, Rhesus (C, c, E), Kell (K) systems and for results of indirect antiglobulin test.

Conflict of interest: None of the authors has anything to disclose.

P009

Assessing the prognostic impact of FLT3-ITD in acute myeloid leukemia patients undergoing haploidentical stem cell transplantation. An Acute Leukemia Working Party/EBMT analysis

Jonathan Canaani ¹ , Myriam Labopin ² , Xiao-Jun Huang ³ , William Arcese ⁴ , Fabio Ciceri ⁵ , Didier Blaise ⁶ , Giuseppe Irrera ⁷ , Lucia Lopez Corral ⁸ , Benedetto Bruno ⁹ , Stella Santarone ¹⁰ , Maria Teresa Van Lint ¹¹ , Antonin Vitek ¹² , Jordi Esteve ¹³ , Mohamad Mohty ² , Arnon Nagler ^2,14

¹ Chaim Sheba Medical Center, Hematology Division, Ramat Gan, Israel; ² EBMT Paris Study Office, Paris, France; ³ Peking University People's Hospital, Peking, China; ⁴ Tor Vergata University, Rome, Italy; ⁵ San Raffaele Scientific Institute, Hematology Division, Milano, Italy; ⁶ Institut Paoli Calmettes, Marseille, France; ⁷ Azienda Ospedaliera, Centro Unico Regionale Trapianti, Alberto Neri, Reggio Calabria, Italy; ⁸ Hospital Clinic de Salamanca, Salamanca, Spain; ⁹ Citta della Salute e della Scienza di Torino, Torino, Italy; ¹⁰ Ospedale Civile, Pescara, Italy; ¹¹ Ospedale San Martino, Genova, Italy; ¹² Institute of Hematology and Blood Transfusion, Prague, Czech Republic; ¹³ Hospital Clínic de Barcelona, Barcelona, Spain; ¹⁴ Chaim Sheba Medical Center, Ramat Gan, Israel

Background: Fms-like tyrosine kinase 3 internal tandem duplication (FLT3-ITD) mutation designates patients with acute myeloid leukemia as having a particularly high risk for disease relapse. It is currently unclear whether FLT3-ITD retains a significant prognostic role in haploidentical stem cell transplantation (haplo-SCT).

Methods: To assess whether FLT3-ITD is prognostically meaningful in T-cell replete haplo-SCT transplanted AML patients, we performed a comparative analysis between FLT3^wt and FLT3-ITD^mut using the multicenter registry of the acute leukemia working party of the European society for blood and marrow transplantation. Patients included in this study were AML patients over the age of 18 who underwent a first T-cell replete haplo-SCT in first remission.

Results: In all, a cohort consisting of 293 de-novo AML patients (202 FLT3^wt and 91 FLT3-ITD^mut) transplanted in first remission with T-cell replete haplo-SCT between 2005-2016 was analyzed. Both groups did not differ to a significant degree in terms of patient age, donor age, performance status, cytomegalovirus donor-recipient matching, and conditioning intensity. FLT3-ITD^mut patients were more likely to be nucleophosmin1 (NPM1) mutated as well as be in the Medical Research Council (MRC) intermediate risk cytogenetic risk category. In multivariate analysis, patients with FLT3-ITD had comparable rates of relapse incidence [Hazard ratio (HR)=1.34, confidence interval (CI) 95%, 0.67-2.7; P=0.9] and leukemia-free survival (HR=0.99, CI 95%, 0.62-1.57; P=0.9) to those of FLT3^wt patients. Survival was not significantly impacted by FLT3-ITD status (HR=0.96, CI 95%, 0.58-1.59; P=0.8), nor were the incidence of non-relapse mortality (HR=0.78, CI 95%, 0.41-1.46; P=0.44), and graft versus host disease-free/relapse-free survival (HR=0.84, CI 95%, 0.54-1.28; P=0.42). An analysis for patients treated either on anti-thymocyte globulin (ATG) protocols or post-transplant cyclophosphamide (PTCy) did not significantly correlate FLT3-ITD status with clinical outcome in either of these groups. Finally, a focused subset analysis of patients with MRC intermediate risk cytogenetics confirmed the absence of a prognostic impact of FLT3-ITD also for this group of patients.

Conclusions: The FLT3-ITD mutation possibly does not retain its prognostic significance in AML patients undergoing haplo-HCT, suggesting that haplo-SCT may conceivably overcome the negative prognostic impact of FLT3-ITD in AML.

Clinical Trial Registry: N/A

Conflict of interest: All authors declare nothing to disclose.

P010

Azacitidine and escalating doses of DLI is an effective combinatin treatment for patients relapsing post allogeneic bone marrow transplantation

Shereef Elmoamly ¹ , Dimitra Oikonomopoulou ² , Maria Kaparou ³ , Evgenia Xenou ⁴ , Amy Gudger ³ , Rebecca Pryor ⁴ , Hannah Giles ³ , Claire Horgan ³ , Alexandros Kanellopoulos ³ , Shankara Paneesha ³ , Bhuvan Kishore ³ , Richard Lovell ³ , Kathy Holder ³ , Lynn Ryan ³ , Katie O'Collins ³ , Hara Giatra ² , Ioannis Baltadakis ² , Dimitrios Karakasis ² , Emmanouil Nikolousis ³

¹ Heart of England NHS Trust, Haematology, Birmingham, United Kingdom; ² Evaggelismos General Hospital, Athens, Greece; ³ Heart of England NHS Trust, Birmingham, United Kingdom; ⁴ Heart of England NHS Foundation Trust, Birmingham, United Kingdom

Background: Patients with AML have a significant risk of relapse even after undergoing haematopoetic stem cell transplantation.In such patients the outlook is very poor with the majority of the patients surviving less than a year.Donor lymphocyte infusions are unable to tackle the kinetics of haematological relapse while intensive chemotherapy could result in mortality post allogeneic transplant leaving a significant treatment gap for this group of patients

Methods: Between 20011 and 2017, a total 25 patients [15 males, 10 females, median age 59.05 (40-73)] received a myeloablative or RIC allograft (17 from fully matched unrelated donors, and 8 from an identical sibling) for standard or high risk AML in 2 centres and subsequently relapsed.14 patients relapsed within first year while 10 patients relapsed more than a year post allogeneic stem cell transplant.The conditioning regimes included fludarabine, together with melphalan and Campath or ATG (13 patients),Fludarabine Busulphan and Campath (5 patients) and TBI/Cy or Bu/Cy(7 patients). Graft versus host disease prophylaxis was Ciclosporin for patients receiving Alemtuzumab or ATG but for patients who had a T-replete allograft Ciclosporin and low dose methotrexate was the preferred prophylaxis.All but four patients were transplanted in CR1.Patients who relapsed receive 2 courses of Azacitidine of 75mg/m2 followed by DLI in repeated cycles with escalating doses of DLI.Progression free survival(PFS) and overall survival(OS) were assessed using SPSS software

Results: Median follo up was 19.3months (range 3-53).) 7 patients remain alive. Causes of death (18 cases) include relapse or progression of the original disease (13 cases), and infection (4 cases) .12 patients achieved CR while 9 achieved PR and 4 had ongoing disease post treatment with Aza and DLI Median PFS was 11.3 months for patients who relapsed within a year post transplant receiving Azacitidine and DLI but for those who relapsed more than one year median PFS hasn't been reached.The influence of timing of relapse on PFS and OS was analysed, showing a trend towards better PFS for patients relapsing more than 12 months post allogeneic stem cell transplant and a significantly better OS for those patients who developed GvHD after Azacitidine and DLI administration.Disease risk was not significant on PFS or OS post relapse which might be due to the low number of patients in each risk group All patients in ongoing complete remission managed to achieve full donor chimerism in both whole blood and T cells

Conclusions: Our results indicate that the combination of Azacitidine with escalating doses of DLI could potentially restore remission in patients relapsing post allogeneic stem cell transplantation for a reasonable amount of time in this difficult cohort of patients especially for patients who relapse more than a year post transplant.Our findings should be evaluated in prospective clinical trials

Our results indicate that the combination of Azacitidine with escalating doses of DLI could potentially restore remission in patients relapsing post allogeneic stem cell transplantation for a reasonable amount of time in this difficult cohort of patients especially for patients who relapse more than a year post transplant.Our findings should be evaluated in prospective clinical trials.

Clinical Trial Registry: NA

Conflict of interest: No conflicts of interest to disclose

P011

Better outcome in children compared to AYAs and young adults after allogeneic HSCT for AML: a multicenter retrospective study from the SFGMTC

Cecile Pochon ¹ , Marie Detrait ² , Eliane Albuisson ³ , Jean-Hugues Dalle ⁴ , Didier Blaise ⁵ , Gérard Michel ⁶ , Ibrahim Yakoub-Agha ⁷ , Jacques Olivier Bay ⁸ , Patrice Chevallier ⁹ , Mohamad Mohty ¹⁰ , Noel Milpied ¹¹ , Mauricette Michallet ¹² , Yves Bertrand ¹³ , Ali Bazarbachi ¹⁴ , Yves Chalandon ¹⁵ , Yves Beguin ¹⁶ , Stéphanie Nguyen Quoc ¹⁷ , Regis Peffault de Latour ¹⁸ , Marie-Thérèse Rubio ²

¹ University Hospital of Nancy, Pediatric Onco-Hematology, Vandoeuvre-les-Nancy, France; ² University Hospital of Nancy, Hematology, Vandoeuvre-les-Nancy, France; ³ University Hospital of Nancy, Plateforme d'Aide à la Recherche Clinique (PARC), Vandoeuvre-les-Nancy, France; ⁴ University Hospital Robert Debré, Hemato-Immunology, Paris, France; ⁵ Paoli-Calmette University Institute, Hematology, Marseille, France; ⁶ University Hospital La Timone, Pediatric Onco-Hematology, Marseille, France; ⁷ University Hospital of Lille, Claude Huriez Hospital, Hematology, Lille, France; ⁸ Clermont-Ferrand University Hospital (Estaing Hospital), Hematology, Clermont Ferrand, France; ⁹ Nantes University Hospital, Hematology, Nantes, France; ¹⁰ Saint Antoine University Hospital, Hematology, Paris, France; ¹¹ Bordeaux University Hospital (Haut-Lévèque Hospital), Hematology, Pessac, France; ¹² Centre de Lutte contre le Cancer Régional Léon Bérard, Lyon, France; ¹³ IHOP University Hospital, Pediatric Hematology, Lyon, France; ¹⁴ American University of Beirut Medical Center, Beirut, Lebanon; ¹⁵ Geneva's University Hospitals, Hematology, Stem Cell Transplant Unit, Geneva, Switzerland; ¹⁶ University Hospital of Liège, Hematology, Liege, Belgium; ¹⁷ La Pitié Salpétrière University Hospital (AP-HP), Hematology, Paris, France; ¹⁸ Saint Louis Lariboisière University Hospital, Hematology, Paris, France

Background: *The 2 first authors contributed equally

Data on the outcome of AML in adolescents and young adults (AyAs) after allogeneic HSCT are scarce. In this retrospective study from the SFGM-TC, we analyzed the outcome of AML patients classified in 3 groups according to the age at transplantation: children (< 15 years), AyAs (15-25 years) and adults (26-40 years).

Methods: Patients aged between 0 and 40 years who received a first allogeneic HSCT for AML between 2005 and 2015, reported in the SFGM-TC registry, were included.

Results: This study included 2240 patients from 43 centers, of whom 481 children (21.5%), 545 AyAs (24.3%) and 1214 adults (54.2%).

According to the 2016 WHO classification, children had the highest rate (31.8%) of adverse risk disease, followed by AyAs (25.2%) and adults (21.3%) (p=0.0001). Extramedullary disease was observed in 42.8%, 24% and 21.4 % of children, AyAs and adults (p=0.0001). Among patients in complete remission (CR), younger age was associated more frequently with advanced disease (29.9% ≥ CR2 for children, 23% for AyAs and 18.2% for adults) (p=0.0001).

Donors were matched siblings in 36.4% of children, 36.7% of AyAs and 40.3% of adults, and matched unrelated in 29.1%, 33.4% and 35.7% respectively. Children had more mismatched unrelated donors (31.6%), compared to AyAs (25.3%), and adults (19.5%) (p=0.0001). Haploidentical HSCT counted for less than 4% of transplantations in each group.

Stem cell source diverged between the 3 groups (p=0.0001): bone marrow (BM) was the main source for children (60.9 % of grafts) followed by cord blood (CB) (28.9%), whereas peripheral blood stem cells (PBSC) was the main source for adults (60 %), followed by BM (29.7%) and CB (10.2%). AyAs were transplanted with BM in 44.4% of cases, PBSC in 40.4% or CB in 15.2%.

The intensity of the conditioning was myeloablative (MAC) in 93% of children, 79% of AYAs and 77.7% of adults, while a reduced-intensity regimen was used in 4% of children, 12.3% of AyAs and 14.1 % of adults (p=0.0001) at a median time of 150, 166 and 170 days respectively after diagnosis.

With a median follow-up of 7 years, 5y-overall survival (OS) was higher in children (64%), compared to AyAs (54%) and adults (52%) (p=0.001). One-year cumulative incidence (CI) of relapse was comparable in the 3 groups: 27 % for children, 33% for AYAs and 28% for adults (p=0.19). There was no difference in the CI of grade II-IV acute GvHD at day +100 between the 3 groups: 37.7%, 35.8% and 35% respectively (p=0.58). By contrast, one-year CI of chronic GvHD was higher in adults (37%) and AyAs (31%) compared to children (17%) (p=0.0001).

In multivariate analysis, independent factors associated with better OS were younger age (HR =1.37, 95% CI_1.07-1.75, p=0.013), BM (HR =1.46, 95% CI_1. 5-1.87, p=0.002), MAC (HR =1.98, 95% CI_1.5-2.61, p=0.0001) and low-risk cytogenetics (HR =1.93, 95% CI_1.53-2.45, p= 0.0001).

Conclusions: Children have a better OS than AyAs and adults. The other factors associated with a better outcome were MAC regimen, absence of poor cytogenetics, and BM as stem cells source.

Conflict of interest: We have to declare no conflict of interest

[P011 Figure]

$$$$

P012

Abstract previously published

P013

Characterization of Veno-Occlusive Disease (VOD) in Adult Patients With Acute Myeloid Leukemia (AML) Receiving Gemtuzumab Ozogamicin (GO) Before Allogeneic Stem Cell Transplant (SCT)

Vincent T. Ho ¹ , Andrew St. Martin ² , Waleska Perez ² , Patricia Steinert ² , Mei-Jie Zhang ² , Deborah Chirnomas ³ , Caroline J. Hoang ³ , Fausto R. Loberiza, Jr ³ , Wael Saber ²

¹ Dana-Farber Cancer Institute, Boston, MA, United States; ² Center for International Blood & Marrow Transplant Research, Milwaukee, WI, United States; ³ Pfizer Inc, New York, NY, United States

Background: Hepatic VOD has been reported in AML patients exposed to the CD33-directed antibody-drug conjugate GO before SCT.

Methods: To determine VOD risk and outcomes after GO exposure and subsequent SCT, we analyzed data from a subset of patients randomly selected for research level reporting to the Center for International Blood & Marrow Transplant Research (CIBMTR). Adults with AML and GO exposure at any time before first myeloablative allogeneic SCT were matched 1:4 by age at SCT and disease status before SCT to patients without GO exposure. Marginal Cox regression and marginal logistic regression modeling were used to determine overall survival and incidence of VOD at 100-days, respectively. A stepwise model building approach was used to identify risk factors associated with VOD and death.

Results: 141 patients with and 564 patients without GO exposure underwent SCT between 2008 and 2011. Median (range) age for patients with and without GO exposure before SCT was 42 (18-73) years and 38 (18-74) years, respectively. Myeloablative conditioning (MAC) with chemotherapy (with GO, 61%; without GO, 57%) was utilized more frequently than MAC with total body irradiation (with, 39%; without, 43%). Incidence of VOD at 100-days between patients with/without GO exposure was similar (4%/3%), as was the incidence of severe VOD (3%/1%). Survival probabilities were similar in patients with/without GO exposure at 100 days (81%/81%), 6 months (67%/69%), and 1 year (52%/58%) from SCT and 100-days (50%/43%), 6-months (33%/30%), and 1-year (33%/26%) from onset of VOD. In multivariate analysis (MVA), GO exposure was not associated with an increased risk of VOD (OR, 1.05; 95% CI, 0.41-2.67) or death (HR, 1.08; 95% CI, 0.86-1.36).

Conclusions: The incidence of VOD was similar in AML patients with and without GO exposure before SCT. Moreover, GO exposure was not associated with an increased risk of VOD or death. Limitations of the study include selection bias associated with the retrospective design. Future analyses will evaluate the impact of dose on VOD risk and survival outcomes in this cohort.

Encore of BMT Tandem 2018. Funding: Pfizer. Data from the Coordinating Center of the CIBMTR are preliminary.

Conflict of interest:

V. Ho: advisory boards and consultant for Jazz Pharmaceuticals

D. Chirnomas, C. Hoang, F. Loberiza: employees of and own stocks in Pfizer Inc.

A. Martin, W. Perez, P. Steinert, M. Zhang, and W. Saber: nothing to disclose

P014

Chemotherapy, donor lymphocyte infusions or second allogeneic stem cell transplantation (SCT) for post-allograft relapsed acute myeloid leukemia (AML) in adults

Matthias Eder, Michael Stadler, Elke Dammann, Catherina Lueck, Victoria Panagiota, Lothar Hambach, Souheila Tayeb, Maleen Beck, Munir Rababah, David Dorn, Christian Koenecke, Steve Ehrlich, Arnold Ganser, Gernot Beutel

Hannover Medical School, Dept. of Hematology, Hemostasis, Oncology, and Stem Cell Transplantation, Hannover, Germany

Background: Relapse rates of acute myeloid leukemia (AML) after allogeneic stem cell transplantation (SCT) remain high with limited treatment options and low survival rates. We performed a retrospective single center analysis of adult AML patients with post- allograft relapse treated with (1) chemotherapy only without cellular therapy (CHX), (2) donor lymphocyte infusions (DLI) or (3) second SCT (2.TX).

Methods: One hundred seventy five AML patients transplanted between 01/1994 and 08/2017 with systemic (n=123), both systemic and extramedullary (EM) (n=31), and EM only (n=21) post-allograft relapse were included in this analysis. Patients were treated with chemotherapy +/- irradiation without further cellular therapy (CHX, n=61), with DLI (n=90) alone (n=12) or after chemo/radiotherapy (n=78), or received a second SCT with (n=15) or without previous DLI (n=9) from the same (n=8) or an alternative (n=16) donor (2.TX). Chemotherapies included conventional cytarabine-based regimens +/- targeted therapies or hypomethylating agents. Treatment groups were compared for age, de novo or secondary AML, EM manifestation, CR rate at SCT, donor type, myeloablative (MAC) or reduced intensity conditioning (RIC) at initial SCT, and kind of chemotherapy for post-allograft relapse. With a median follow-up of 0.5 years (range 0-21) we analyzed overall survival (OS) and non relapse mortality (NRM) after first post-allograft relapse, as well as acute and chronic GvHD incidence after start of cellular therapy for the DLI and 2.TX groups using Kaplan-Meier statistics. Variables with impact on OS were identified by Cox regression analysis.

Results: Five year OS was significantly lower in the CHX group (2.0%) as compared to the DLI (11.5%) and 2.TX (21.3%) groups, respectively. The CHX, DLI and 2.TX groups were different for age (mean 50, 49, and 39 years, p=0.001), rate of 1. CR (42.6%, 40% and 75%, p=0.008), rate of matched related donors (39.3%, 20.0% and 58.3%, p=0.002) and rate of MAC (37.7%, 34.5% and 66.7%, p=0.005) all at initial SCT as well as use of hypomethylating agents for therapy at relapse (18%, 24.4%, and 4.2%, p=0.002). In contrast, EM manifestation (41%, 27.8%, and 33.3%, p=0.29) and de novo AML (77%, 67.8%, and 70.8%, p= 0.46) were equally distributed. OS was highest in the 2.TX group (p< 0.001), whereas NRM was not different between the groups (p=0.321) (Figure 1). Incidence of aGvHD (III-IV°) was 18.9% and 8.3% (p=0.198) and of extensive cGvHD 22.2% and 8.3% (p=0.035) in the DLI and 2.TX groups, respectively. In multivariate analysis only MAC as compared to RIC for first SCT (p=0.028) but not age, CR rate at initial SCT, donortype, EM manifestation, de novo AML, and kind of chemotherapy after SCT was relevant for OS after post-allograft relapse.

Conclusions: Our data show improved survival for adult patients with post-allograft relapsed AML treated by second SCT. Obviously, candidates for second SCT are a highly selected group with favorable characteristics. In contrast, patients treated with chemotherapy only have a very poor outcome. With the limitations of a single-center retrospective analysis these data suggest that second SCT may be considered for highly selected adult AML patients with post-allograft relapse.

Conflict of interest: The authors have no conflict of interest to disclose.

[P014 Figure]

$$$$

P015

Comparison of autologous stem cell transplantation versus haploidentical donor stem cell transplantation for favorable- and intermediate-risk acute myeloid leukemia patients in first complete remission

Jia Chen ^1,2,3 , Lingyi Yang ⁴ , Yi Fan ^1,2 , Yang Xu ^1,2,3 , Yue Han ^1,2,3 , Xiaowen Tang ^1,2,3 , Huiying Qiu ^1,2 , Chengcheng Fu ^1,2 , Miao Miao ^1,2 , Feng Chen ^1,2 , Depei Wu ^1,2,3

¹ the First Affiliated Hospital of Soochow University, Hematology, Suzhou, China; ² Jiangsu Institute of Hematology, Suzou, China; ³ Soochow University, Institute of Blood and Marrow Transplantation, Suzhou, China; ⁴ the First Affiliated Hospital of Soochow University, Respiratory, Suzhou, China

Background: Stem cell transplantation (SCT) is an attractive post-remission treatment option for patients with intermediate-risk acute myeloid leukemia (AML) and for some favorable-risk AML patients with additional non-genetic risk factors. Autologous SCT (auto-SCT) and haplo-identical donor SCT (haplo-SCT) are the widely used alternatives in case of a lack of a human leukocyte antigen (HLA)-matched donor. However, limited data have been published on the direct comparison between these two transplant types.

Methods: Based on the transplantation database in our center, we conducted a retrospective study involving patients with favorable- and intermediate-risk AML in first complete remission (CR1), according to the NCCN guideline. Patients with APL, extramedullary disease or those achieving CR by more than two cycles were excluded.

Results: In total, 196 patients were included in the study, 88 of whom underwent auto-SCT and 107 underwent haplo-SCT. Patients in the auto-SCT group were older than those in the haplo-SCT group, with significant difference (P = 0.004). According to NCCN guidelines version 1.2017, 58 patients (65.9%) in the auto-SCT group were classified as patients with favorable-risk AML, while only 30.8% of patients in the haplo-SCT group were classified as patients with favorable-risk AML (P < 0.001). Core binding factor (CBF) AML presented more frequently in the auto-SCT group (P = 0.005). Both the interval from diagnosis to transplantation and from remission to transplantation were longer for patients in the auto-SCT group (P < 0.001).

In the whole cohort analyses, the impact of high relapse incidence in the auto-SCT group was compensated by low non-relapse mortality (NRM), which resulted in a comparable overall survival (OS) (79.0 ± 4.6% versus 80.1 ± 5.0%, P = 0.769) and relapse-free survival (RFS) (66.1 ± 5.2% versus 77.4 ± 4.8%, P = 0.079) compared to those observed in the haplo-SCT group.

However, for patients with intermediate-risk AML, NRM was similar between the groups, and haplo-SCT exhibited superior survival. In case of post-SCT relapse, patients with intermediate-risk AML showed markedly inferior 3-year OS compared to that shown by patients with favorable-risk AML (23.3 ± 9.8% versus 60.8 ± 14.3%, P = 0.011). In the multivariate analyses, minimal residual disease (MRD) measured by flow cytometry, and gene mutation status before transplantation were independent predictors for both OS and RFS.

Conclusions: We concluded that both auto-SCT and haplo-SCT were acceptable options for post-remission treatment of patients with favorable- and intermediate-risk AML. Haplo-SCT yielded a better outcome in patients with intermediate-risk AML, but the relapse after SCT still led to a poor outcome. Clearance of MRD before SCT could improve the prognosis after transplantation.

Conflict of interest: None of the authors has anything to disclose.

		Whole cohort			Favorable-risk patients			Intermediate-risk patients
	Auto-SCT	Haplo-SCT	P value	Auto-SCT	Haplo-SCT	P value	Auto-SCT	Haplo-SCT	P value
3y-NRM	5.3±2.6%	14.1±4.1%	0.032	2.3±2.2%	6.9±4.7%	0.015	12.1±6.6%	16.3±5.0%	0.647
3y-CIR	31.0±5.1%	9.2±3.2%	<0.001	25.0±5.8%	4.3±4.3%	0.037	41.7±9.3%	10.9±4.0%	<0.001
3y-OS	79.0±4.6%	80.1±5.0%	0.769	88.3±5.2%	93.1±4.7%	0.318	56.1±9.2%	76.7±5.9%	0.018
3y-RFS	66.1±5.2%	77.4±4.8%	0.079	75.0±5.8%	89.0±6.0%	0.448	50.0±9.1%	73.8±5.8%	0.006

[ [P015 Table] Comparisons of transplantation outcomes]

[P015 Figure]

$$$$

P016

Comparison of TBI- and busulfan in the occurrence of long term side effects after HSCT in pediatric acute leukaemia: a report from the AIEOP Registry

Paola Quarello ¹ , Franca Fagioli ¹ , Franco Locatelli ² , Claudio Favre ³ , Maura Faraci ⁴ , Chiara Messina ⁵ , Arcangelo Prete ⁶ , Mimmo Ripaldi ⁷ , Attilio Rovelli ⁸ , Marco Zecca ⁹

¹ Pediatric Onco-Hematology and Stem Cell Transplantation Division, Torino, Italy; ² Department of Pediatric Hematology-Oncology, Istituto di Ricovero e Cura a Carattere Scientifico (IRCCS), Bambino Gesù Children's Hospital, Roma, Italy; ³ Department of Pediatric Oncology, Hematology, and Transplants, Anna Meyer Children's Hospital, Firenze, Italy; ⁴ Department of Pediatric Hematology Oncology, G. Gaslini Institute, Genova, Italy; ⁵ Department of Pediatric Hematology and Oncology, University of Padova, Padova, Italy; ⁶ Paediatric Oncology and Haematology Unit “Lalla Seragnoli”, Department of Paediatrics, University of Bologna Sant'Orsola-Malpighi Hospital, Bologna, Italy; ⁷ BMT Unit, Department of Pediatric Hemato-Oncology, Santobono-Pausilipon Hospital, Napoli, Italy; ⁸ Department of Pediatric Hematology, San Gerardo Hospital, Monza, Italy; ⁹ Department of Pediatric Onco-Hematology, IRCCS, Policlinico San Matteo Foundation, Pavia, Italy

Background: The outcome of hematopoietic stem cell transplant (HSCT) as treatment of pediatric patients with high-risk acute leukemia is increased in the last decades. However, this group of patients requires a long term follow-up due to late complications and second malignancies. An increased morbidity has often been attributed to the late toxicity of myeloablative conditioning regimens that include either Total Body Irradiation (TBI) or Busulfan (BU) in the majority of pediatric patients (Cohen, BMT 2008, Faraci, BMT 2008). Recently, Bernard et al. showed that late events seem less frequent after BU when compared with TBI (Bernard, BMT 2014).

The purpose of our retrospective multicenter study was to compare the long-term impact of BU or TBI on occurrence of late complications and second malignancies in childhood acute leukemia treated with HSCT.

Methods: We retrospectively evaluated the incidence of late effects in 646 pediatric patients (median age 10 years) undergoing HSCT in a AIEOP transplant Centre for acute lymphoid or myeloid leukemia (ALL or AML) between 2000 and 2013. All patients received TBI- or BU-based conditioning regimen and underwent HSCT in complete morphological remission.

Late effects incidences were calculated as a cumulative incidence to adjust the analysis for competing risks. Association between categorical variables were assessed with Fisher's exact test. P values < 0.05 were considered to be statistically significant. Statistical analysis was performed using NCSS (Hintze, 2001; NCSS PASS, Number Crunched Statistical System, Kaysville, UT, USA).

Results: In total, 472 patients (73%) had received TBI (95% ALL and 5% AML), when 174 (27%) had a BU-based conditioning regimen (80% AML and 20% ALL). Median follow up from HSCT was 7 years. The 38% (247/646) of patients developed at least one long-term late effects and 14% (88/646) of patients showed more than two complications. The incidence of at least one late complication was higher in TBI patients than in BU patients (48% vs 28%, p=0.001). Similarly, the occurrence of more than two late effects was higher in the TBI group (16% vs 7%, p=0.006).

Specifically, univariate analyses revealed that adverse effects on height and cataract were significantly more frequent after a TBI-based conditioning regimen (12% vs 7%, p=0.04, 19% vs 2%, p< 0.0001) while no significant differences were observed in occurrence of gonadal dysfunction (33% vs 20%), hypothyroidism (17% vs 12%) and pulmonary function impairment (6% vs 6%) between TBI- and BU- group. A significant higher incidence of second tumors was observed in TBI group (19% vs 0%, p=0.02).

Conclusions: Our retrospective analysis in a large cohort of patients suggests that late complications seem more frequent after TBI-based conditioning regimen in comparison with BU-based regimen and that the pattern of late side effects is different between these two groups. These data are important both in the choice of HSCT conditioning regimen and in the development of follow up measures. Thus, prospective and randomized studies aimed at better understanding the role of conditioning regimens on occurrence of different late effects are warranted.

Clinical Trial Registry: NA

Conflict of interest: The authors have no commercial, proprietary or financial interests in the products or companies described in this article.

P017

Dasatinib-induced Cytomegalovirus Reactivation Induces Earlier Molecular Response in Patients with Philadelphia Chromosome-Positive Acute Lymphoblastic Leukemia without Improvement of Transplant Outcome

Masahide Osaki, Motohito Okabe, Yuka Kawaguchi, Yoonha Lee, Marie Ohbiki, Miyo Yoshino, Hiroaki Araie, Seara Ikeno, Takahiko Sato, Takanobu Morishita, Yukiyasu Ozawa, Koichi Miyamura

Japanese Red Cross Nagoya Daiichi Hospital, Hematology, Nagoya, Japan

Background: Cytomegalovirus (CMV) infection during dasatinib therapy is reported to be correlated with a better prognosis in patients with chronic myeloid leukemia. However, this observation in patients with Philadelphia chromosome-positive acute lymphoblastic leukemia (Ph+ ALL) who need hematopoietic stem cell transplantation is not yet clarified.

Methods: We retrospectively assessed 46 patients who were newly diagnosed with Ph+ ALL between 2008 and 2017 in our hospital. All patients received TKI combined chemotherapies (JALSG Ph+ALL208 (imatinib-based therapy) and JALSG Ph+ALL213 (dasatinib-based therapy)), respectively depends on the period). All protocol was composed of induction phase and 4 consolidation phases. We define early molecular complete remission (eCMR) as one that entered CMR after the first consolidation. According to the CMV infection status (seropositive or not) at diagnosis and reactivation/infection during therapies, we divided patients into 4 groups: CMV seropositive and reactivated group (+/+) (n=5), not reactivated but seropositive group (+/-) (n=33) and CMV seronegative and no infection (group (-/-) (n=2)), CMV status was not tested (unknown group, n=6).

Results: Median age at diagnosis was 42 years. Eighteen patients were treated with dasatinib-based therapy and 28 patients were treated with imatinib-based therapy.

Neither of 2 seronegative patients developed CMV infection; group (-/-). Five of 16 seropositive patients who received dasatinib-based therapy developed CMV reactivation (CMV antigenemia in 3 and colitis in 2). On the other hand, none of 21 patients who were seropositive during imatinib treatment (31% vs 0%, p=0.01).

We analyzed the time of achieving CMR among patients treated with dasatinib-based therapy. All patients with CMV reactivation (group (+/+)) achieved eCMR. Seven of 11 patients with infected CMV without reactivation (group (+/-)) during dasatinib-based therapy achieved eCMR. Neither of 2 patients never infected with CMV (group(-/-)) achieved eCMR. Group(+/+) tends to advance CMR in comparison with group(-/-) (p=0.0668, Fisher's exact test) among patients who received dasatinib-based therapy.

34 patients received allogeneic hematopoietic stem cell transplantation. Overall survival at 2 years was not different between the patients with eCMR and patients without eCMR (76.2% vs 81.8%p=0.903). DFS and NRM at 2 years were 76.5 vs 72.7 (p=0.636) and 14.2 vs 18.2 (p=0.661), respectively. There is no difference between dasatinib patients and imatinib patients about CMV reactivation after transplantation (50% vs 65%, p=0.487). A patient of group (+/-) died of CMV colitis after allogeneic hematopoietic stem cell transplantation.

Conclusions: We have demonstrated that CMV reactivation can achieve CMR rapidly exclusively among dasatinib-based therapy patients. Some study reported that earlier CMR induced better OS after allogeneic hematopoietic stem cell transplantation. But our study found that eCMR did not yield a favorable prognosis. The CMV reactivation group can develop CMV infection after HSCT and that may reduce overall survival.

Conflict of interest: K. Miyamura: Novartis, Bristol-Myers Squibb

[P017 Figure]

$$$$

P018

Decitabine plus DLI is effective for TP 53 Mutated Patients with Relapse after Allogeneic Stem Cell Transplantation

Christine Wolschke, Raissa Adjalle, Francis Ayuk, Ute-Marie von Pein, Gaby Zeck, Nikolaus Kröger

University Hospital Hamburg Eppendorf, Clinic for Stem cell Transplantation, Hamburg, Germany

Background: Patients with AML or MDS harbouring TP53 mutation have a risk of up to 60% for relapse after allogeneic stem cell transplantation (alloHSCT). Relapse after alloHSCT have a poor outcome, resulting in overall survival of 10-20%. Remission appears to rely on therapies that maximise the graft versus leukemia effect. It is unclear whether chemotherapy is a prerequisite to donor cell therapy. Patient with myeloid neoplasm and TP 53 mutation have a poor remission rate after intensive chemotherapy. More recent data suggest efficacy of decitabine in TP53 mutated AML/MDS patients.

Here we describe 7 AML/MDS pts with TP53 mutation who relapse after alloHSCT and received decitabine (DAC) in combination with donor lymphocyte transfusion (DLI) as salvage therapy.

Methods: We retrospectively analyzed data of 7 TP53 mutated pts. with acute myeloid leukemia (n=3) or MDS (n=4) who relapsed after alloHSCT and were treated with a median of 4 cycles DAC (range 1 to 9). 6 pts. received a dose of 20 mg/m² for 5 days every 28 days, 1 pt. 20 mg/m² for 3 days every 21 days. DAC was the first salvage therapy in 5 pts. after allogeneic stem cell transplantation. 2 pts failed after a therapy with azacytidine. In 5 pts donor lymphocyte transfusion was administrated in addition to DAC. The median of DLI was 2 per patient (range 1 to 2) with a median T cell dosage of 5 x 10⁵ to 5 x 10⁶ CD3+ cells/kg pt. 1 pat. also received lenalidomide in combination.

Results: As a result, overall response rate was 71%, including 3 complete cytogenetic remission (cCR, 42%) and 2 partial remission (PR, 29%). 2 patients within the first line group achieved CR, 1 patient received DAC as second line treatment after previous Aza failure reached CR. Median duration of CR was 6 months (range 5 to 7) and 2 patients relapsed so far. Incidence of acute or chronic graft-versus-host disease was low.

Conclusions: DAC in combination with exerts clinical efficacy in patients with AML or MDS with TP 53 Mutation and is able to induce complete remission in nearly half of the patients. These data suggest an alternative relapse regime in combination with DLI after an allogeneic stem cell transplantation for patients with TP 53 mutation. The role of decitabine as pre-transplant and maintenance strategy post transplantation for patients with high risk AML or MDS and TP53 Mutation should be evaluate in prospective clinical trials.

Conflict of interest: C. Wolschke: nothing to disclose

P019

Abstract previously published

P020

Disease Progression is Main Barrier to Allogeneic Hematopoietic Stem Cell Transplantation (HCT) in Patients with Newly Diagnosed and Relapsed Acute Leukemia

Elaina V. Preston ¹ , Stephanie Chinapen ¹ , Taylor Borrill ¹ , Eric Davis ² , Brian C. Shaffer ^1,3 , Roni Tamari ^1,3 , Ann A. Jakubowski ^1,3 , Miguel-Angel Perales ^1,3 , Doris M. Ponce ^1,3 , Gunjan Shah ^1,3 , Parastoo B. Dahi ^1,3 , Juliet N. Barker ^1,3 , Ellin Berman ^3,4 , Martin Tallman ^3,4 , Sergio A. Giralt ^1,3 , Eytan Stein ^3,4 , Boglarka Gyurkocza ^1,3

¹ Memorial Sloan Kettering Cancer Center, Dept. of Medicine, Adult Bone Marrow Transplant Service, New York, NY, United States; ² Memorial Sloan Kettering Cancer Center, Dept. of Pediatrics, Bone Marrow Transplant Service, New York, NY, United States; ³ Weill Cornell Medical College, Dept. of Medicine, New York, NY, United States; ⁴ Memorial Sloan Kettering Cancer Center, Dept. of Medicine, Leukemia Service, New York, NY, United States

Background: Multiple studies demonstrated the curative effect of HCT for patients with acute leukemia (AL), but it is estimated that only 1 of 3 eligible patients undergo this procedure. Historically, lack of available donor was among the most common reasons for not proceeding to allogeneic HCT, but this is no longer a major limitation to transplant with widespread use of cord blood and haploidentical HCT.

Our aim is to prospectively determine the rates in which AL patients proceed to HCT based on disease risk using National Comprehensive Cancer Network guidelines and to identify reasons for not proceeding to HCT in the era of alternative donor sources.

Methods: All patients with newly diagnosed or relapsed AL admitted to our center's Adult Leukemia Service for induction or re-induction therapy were approached to enroll on this prospective observational study.

Results: Between April 2016 and December 2017, 90 patients were enrolled and have sufficient follow up (180 days from initiation of induction or re-induction).

While 52% of patients identified as White European, there was substantial representation from groups at high risk of not finding an 8/8 HLA-matched donor in the international registries.

Twenty-two patients had HLA-identical sibling donors. Fifty-three patients had at least one potential 8/8 HLA-matched unrelated donor identified through preliminary donor searches. In 4 cases suitable umbilical cord units or HLA-haploidentical donors were identified. The remaining 11 pts did not have searches due to early death or known matched related donor.

Allogeneic HCT was recommended in 71 cases (78%) based on disease-related risk via a consensus decision of Leukemia and Bone Marrow Transplant Services. Of the 71 patients, 46 (65%) underwent HCT. The median time between starting induction or re-induction chemotherapy and transplant was 109 (range 41-545) days. Donors were HLA-identical siblings (n=10), HLA-matched unrelated donors (n=27), mismatched unrelated donor (n=1), and umbilical cords (n=8).

Of the 25 patients for whom HCT was recommended but was not performed, 16 had progressive disease (11 have since died), 1 patient died of induction-related complications, 3 had severe comorbidities, 4 patients declined HCT and 1 patient was lost to follow up. There weren't any cases where donor availability was the barrier to transplant.

Conclusions: Based on interim results of our prospective observational trial 65% of patients underwent allogeneic HCT in whom it was recommended based on disease-related risk factors. A suitable donor was identified for all patients in whom allogeneic HCT was recommended. The main barrier to transplant was early disease progression and death.

Conflict of interest: None of the authors has anything to disclose.

Male Patients	53	59%
Female Patients	37	41%
Median Age	58
Age Range	24-74
DISEASE STATUS AT DIAGNOSIS
Newly Diagnosed	66	73%
RES	7	8%
First Relapse	15	17%
Relapsed Refractory	2	2%

[ [P020 Table] Patient Characteristics]

[P020 Figure]

$$$$

P021

Donor lymphocyte infusion after salvage chemotherapy in relapsed acute myeloblastic leukemia patients after allogeneic stem cell transplantation

Atilla Uslu, Guldane Cengiz Seval, Sinem Civriz Bozdag, Gunhan Gurman, Taner Demirer, Onder Arslan, Muhit Ozcan, Selami Kocak Toprak, Pervin Topcuoglu, Osman Ilhan

Ankara University School of Medicine, Department of Hematology and Stem Cell Transplant Unit, Ankara, Turkey

Background: Although allogeneic HSCT(AHSCT) can offer cure to acute myeloblastic leukemia (AML) patients, relapse still remains to be a limiting factor. The optimal treatment to improve the outcome of this poor prognostic group is controversial. We aimed to search the impact of donor lymphocyte infusion (DLI) after salvage treatment in relapsed AML patients after allogeneic HSCT.

Methods: We retrospectively included twenty AML patients who had hematological relapse after AHSCT, in Ankara University Faculty of Medicine Bone Marrow Transplantation Unit. Donor lymphocyte infusions have been scheduled at the end of salvage treatments. We statistically analyzed the relationship between response to DLI and overall remission duration with patient/ DLI characteristics.

Results: Demographic features of the patients have been summarized in Table 1. All but two patients have been treated with salvage chemotherapy. Donor lymphocyte infusion has been performed at the end of chemotherapy. Two patients have been treated with azacytidine and DLI has been given at the end of 2 cycles. DLI features of the patients have been shown in Table 2. Eight patients recieved DLI in complete remission whereas 5 out of 12 patients achieved remission after DLI. Response duration was not found to be related with gender, CMV status, cytogenetic risk factor, blood percent at relapse, remission duration before relapse, pretransplant acute or chronic GVHD before relapse. Also, donor age, gender, sex mismatch or blood group mismatch with recipient did not have a statistically significant effect on response duration after DLI. The only statistically significant factor was remission status before donor lymphocyte infusion. Acute GVHD was observed in 32% of the patients. Median survival in patients who achieved remission and refractory disease after DLI treatment were 2.2 years and 1.4 months respectively (p=0.93) Estimated overall survivals at 1 year who achieved remission were 60% respectively.

Conclusions: In AML patients who relapsed after AHSCT, donor lymphocyte infusion after salvage chemotherapy is a feasible treatment choice if remission can be achieved before infusion.

Conflict of interest: All authors: nothing to disclose

[P021 Figure]

$$$$

Disease status at DLI (n (%)) Remission Active disease	8 (40%) 12 (60%)
Median Infused CD3+ cell count (x108)(median)	1,38 (0,27-11,60)
Total DLI number(n (%)) 1 2 3 4	9(52,0%) 4(20,0%) 6(24,0%) 1 (4,0%)

[ [P021 Table] Table 2: DLI features of the patients]

P022

Early molecular dagnosis and preventive therapy of expected cytological post-transplant relapses in patients with acute myeloid leukemia

Yana Gudozhnikova, Nikolay Mamaev, Sergey Bondarenko, Ildar Barkhatov, Valeriya Katerina, Alena Shakirova, Tatiana Gindina, Valentina Kravtsova

Pavlov First St. Petersburg State Medical University, R.M. Gorbacheva Memorial Institute of Children Oncology, Hematology and Transpantation, St. Petersburg, Russian Federation

Background: Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is now considered the best treatment of AML with poor prognosis, however, often followed by post-transplant relapses (PTR). Among PTR, molecular and cytological variants (mPTR and cPTR, respectively) were recognized, and WT1 gene was an important independent prognostic indicator of acute leukemia following allo-HSCT.

Methods: Ninty AML patients aged 2-68 years (median - 38.8) who underwent allo-HSCT at our University were enrolled in the study. WT1 gene expression levels by real-time PCR as well as bone marrow blast counts were measured before transplantation as well as on days D30+, D60+, and D100+ after HSCT. The WT1 threshold value was determined as 250 WT1 copies/10⁴ copies of BCR gene. Overall survival (OS), relapse-free survival (RFS), and cumulative incidence of relapse (CIR) were determined for the WT1+ and WT1- patient groups. Besides, we have allocated time intervals between mPTR and cPTR in about one-third of patients.

Results: The levels of WT1 gene expression at allo-HSCT ranged from 0 to 56884 WT1/10⁴ ABL copies (median - 209). There was a significant difference in 2-year OS, RFS, and CIR between WT1-positive and WT1-negative groups (34.1%, 24.4%, 43.9% vs 74.5%, 66%, and 72.3%, at p< 0.0001, p< 0.0001 and p=0.003, respectively). On the day D30+, the 250-copy threshold recognized difference only for CIR (46.2%, 38.5%, 38.5% vs 60.9%, 50%, and 59.4% at p=0.18, p=0.17 and p=0.04, respectively). Meanwhile, the differences in these parameters between WT1- and WT1+ groups were maximal at the time points of D60+ (30.8%, 30.8% and 30.8% vs 66.1%, 53.2% and 61.3%, at p=0.005, р=0.009 and 0.006, respectively), and D100+ (25%, 15% and 20% vs 74.5%, 63.6% and 69.1%, at p< 0.0001, р< 0.0001 and р< 0.0001, respectively). In general, these data showed a prognostic significance of WT1 level normalization measured on D0, as well as on days D60+ and D100+ after allo-HSCT. In our cases, mPTR and cPTR were registered in 57/86 (66.2%) and 33/86 (38.4%) patients, respectively. In 33.3% of the patients (11/33), mPTR and cPTR were noted simultaneously, whereas in 22 cases, cPTR occurred with a delay after mPTR. This period ranged from 13 up to 321 days with a median of 35 days. It should be mentioned that our patients, upon detection of overexpressed WT1 obtained different preventive treatments that should be kept in mind explaining the increased interval between mPTR and cPTR. However, the prolonged interval also may be due to molecular and cytogenetic features of these leukemias. New investigations on the topic are needed to test this hypothesis.

Conclusions: Thus, in AML patients, WT1 levels in bone marrow cells at time points of D0, D60+ and D100+ were predictive for OS, EFS and CIR, which agrees well with recently published data. The unusual phenomenon of cPTR delay after mPTR detection may be useful in clinical practice for the controlled molecular study of new medical drugs and therapeutic approaches.

Conflict of interest: The authors have stated that they have no conflicts of interest.

P023

Efficacy of active surveillance in reducing mortality due to multidrug resistant bacteria in acute leukemia

Christelle Castañon ¹ , Alejandro Zuazua ² , Adan Rodriguez ² , Teresa Bernal ¹ , Sara Alonso ¹ , Enrique Colado ¹

¹ Hospital Universitario Central de Asturias, Oviedo, Spain; ² Universidad de Oviedo, Oviedo, Spain

Background: Multidrug resistant infections (MRD), specially those caused by gram negative bacteria (GNB), are associated with poor outcomes in acute myeloblastic leukemia (AML) and could make hinder patients from completing their optimal treatment, such stem cell transplantation. Active surveillance (AS), where patients are screened for colonization with MDR-GNB, may serve as a tool to decrease bacterial transmission and guide the choice of empirical antibiotic treatment. The aim of the study was to analyze the impact of active surveillance in the incidence of MDR and GNB infections and infection related mortality in AML patients.

Methods: This was a prospective observational study including all consecutive adult patients with an AML diagnosis, treated with intensive chemotherapy at Hospital Universitario Central de Asturias (HUCA) from January 2015 to October 2017. The cohort of patients in whom active surveillance was performed (cohort 1) was compared to a cohort of AML patients in whom AS was not done (cohort 2). Empiric antibiotic therapy was considered appropriate if at least one of the empiric antibiotics was effective for the bacteria detected in surveillance cultures. All patients received fluoroquinolone prophylaxis. Data are presented as mean ± standard deviation or as median (range). Univariate comparisons were done using the chi-square test (categorical variables) or Fisher´s exact test (for contingency tables with frequencies below 5).

Results: From January 2015 to October 2017, seventy-three AML patients received 172 intensive chemotherapy cycles. In cohort 1, active surveillance was performed along 86 cycles. General characteristics were similar in both cohorts: median age 57 (43-64) years, 48% females, European-Leukemia-Net genetic risk favorable in 44%, intermediate in 27% and high in 29% patients. Gram negative multidrug resistant bacteria colonization was detected in 55% of the cycles, without differences between the different phases of treatment (46% during induction cycle-1, 70% during induction-2, 70%, during consolidation-1, 44% during consolidation-2 and 50% during consolidation-3). Regarding the epidemiology of bacterial infections, the more frequent isolates in both cohorts were Gram negative (45%) followed by gram positive (33%) and polimicrobial infections (15%). Infection due to multidrug resistant bacteria was detected in 50% of cycles. Empirical antibiotic treatment was considered appropriate according previous surveillance cultures in 86% of infectious episodes. Gram negative infections were statistically associated with higher mortality compared to other infections (21% vs 6.6%, p= 0.03) and the main cause of death in complete remission prior to stem cell transplant. A non-significant reduction in thirty-day mortality was observed between cohort 1 and cohort 2 (7,5% vs 12%).

Conclusions: In our population of acute leukemic patients, colonization due to GN resistant bacteria occurs in a high proportion of patients and is an early event during treatment. Gram negative bacterial infections remain the most frequent cause of morbidity and mortality. Although active surveillance may help in the choice of appropriate empiric antibiotic treatment in the era of multidrug resistance, larger studies are needed to demonstrate its efficacy in reducing mortality.

Conflict of interest: None of the authors has anything to disclose.

P024

Efficacy of allogeneic haematopoietic transplantation in older patients with myeloid malignancies assessed by Graft-versus-Host Disease Free, Relapse-Free Survival (GRFS). Multivariable Analysis from a Single Centre

Pavel Jindra, Michal Karas, Daniel Lysák, Alexandra Jungova, Katerina Steinerova, Hrabětová Marcela, Jiří Šrámek, Tereza Zábranská, Lekka Mohammadová, Veronika Bergerová

University Hospital Pilsen, Haematology & Oncology, Pilsen, Czech Republic

Background: Older patients with high risk myeloid malignancies (AML/MDS) have dismal prognosis and the only curative option is alloHCT. The decision to offer HCT for such older pts is challenging given their high rate of transplant related morbidity/mortality. Recently, a composite endpoint of GRFS was proposed to incorporate the most significant complications of allogeneic transplantation and can be used as a meaningful parameter to define transplant success. However, the main variables affecting this endpoint in this particular population are poorly defined. To address this question, we have performed single centre retrospective analysis with long-term follow-up.

Methods: 91 consecutive pts≥60 years with AML (n=82) or MDS (n=9) who received RIC alloHCT in our institution between 11/2001 and 11/2016 were analysed. GRFS events were defined as acute GVHD ≥ gr. III, chronic GVHD (requiring systemic therapy or extensive stage), relapse, or death whichever comes 1st during follow-up. Donors were matched sibling (MSD, n=25, 27%), matched unrelated donors (MUD, n= 41, 45%), mismatched MUD (n=22, 24%) or haploidentical donors (n=3%). The median age of cohort was 64 years (range 60-74), 62% were males. 32 pts (35%) had advanced disease (= >CR1/PR1) at HCT. Disease risk index (DRI) at HCT was low/intermediate in 47pts (52%) and high in 44 pts (48%). GVHD prophylaxis was uniformly based on calcineurin inhibitors and short MTX.

Results: With median follow-up for survivors of 52 months (range 7-144) a total of 54 (59%) died, 20 (22%) from relapse and 34 (37%) from NRM. Estimated OS, RFS and GRFS at 1 year were 61%, 60%, and 49 %, whereas corresponding probabilities at 5-years were 36%,32% and 24%, respectively (Figure 1). Cox regression hazard model identified high DRI as the strongest factor predicting extremely poor OS/RFS (HR 2.121, p=0.0049/HR 1.924, p=0.0123), whereas recipient-donor sex mismatch was protective for both the outcomes (OS:HR 0.513, p=0.0131/RFS: HR 0.555, p=0.0249). A donor age≥62 year was associated with adverse OS (HR 2.110, p= 0,0345) as well as (yet borderline) CMV match (HR 1.791 p=0.0392). Recipient age ≥69 years and < 2,5 CD34+ cells/kg transplanted increase NRM (HR 4.493, p=0.0068/HR 2.960, p=0.0078).

In a multivariate analysis for GRFS endpoint, high DRI retains the most significant impact (HR 2.652, 95% CI 1.562-4.563, p=0.0003), followed by donor age ≥ 62years ((HR 2.304, 95% CI 1.106-4.797, p=0.0268).

Conclusions: Our data are limited by retrospective nature and inherent selection bias. Despite this we found that with long-term follow up, significant portion of older patients with myeloid malignancies survive alloHCT without experiencing at least 1 GFRS event - 49% at 1year and 24% at 5years. In addition, we observe clear trend for plateau since 3-year time point. The negative impact of DRI and high donor age suggests that HCT should be performed early during the course of disease with the preference of young donor (regardless whether related or unrelated).

Conflict of interest: nothing to disclose

[P024 Figure]

$$$$

P025

Flag-Ida and donor lymphocyte infusion (DLI) for acute leukemia relapse after allogeneic stem cell transplantation (alloSCT) - single center experience

Jakub Radocha, Alzbeta Zavrelova, Miriam Lanska, Benjamin Visek, Jan M Horacek, Pavel Zak

University Hospital Hradec Kralove and Faculty of Medicine Hradec Kralove, Charles University, Department of Internal Medicine - Haematology, Hradec Kralove, Czech Republic

Background: Prognosis of patients with relapsed acute myeloid (AML) or lymphoblastic (ALL) leukemia after alloSCT remains dismal. The treatment options are few with short overall survival. We analyzed the outcome of patients with AML and ALL who received Flag-Ida regimen for relapse after alloSCT.

Methods: All patients with relapse of AML and ALL who received at least one cycle of Flag-Ida (Idarubicin 12 mg/m2 D1-3, Cytosinarabinoside 2000 mg/m2 D1-5 and Fludarabine 30 mg/m2 D1-5) from 2006 to 2017 were eligible for analysis. The outcome in terms of progression free and overall survival was analyzed. The impact of DLI administered after Flag-Ida was analyzed.

Results: Overall 36 patients received 42 cycles of Flag-Ida. There were 20 (55.5%) females and 16 (44.5%) males with mean age 41 years (18-64 years). 25 (69%) patients had AML and 11 (31%) ALL. 20 (55.5%) patients received Flag-Ida for 1^st relapse and 16 (44.5%) for 2^nd and further relapse. CR was reached after 34 cycles (81%), PR after 2 cycles (5%), progressive disease after 4 cycles (9%), 2 (5%) were not evaluable due to early death. 10 (28%) patients received 2^nd allogeneic transplantation from different donor. 22 (61%) patients received DLI after Flag-Ida. Median follow-up from administration of treatment was 8.9 months (1.0-90.3 months). There was no difference in PFS and OS in patients with AML and ALL (14.3 months and 29.5 months, p=0.34 for OS, 10.7 months and 12.1 months, p=0.51 for PFS). Patients who received post treatment DLI after Flag-Ida showed significantly better OS (6.7 versus 23 months, p=0.0036).

Conclusions: Flag-Ida plus DLI remain a valid clinical option for treatment of relapse after alloSCT with some patients reaching long term remission when post treatment DLI can be administred.

Conflict of interest: All authors: nothing to disclose

P026

Haploidentical Stem Cell Transplantation with Post-Transplantation Cyclophosphamide for Pediatric Acute Lymphoblastic Leukemia: A Curative Option

Satya Prakash Yadav, Sagar Nivargi, Prashant Chabra, Anil Sharma, Neha Rastogi

Medanta Medicity Hospital, Pediatric Hematology Oncology & Bone Marrow Transplant, Gurgaon, India

Background: T-cell replete haploidentical stem cell transplantation(SCT) with post-transplantation cyclophosphamide(PTCy) has shown encouraging results for the treatment of hematologic malignancies in adults. However data regarding its application for curing pediatric acute lymphoblastic leukemia(ALL) is meagre. Here we present our experience using the same in 8 children suffering from ALL.

Methods: The conditioning was Non-myeloablative in 4 (Fludarabine 160 mg/m², Cyclophosphamide 29 mg/kg, total body irradiation TBI 2Gy) and Myeloablative in 4 (Thiotepa10 mg/kg, Fludarabine 160 mg/m², Cyclophosphamide 29 mg/kg, TBI 2Gy in 3 and Fludarabine 160 mg/m², Busulfan 12.8 mg/kg in 1). All received PTCy 50 mg/kg on day 3 and 4 as graft vs. host disease(GVHD) prophylaxis along with tacrolimus and mycophenolate mofetil. After informed consent, 8 children with ALL were allografted; median age was 10 years (range 4-14); Male: Female=1:1. All were in complete remission(CR) prior to SCT: CR1-1, CR2-5 and CR3-2. Donors were mobilized with Granulocyte colony stimulating factor 10 microgram/kg daily for four days, the PBSC were collected with one large volume apheresis procedure. All the donors shared 5 out of 10 alleles with the recipient; all donors were parents except one who was a younger sibling. Only one was sex-mismatched SCT. A median of 6 million of CD34cells/kg was infused (range 5-23 million/kg).

Results: The engraftment rate was 100%, median time of neutrophil engraftment was 18 days (range 13-23) and platelet engraftment was 15 days (range 9-20). Chimerism at day +30 was fully donor in all 8 children. One child relapsed in bone marrow on day+48. He did not respond to Inotuzumab and died with progressive disease. Remaining 7 patients are alive at the median follow-up of 20 months (range 6-39). Three children relapsed at 6 months,11months and 15 months post SCT. One with isolated central nervous system (CNS) relapse achieved CR4 with chemotherapy followed by donor lymphocyte infusions (DLI) and is alive and disease free at follow up of 10 months . Escalating doses of DLI were given and dose of CD3 cells 1 million/kg induced acute GVHD grade II in this patient. Second child with isolated CNS relapse failed to respond to DLI but achieved CR4 with cisplatin 75 mg/m2 and Bevacizumab 5mg/kg once a month therapy and is alive and disease free at 4 months. Third child with isolated bone marrow relapse with blasts expressing CD20 achieved CR3 after treatment with a single dose of Rituximab 100 mg/m2 and is in CR at 15 months on maintenance Rituximab therapy. The cumulative incidence of graft versus host disease (GVHD) acute and chronic extensive was 12.5% and 36.5% respectively. Grade-I acute GVHD was seen in 1 patient. All three children with chronic graft vs. host disease are alive and disease free. Three out of 4 children receiving non-myeloablative conditioning relapsed. Overall survival is 88% and event free survival is 50% at a median follow up of 20 months.

Conclusions: Haploidentical SCT with PTCy is safe for children with ALL. Myeloablative conditioning, post transplant DLI and/or targeted therapies can help improve survival.

Conflict of interest: None of the authors has anything to disclose.

P027

High efficacy of R-HyperCVAD plus Dasatinib followed by allogeneic stem cell transplantation then maintenance with Dasatinib in adults with philadelphia postive acute lymphoblastic leukemia

Ahmad Ibrahim ¹ , Rima Moghnieh ² , Charbel Khalil ³ , Ali Youssef ⁴ , Ghada Awad ⁵ , Kamal Zahran ⁶ , Tamima Jisr ⁷ , Anas Mogharbel ⁴

¹ Makassed and Middle East Institute of Health University Hospitals-Lebanese University, Hematology Oncology-BMT, Beirut, Lebanon; ² Makassed and Middle East Institute of Health University Hospitals-Lebanese University, Internal Medicine, Beirut, Lebanon; ³ Middle East Institute of Health University Hospital-Saint Joseph University-Lebanese University, Cellular Therapy, Beirut, Lebanon; ⁴ Makassed University Hospitals, Hematology Oncology-BMT, Beirut, Lebanon; ⁵ Makassed University Hospitals, BMT, Beirut, Lebanon; ⁶ Middle East Institute of Health University Hospital, BMT, Beirut, Lebanon; ⁷ Makassed University Hospitals, Laboratory Medicine, Beirut, Lebanon

Background: Philadelphia (Ph+) chromosome acute lymphoblastic leukemia (ALL) represents approximately 25-40% of adults ALL patients (pts). Ph+ ALL is historically considered as having poor prognosis. Allogeneic hematopoietic stem cell transplantation (AHSCT) has been the gold standard therapy for maintenance of complete remission (CR) in Ph+ ALL pts. Pre and post AHSCT administration of first generation thyrosine kinase inhibitors (TKIs) Imatinib has dramatically improve the outcome of Ph+ ALL with reducing disease recurrence with a long term leukemia-free-survival reaching approximately 60%, with no evidence that Imatinib had an adverse effect on transplant-related morbidity and mortality. The usage of second generation TKIs in the same setting improved the results in lower number of pts. The role of Ponatinib, a third generation of TKIs given as bridge to AHSCT or after AHSCT is under investigation. The achievement of molecular remission before AHSCT in adults' pts with Ph+ALL had a positive impact on relapse and long term outcome after AHSCT. We present the results of our experience in adults' pts with Ph+ALL who underwent AHSCT in complete molecular response after induction therapy including Dasatinib, a second generation TKIs, and receiving a maintenance therapy after AHSCT with Dasatinib

Methods: All the pts received induction therapy according to R-HyperCVAD regimen associated to Dasatinib. AHSCT was performed from a related donor. Dasatinib was given in pre and post-transplant at the dose of 70mg/day. Evaluation of molecular response was done by qPCR, In particular every 3 to 6 months after AHSCT.

Results: From 1/2010 to 3/2017, 6 pts were included in the study. Median age was 40yo (23-43).There were 3 males and 3 females. Four pts underwent allogeneic peripheral stem cell transplantation from an identical donor with a conditioning associating total body irradiation and Cyclophosmamide in 2 and TBF (Thiotepa, Bisulfex, Fludarabine) in 2 other pts. Two pts underwent haploidentical allogeneic bone marrow transplantation from their mothers after a conditioning with TBF. The pts started Dasatinib as maintenance therapy approximately 3 months after AHSCT except one who died at 1 month after AHSCT of CMV pneumonia. Median duration of Dasatinib administration as maintenance therapy was 19 months (6-48). No modification of dose of Dasatinib was done because of no major toxicity. Five pts are alive in complete molecular response with a median of 21m (9-95).

Conclusions: AHSCT after R- HyperCVAD with Dasatinib followed by Dasatinib maintenance is highly efficient and well tolerated in pts with Ph+ ALL.

Conflict of interest: Nothing to disclose

P028

High TAL1 expression predicts a poor clinical outcome in pediatric T-cell acute lymphoblastic leukemia

Xinran Chu, Sh H

Children's Hospital of Soochow University, Suzhou, China

Background: To explore the clinical characteristics and prognosis of pediatric T-cell acute lymphoblastie leukemia(T-ALL) enrolled in protocol CCLG-2008. The present study was also designed to investigate the clinical significance of TAL1 noncoding mutation and TAL1 expression in childhood T-ALL.

Methods: From July 2008 to November 2016, 150 consecutive patients aged≤15 years were treated with CCLG-2008 in Children's Hospital of Soochow University and Department of Pediatric, Institute of Hematology and Blood Diseases Hospital. Predictive values of clinical characteristics and early treatment responses,including prednisone response, bone marrow morphology on day 15 and 33 during induction chemotherapy and week 12 before consolidation therapy were analyzed. The incidence of TAL1 noncoding mutation was analyzed by polymerase chain reaction(PCR) and sanger sequencing in 89 children. TAL1 expression was anlayzed by real-time quantitative PCR in 67 patients.

Results: The 5-year overall survival (OS) and event free survival (EFS) rates for 150 patients were 64.5 % (SE,4.6) and 58.7 % (SE, 4.5), respectively. Prednisone poor responder was strongly associated with decreased survival rates (5-year EFS rate, 47.1 % (SE,7.4)). And patients with ≥5% blast cells in bone marrow morphology at D33 tend to have lower 5-year EFS. Patients received bone marrow transplantation in high risk group showed no significant difference in overall survival rates than others who didn't (P=0.146). We have eight patients (9%) presented with heterozygous somatic mutation in noncoding elements of TAL1. All eight patients showed resistance to prednisone treatment. In TAL1 expression analysis, all TAL1 mutation patients were in high expression group(P75-P100). Cox analysis for 3-year OS identified high TAL1 expession and prednisone response as independent prognostic factors(Table 1). Survival analysis showed patients with high TAL1 expression had lower OS and EFS (Figure 1,2).

Conclusions: It is concluded that TAL1 noncoding mutation is not rare in pediatric T-ALL patients. And patients with high TAL1 expression (P75-P100) tend to have a poor clinical outcome.

Conflict of interest: The authors declared that they have no conflicts of interest to this work.We declare that we do not have any commercial or associative interest that represents a conflict of interest in connection with the work submitted.

P029

HLA-haploidentical peripheral blood stem cell transplantation following reduced-intensity conditioning (haplo RIC-PBSCT) with very low-dose ATG for refractory/relapsed pediatric acute leukemia

Kohei Higuchi, Akihisa Sawada, Mariko Shimizu, Aya Ioi, Tatsuro Nakanishi, Hiroshi Tsujimoto, Maho Sato, Masahiro Yasui, Masami Inoue

Osaka Women's and Children's Hospital, Department of Hematology/Oncology, Izumi, Osaka, Japan

Background: The prognosis of refractory/relapsed pediatric acute leukemia is dismal. To enhance graft-versus-leukemia effect, HLA-haploidentical peripheral blood stem cell transplantation has been performed in our institute. In this setting, it is difficult to determine the optimal GVHD prophylaxis.

Methods: We retrospectively reviewed 17 pediatric patients with refractory/relapsed acute leukemia who underwent HLA-haploidentical peripheral blood stem cell transplantation following reduced-intensity conditioning (haplo RIC-PBSCT) with very low-dose ATG between 2012 and 2016.

Results: Six patients had acute lymphoblastic leukemia and 11 had acute myeloid leukemia. Four patients who relapsed after allogeneic stem cell transplantation (allo-SCT) were in complete remission and 13 were on disease at haplo RIC-PBSCT. The median age at haplo RIC-PBSCT was 8 years (range, 1 to 17 years). The median follow-up duration was 7 months (range, 0 to 68 months). All patients received HLA-haploidentical graft from related donors, and received reduced-intensity conditioning consisting of fludarabine/clofarabine, melphalan, and etoposide. GVHD prophylaxis consisted of cyclosporine/tacrolimus, short-term methotrexate, methylprednisolone, and ATG 1.25mg/kg on day -2. Median follow-up for survivors was 44 months (range, 15-68 months). The 2-year overall survival (OS) rate and event free survival (EFS) rate were 49.4% and 34.3%, respectively. The 2-year cumulative transplant-related mortality (TRM) rate and relapse rate were 11.8% and 60.7%, respectively. Among 17 patients, 12 (70.5%) developed grade3-4 acute GVHD, 2 developed mild chronic GVHD, and 2 developed moderate chronic GVHD.

Conclusions: Although the sample size is small and follow-up duration is short, the 2-year OS/EFS of patients in this analysis including those who relapsed after allo-SCT was encouraging. The incidence of severe acute GVHD was high, but we could manage severe GVHD by adding therapeutic dose of immunosuppressant. For further improvement of outcome, it should be considered using combination therapy to enhance anti-leukemic effect such as molecular target therapeutic agents before and/or after transplantation.

Conflict of interest: nothing to disclose.

P030

Implementation of EBMT risk score in patients with Acute myeloid Leukemia undergoing allogeneic hematopoietic stem cell transplantation. What can we expect? - Single center experience

Lazar Chadievski, Irina Panovska, Aleksandra Pivkova Veljanovska, Zlate Stojanoski, Sonja Genadieva Stavric, Lidija Cevreska, Marija Popova Labacevska, Gazmend Amzai, Borce Georgievski

University Clinic for Hematology, Skopje, Macedonia, the Former Yugoslav Republic of

Background: Allogeneic hematopoietic stem cell transplantation (SCT) is the treatment that has the highest curative potential in treatment of patients with hematological malignancies, or to be more precise in patients with acute myeloid leukemia (AML). But as it is known to be probably the most complicated biological procedure in the medicine, it has own risks of complications, some of which can have worst outcome. Transplant related mortality (TRM) is higher in some patients depending on various factors. The European Group for Blood and Marrow Transplantation (EBMT) provides a simple and usable risk score that can be efficiently used to evaluate the risk in patients with AML undergoing allogeneic SCT.

Methods: We used the EBMT score to evaluate the pretransplantation risk in patients diagnosed with AML undergoing allogeneic SCT. We analyzed five factors, age of the patient (< 20 years 0 points, 20-40 years 1 point; over 40years 2 points), stage of the disease (early 0 points, intermediate 1 point, late 2 points), time from diagnosis (< 12 months 0 points, > 12 month 1 point), donor type (sibling 0 point, unrelated 1 point), and donor recipient gender combination (female donor male recipient 1 point, any other combination 0 points). The result could augment the risk for an individual patient with increasing score from 0 as best to 7 as worst. We made a retrospective analysis of 30 patients with AML in the last 5 years all treated with allogeneic SCT. In our group, 15 were male (50%), 15 were female patients (50%). The average age of the patients was 38.7 (18-67).

Results: In 27 (90%) patients we performed sibling allogeneic SCT, while in 3 (10%) matched unrelated donor (MUD) allogeneic SCT. The mostly used conditioning regimen was Bu-Cy (43.3%), then Bu-Cy-ATG (16.7%), Bu-Cy-Mel (10%). We used unmanipulated allografts. The median number of used stem cells was 3.41x10⁸/kg TT mononuclear cells (1.5-8.0). The average day to engraftment was 13.8 (10-20). In the majority of patients the period to allogeneic SCT was < 12 months (25 patients; 83%). Using the EBMT score we divided the patients into two groups: those with score < 2 (lower risk, favourable, 12 (40%) and ≥ 3 (higher risk, unfavourable, 18 (60%)). In the low risk group Transplant related mortality (TRM) was 16% (2 patients), while in the higher risk group 38% (7 patients), but the Non relapse mortality was NRM 33% (6 patients). The Overal Survival (OS) for 3 years in the group with score < 2 was 85%, while in the group with score ≥ 3 was 57%. But we must note that in the latter group two patients were transplanted in active disease, one of them was with haploidentical donor. Two of them died because of acute steroid refractory GvHD.

Conclusions: The EBMT score is a valuable tool to make adequate identification of patients with AML and favourable risk score that could benefit from allogeneic SCT as treatment option.

Conflict of interest: No disclosures

P031

Improvement of relapse-free survival by allogeneic stem cell transplantation in first complete remission in ELN favorable-risk adult AML patients with initial hyperleukocytosis

Feng-Ming Tien ^1,2 , Cheng-Hong Tsai ^1,2 , Hsin-An Hou ² , Jih-Luh Tang ^1,2

¹ Tai‐Cheng Stem Cell Therapy Center, National Taiwan University, Taipei, Taiwan, Republic of China; ² National Taiwan University Hopsital, Taipei, Taiwan, Republic of China

Background: Acute myeloid leukemia (AML) patients of the 2017 European Leukemia Net (ELN) favorable-risk group generally do not require allogeneic stem cell transplantation (allo-SCT) in first complete remission (CR1). However, a substantial portion of these patients still relapses and dies of disease progression. In this study, we aimed to refine the risk assessment in ELN favorable-risk patients and to identify a subgroup of patients who might benefit from allo-SCT in CR1.

Methods: The study population comprised 176 ELN favorable-risk AML patients. Hyperleukocytosis (HL), defined as initial white blood cell counts above 50,000/uL, was observed in 35.2% of patients (n=62). We compared postremission treatment with chemotherapy and allo-SCT in patients with HL who achieved CR1 after induction chemotherapy. To minimize time to transplantation selection bias, survival analysis was limited to patients who remained in CR1 for at least 3 months (median time to allo-SCT) after achieving CR1.

Results: Among the HL and non-HL groups, there was no significant difference in gender, age, hemoglobin level and platelet counts at initial diagnosis. HL patients had higher incidences of inv(16), NPM1+/FLT3-ITD- and CEBPA double mutations, but lower incidence of t(8;21) than non-HL patients. HL patients had a lower CR rate (85.5% vs. 95.6%, P=0.037) and a higher relapse rate (54.7% vs. 29.4%, P=0.002). The median follow-up from the date of initial diagnosis was 76.8 months. HL patients had an inferior overall survival (OS) (median, 103.1 months vs. not reached (NR), P=0.024) and relapse-free survival (RFS) (median, 15.2 months vs. NR, P< 0.0001) than non-HL patients (Figure 1A). Among the HL patients who attained CR1 for at least 3 months (n=51), 8 patients receive allo-SCT and 43 patients had postremission chemotherapy. The relapse rate was higher in chemotherapy group than in allo-SCT group (65.1% vs. 0%, P=0.001). Allo-SCT in CR1 was associated with a significantly better RFS (median, NR vs. 15.5 months, P=0.009) (Figure 1B) and a trend toward better OS (median, NR vs. 103.3 months, P=0.092). On the contrary, in non-HL group, patients receiving postremission chemotherapy alone had similar OS (P=0.816) and RFS (P=0.402) when compared to those with allo-SCT in CR1.

Conclusions: HL predicts a higher relapse rate and poorer survival in ELN favorable-risk AML patients. Hyperleukocytotic AML patients of the ELN favorable-risk group benefit from allo-SCT in CR1 with respect to RFS. Our results provide evidence for new transplantation strategies in hyperleukocytotic ELN favorable-risk patients.

Conflict of interest: The authors declare no relevant conflict of interest.

[P031 Figure]

$$$$

P032

Influence of FLT3-ITD and NPM1 status on allogeneic hematopoietic cell transplant outcomes in patients with cytogenetically normal AML

Ivan Pasic ¹ , Waleed Da'na ² , Wilson Lam ¹ , Arjun Law ¹ , Jeffrey Lipton ¹ , Auro Viswabandya ¹ , Dennis Kim ¹ , Santhosh Thyagu ¹ , Hans Messner ¹ , Fotios Michelis ¹

¹ Princess Margaret Cancer Centre, Medical Oncology & Haematology, Toronto, Canada; ² King Hussein Cancer Center, Amman, Jordan

Background: In patients with cytogenetically normal (CN) AML, prognostic information based on the molecular status of NPM1 and FLT3 genes is used to guide treatment. Individuals with an NPM1 mutation but without FLT3 internal tandem duplication (NPM1⁺/FLT3-ITD^-) have favourable risk disease and are offered allogeneic HCT following relapse, while those with NPM1^-/FLT3-ITD^- or FLT3-ITD⁺ are considered at higher risk of relapse and are offered transplant in CR1. However, data regarding the impact of NPM1 and FLT3-ITD status post-allogeneic HCT are limited. We examined the effect of NPM1 and FLT3-ITD status on outcomes post allogeneic HCT in patients with CN AML.

Methods: This retrospective single-center study included CN AML patients who underwent allogeneic HCT at the Princess Margaret Cancer Centre between 2006 and 2017. Overall survival (OS) was calculated using Kaplan-Meier analysis and multivariable Cox proportional hazards regression was performed. Cumulative incidence of relapse (CIR) and non-relapse mortality (NRM) were calculated using competing risk regression (Fine and Gray method). This study was approved by the University Health Network Research Ethics Board.

Results: A total of 131 patients were included: of these 63 (48%) were NPM1^-/FLT3-ITD^- and 51 (39%) FLT3-ITD⁺ (irrespective of NPM1 mutational status, all transplanted in CR1) while 17 (13%) were NPM1⁺/FLT3-ITD^- who had relapsed following conventional chemotherapy and were transplanted in CR2. Seventy-two patients (55%) were female. Median age at HCT was 54 (range 19-72) and median follow-up of survivors was 36 months (range 2-120). AML diagnosis was de novo in 117 (89%) and secondary to another hematological disorder or other conditions in 14 (11%) patients. Sixty-two (47%) patients had related donors, 64 (49%) had matched unrelated and 5 (4%) had haploidentical donors. Peripheral blood stem cells were used for all patients. CMV serostatus was negative for both donor and recipient in 20 (15%) of transplants. Myeloablative conditioning was used in 69 (53%) patients, 62 (47%) patients received reduced intensity regimens.

Univariate analysis demonstrated a 3-year OS of 51% (95% CI=41-60) for the entire cohort. When stratified by molecular status, NPM1^-/FLT3-ITD^- and FLT3-ITD⁺ patients demonstrated 3-year OS of 53% (95% CI=39-66) and 46% (95% CI=31-60), respectively, while NPM1⁺/FLT3-ITD^- patients transplanted in CR2 demonstrated 3-year OS of 56% (95% CI=29-76) (p=0.52) (Figure 1). CIR at 3-years for NPM1^-/FLT3-ITD^- and FLT3-ITD⁺ patients was 14% (95% CI=6-26) and 15% (95% CI=6-27) respectively, while no relapses occurred in the NPM1⁺/FLT3-ITD^- group. NRM at 3-years for NPM1^-/FLT3-ITD^- and FLT3-ITD⁺ patients was 38% (95% CI=25-50) and 43% (95% CI=28-57) respectively, while for NPM1⁺/FLT3-ITD^- patients 3-year NRM was 44% (95% CI=19-67) (p=0.89). Multivariable analysis for OS and NRM confirms that there is no significant difference between the molecular subgroups.

Conclusions: In patients with CN AML who undergo allogeneic HCT, the NPM1-/FLT3-ITD- and FLT3-ITD+ subgroups have similar post-transplant outcomes. Favourable risk patients (NPM1⁺/FLT3-ITD^-) that undergo transplant in CR2 have similar outcomes to the NPM1-/FLT3-ITD- and FLT3-ITD+ subgroups transplanted in CR1.

Clinical Trial Registry: Not applicable

Conflict of interest: None of the authors has anything to disclose.

[P032 Figure]

$$$$

P033

Inotuzumab-Ozogamicin induces remissions and enables allogeneic SCT in patients with Blinatumomab-refractory B-precursor acute lymphoblastic leukemia

Veit Bücklein ^1,2 , Alessia Fraccaroli ¹ , Dusan Prevalsek ¹ , Sarah Häbe ¹ , Christoph Schulz ¹ , Anna-Katharina Zoellner ¹ , Karsten Spiekermann ¹ , Marion Subklewe ^1,2 , Johanna Tischer ¹

¹ University of Munich, Department of Medicine III, University Hospital, Munich, Germany; ² University of Munich, Translational Cancer Immunology, Gene Center, Munich, Germany

Background: Allogeneic stem cell transplantation (allo-SCT) so far is the only curative approach for relapsed/refractory B-precursor ALL. Outcomes of allo-SCT improve with deep remission prior to conditioning. In the past years, treatment options for r/r B-ALL have been widened significantly by the approval of Blinatumomab (Blina) and Inotuzumab-Ozogamicin (InO), immunotherapeutic agents with impressive activity in chemotherapy-refractory disease. However, a significant proportion of patients still shows refractory disease or early relapse, and the most effective sequence of treatments is yet unknown. Here, we assessed the activity of InO in patients with Blina-refractory r/r ALL as bridge-to-transplant salvage concept.

Methods: Patients ≥ 18 years eligible for allo-SCT with r/r B-precursor ALL after Blina treatment failure were treated with InO and included in this retrospective analysis. Response to treatment, toxicity (with special attention to veno-occlusive disease), frequency of subsequent allo-SCT, relapse-free and overall survival were evaluated.

Results: Eight patients (7 male, 1 female, median age 36 years, range 22-56 years) with BCR-ABL negative B-precursor ALL were treated with InO after a median of four (range 3-5) prior anti-neoplastic treatment lines. Two of the patients had undergone previous allo-SCT. All patients had been treated with Blina prior to InO. Seven patients had Blina-refractory disease with no response to treatment, whereas one patient achieved a MRD-negative CR with two cycles of Blina, but had early relapse after 95 days. Median time from Blina initiation to start of InO treatment was 79 days (range 34-295 days). Short-term toxicity of InO was mainly hematologic and manageable, with thrombocytopenia (grade 1 in one, grade 2 in one, grade 4 in six patients) and neutropenia (grade 2 in one, grade 4 in seven patients) as most prominent adverse events. Apart from atypical pneumonia in one patient, no further infectious complications were recorded. VOD was suspected in two patients undergoing subsequent allo-SCT and successfully treated with defibrotide. No liver-associated consecutive morbidity or mortality was observed. Five of the eight patients (62,5%) responded to InO treatment (median 1 cycle, range 1-2 cycles). CR was achieved in one (MRD+), CRi in three (2 MRD-, 1 MRD+) and PR in one patient. All responding patients proceeded to allo-SCT (62,5%). Three of five patients responding to InO relapsed, with two relapses occurring before (56 and 60 days after InO initiation, respectively) and one relapse after subsequent allo-SCT (265 days after InO initiation). Median relapse-free survival was 58 days (range 16-367 days). Of the two patients with ongoing remission after InO, one remains relapse-free after 97 days, and one patient died relapse-free due to sepsis on day 367 after InO initiation. At the time of analyses, three patients are alive. Median overall survival was 121 days (range 27-550 days).

Conclusions: By inducing remissions in a significant proportion of patients with Blina-refractory B-precursor ALL, InO represents a valuable therapy option for those patients with otherwise fatal prognosis. Subsequent allo-SCT can be enabled in heavily pretreated patients with limited hematological toxicity, while observed VOD was manageable. If InO can induce long-lasting remissions without subsequent allo-SCT remains to be evaluated.

Conflict of interest: None of the authors has anything to disclose.

P034

Isolated extramedullary relapse after hematopoietic cell transplantation in acute leukemia patients

Annalisa Natale, Stella Santarone, Paola Olioso, Mauro Di Ianni, Gabriele Papalinetti, Corrado Colasante, Simone Ciavarella, Stefano Angelini, Paolo Di Bartolomeo

UTIE, Ospedale Civile, Dipartimento di Ematologia, Medicina Trasfusionale e Biotecnologie, Pescara, Italy

Background: Isolated extramedullary (IEM) relapse is a rare event after allogeneic HCT and its characteristics and prognosis are poorly defined. The aim of this study is to evaluate incidence, risk factors and outcome of IEM relapse in acute leukemia (AL) patients after HCT.

Methods: We retrospectively analyzed 458 consecutive AL patients (294 AML, 164 ALL) who underwent allogeneic HCT in our Center between March 1983 and June 2017.

Results: After a median follow-up of 140 months (range 7-420), 191 (41%) patients are living. The 35-yr overall survival (OS) was 33.7+3%. 178 patients (39%) showed AL relapse after a median follow-up of 188 days (range 18 -3980). Among them, 155 (87%) died for disease progression. The 30-yr cumulative incidence (CI) of relapse was 40.4+0.06% (33.6+0.08% in AML and 52.4+0.16% in ALL, P < 0.0001). IEM relapse was histologically documented in 15 patients (8%). IEM relapse involved testis (1), CNS (2), bone (2), breast (1), skin (1), lymphonode (2) in AML patients and CNS (2), skin (2), lymphnode (1), eye (1) in ALL patients. The 30-yr CI of IEM and bone marrow (BM) relapse was 3.4+ 0.008% and 36.9+ 0.05% respectively. The median age of patients with IEM relapse and BM relapse was 22 years (2-66) and 32 (1-67) respectively. Table 1 shows some characteristics of patient population.

We analyzed the impact of age, gender, AL type (AML vs ALL), presence of EM disease at diagnosis, CIBMTR risk factor, disease status, conditioning (chemotherapy vs TBI and MAC vs RIC), graft source (BM vs PBSC), donor (sibling, MUD, haploidentical), acute and chronic GvHD on the occurence of IEM relapse. EM disease at diagnosis was the only factor associated with increased risk of IEM relapse both in univariate and multivariate analysis (HR 4.14; P=0.0257). The 13 ys OS rate was 22.2+ 12% for IEM relapse and 9+2.4% in BM relapse (P=0.141). Among 15 patients with IEM relapse, 11 died (10 for disease progression and 1 for acute GvHD after DLI). 4 patients are living. Of them, 2 received a second allogeneic HCT, 1 chemotherapy and 1 sorafenib.

Conclusions: IEM relapse is a rare event compared to BM relapse in AL patients, seems to occur in younger patients and is associated with poor prognosis. Patients with EM disease at diagnosis are at higher risk of developing IEM relapse and need a close follow-up after HCT to detect early evidence of EM disease.

Conflict of interest: No conflict of interest.

	BM relapse (n=163)	IEM relapse (n=15)
Median age,years; Gender,M/F; EM at AL diagnosis	32; 85/78; 11	22; 5/10; 3
Disease risk: CIBMTR2,CIBMTR3	77,60	10,3
Status at HCT:CR1, CR2, CR>2, Active disease	79,40,11,33	6,8,0,1
Time to relapse,days	185(18-3980)	208(56-1900)
Chemotherapy/TBI; MAC/RIC	105/58; 153/10	12/3; 13/2
BM/PBSC; Sib/MUD/Haplo	107/53; 101/37/25	9/6; 9/3/3
aGvHD:no,grade1/2,grade3/4	102,38/20,2/1	11,0/4,0/0
cGvHD:yes,no,NV	19,133,11	1,13,1
Living/Dead	17/146	4/11

[ [P034 Table] Table 1]

P035

Leukemia in Kosovo: a post-war concern

Hatixhe Latifi-Pupovci, Miranda Selmanaj, Driton Vela, Violeta Grajcevci

University of Prishtina, Prishtina, Kosovo, Republic of

Background: During the Kosovo conflict, March through June 1999, NATO air-forces fired around 30 000 rounds with depleted uranium (DU) in 85 locations. The 4-fold increased incidence of lymphoblastic leukaemias in Iraqi children during the war period 1989-1993, attributed to the use of DU rounds, rised concerns about the posible link between the use of DU rounds and post-war incidence of leukemia in Kosovo.

The aim of this study was to compare the incidence rate of acute leukemia in the seven regions of Kosovo (Prishtina, Prizren, Peja, Gjakova, Mitrovica, Ferizaj and Gjilan) in pre-war period 1996-1998, with periods 2001-2003 and 2013-2015, and explore possible links between the number of sites hit by DU rounds and incidence of acute leukemia in Kosovo regions.

Methods: Acute leukemia cases diagnosed during period 1996-2015 were extracted from patients records in the University Clinical Centre of Kosovo. The number of sites hit by DU rounds by region was determined from accurate NATO records containing geographic coordinates. Incidence rates per 100,000 persons were calculated for each region and three periods (1996-1998, 2001-2003 and 2013-2015) based on population census data.

Results: Incidence rate of acute leukemia in Kosovo increased from 2.8 in 1996-1998 to 4.0 in 2001-2003 and 5.9 in 2013-2015. Also, there were significant differences between the regions. The biggest difference between incidence rates in the periods 1996-1998 and 2001-2003 were found in the Gjakova Region were incidence increased from 0 to 6.3, Prizren Region from 1.1 to 3.2, Peja Region from 1.4 to 2.7 and Mitrovica Region from 2.6 to 6.2. No significant changes were observed in Prishtina, Gjilan and Ferizaj. When differences between incidence rate for periods 1996-1998 and 2013-2015 were analysed, again, the biggest difference was found in Gjakova, from 0 to 6.3 followed by Peja from 1.4 to 6.8 and Mitrovica from 2.6 to 8.3, in Prizren from 1.1 to 3.2, Gjilan from 2.7 to 5.8, Ferizaj 3.7 to 5.8 and Prishtina from 5.8 to 6.5.

The regions with the largest number of sites hit by DU rounds were the Gjakova Region - 25, Prizren Region - 25, Peja Region - 16 whereas with smaller number of sites hit were regions of Mitrovica - 5, Prishtina - 4, Gjilan - 3, Ferizaj - 10.

Conclusions: The biggest difference in incidence of acute leukemia in the periods 1996-1998, 2001-2003 and 2013-2015 was observed in regions with the biggest number of sites hit by DU rounds. One exception is the Mitrovica Region which is known for high concentration of lead in the air and topsoil.

Conflict of interest: Hatixhe Latifi-Pupovci have nothing to disclose

P036

Long-term outcomes of hematopoietic cell transplantation in adults with acute lymphoblastic leukemia during first remission according to the available donor types

Jae-Ho Yoon, Nack-Gyun Chung, Sung-Soo Park, Young-Woo Jeon, Sung-Eun Lee, Byung-Sik Cho, Ki-Seong Eom, Yoo-Jin Kim, Hee-Je Kim, Chang-Ki Min, Seok-Goo Cho, Dong-Wook Kim, Jong-Wook Lee, Woo-Sung Min, Seok Lee

Catholic BMT Center, The Catholic University of Korea, Hematology, Seoul, Korea, Republic of

Background: Adult patients with high-risk acute lymphoblastic leukemia (ALL) have a poor outcome with standard chemotherapy and usually undergo hematopoietic cell transplantation (HCT) from a matched sibling donor (MSD) or matched unrelated donor (MUD). When a matched donor is not available, HCT from a partially-mismatched unrelated donor (MMUD), cord blood (CB), or familial haploidentical donor is considered. We tried to analyze the long-term HCT outcomes according to the donor types including MSD, MUD, MMUD, CB, and autologous (AUTO) sources.

Methods: We enrolled 661 consecutive adult ALL patients who underwent transplantation from 1995 to 2015 in their first remission (CR1). Median age was 33 years (range, 15-65 years). The myeloablative conditioning regimen consisted of cyclophosphamide (120 mg/kg) and total body irradiation (TBI; 13.2 Gy), and for double CB transplantation (DCBT), TBI (12 Gy), cytarabine (9 g/m²), and fludarabine (150mg/m²). Patients aged ≥50 years or those with comorbid conditions were given an identical reduced-intensity conditioning regimen consisting of fludarabine (150 mg/m²) and melphalan (140 mg/m²). Conditioning regimen for AUTO was TBI (12 Gy), cytarabine (9 g/m²), and melphalan (100 mg/m²). Graft-versus-host disease (GVHD) prophylaxis was attempted by administering calcineurin inhibitors (cyclosporine for MSD transplants, tacrolimus for unrelated donor transplants) plus methotrexate or mycophenolate mofetil (for DCBT). Antithymocyte globulin (ATG) was administered to the patients who received MMUD grafts.

Results: After a median follow-up of 80.4 months (range, 6.5-259.1 months), the 7-year overall survival (OS) for the entire 661 patients was 57.0%. In detail, the 7-year OS for MSD, MUD, MMUD, DCBT, and AUTO were 56.9%, 56.8%, 62.6%, 65.1%, and 48.6%, respectively (P=0.423). MSD, MUD and MMUD showed similar relapse rate (27.5%) and non-relapse mortality (NRM) rate (20.6%), but DCBT showed higher NRM rate (29.6%) and lower relapse rate (7.2%). AUTO showed higher relapse rate (41.6%). Interestingly, calculated GVHD and relapse-free survival (GRFS) at 7 years for MSD, MUD, MMUD, and DCBT were 34.1%, 15.0%, 40.9%, and 50.3%, respectively, which was related with higher incidence of severe chronic GVHD in MUD (27.6%) and MSD (14.1%) compared to DCBT (4.2%, P< 0.001). In high-risk subgroup of patients (n=517) including 233 Philadelphia chromosome-positive ALL, the 7-year OS for MSD, MUD, MMUD, DCBT, and AUTO were 54.9%, 58.1%, 57.8%, 60.7%, and 33.3%, respectively. The 7-year GRFS for MSD, MUD, MMUD, and DCBT were 30.7%, 14.5%, 35.3%, and 47.7%, respectively, which revealed lower relapse (8.7%) and severe chronic GVHD (2.6%) rates of DCBT.

Conclusions: Our long-term data showed that outcomes are similar for transplantation using MSD, MUD, MMUD, or CB sources in adult patients with ALL in CR1. However, GRFS was rather superior in patients treated with DCBT than those received other donor sources. Further studies with a sizable population are needed to clarify the role of alternative donor transplantation for adult ALL in the context of GRFS.

Conflict of interest: The authors declare no conflict of interest.

P037

Long-term results of allogeneic hematopoietic stem cell transplantation in 533 patients with various hematological malignancies: single centre experience

Marta Krejci, Michael Doubek, Miroslav Tomiska, Zdenek Racil, Andrea Janikova, Blanka Robesova, Jirina Prochazkova, Andrea Zmijakova, Katarina Ksenakova, Zdenek Kral, Jiri Mayer

University Hospital Brno, Dpt. of Internal Medicine, Hematology and Oncology, Brno, Czech Republic

Background: Allogeneic stem cell transplantation (SCT) has been considered as the treatment of choice for many hematological disorders, mainly for hematological malignancies. Role of allogeneic SCT is widely discussed in era of novel drugs. Here we present our long-term experience with allogeneic SCT in cohort of 533 pts with various hematological malignancies.

Methods: We analyzed 533 pts undergoing allogeneic SCT in our centre from November 1996 to June 2017. The diagnoses of pts were as follows: AML (187 pts; 35%), ALL (73 pts, 14%), lymphomas (66 pts, 12%), MDS+MPN (52 pts, 10%), CML (74 pts, 14%), CLL (47 pts, 9%), other diagnoses (34 pts, 6%). Median age of pts was 43 years (range 18-64). Types of donors and used grafts were as follows: HLA identical sibling, n=256 (48%); unrelated donor, n=275 (52%); PBSCs, n=492 (92%); BM, n=41 (8%). Median follow-up from SCT was 26.6 months (range 0-254), median follow-up from SCT for surviving pts was 83.2 months (range 5-254). Myeloablative conditioning (MAC) was used in 253 pts (47%), reduced intensity conditioning (RIC) was used in 280 pts (53%). Disease status at SCT was remission in 351 pts (67%) and active disease in 182 pts (33%). Median time from diagnosis to SCT was 7.4 months.

Results: The overall response rate after allogeneic SCT was 85%, including CR in 78% of pts and PR in 7% of pts. The incidence of acute GvHD was 42% (grade I+II in 32%, grade III+IV in 10%), the incidence of chronic GvHD was 40% (limited in 30%, extensive in 10%). Non-relapse mortality (NRM) was 18%, 19% and 20% at 1 year, 2 years and 4 years from SCT, respectively. Relapse incidence was 22%, 27% and 31% at 1 year, 2 years and 4 years from SCT, respectively. Median PFS was 31.5 months, median OS was 85.6 months for all 533 pts. Gender, age at SCT, type of donor, type of graft, time from diagnosis to SCT and type of conditioning did not significantly influence PFS and OS in our cohort of pts. Patients with AML have got significantly shorter PFS than pts with other diagnoses (median PFS 18.9 months versus 43.2 months, p = 0.031, HR 1.32), pts with active disease at SCT have got significantly shorter PFS than pts with remission at SCT (median PFS 9.6 months versus 73.7 months, p under 0.001, HR 1.79). In a multivariate analysis, diagnosis of AML versus other diagnoses (median OS 29.7 months versus 197.2 months, p under 0.001, HR 1.63) and active disease at SCT versus remission at SCT (median OS 25.0 months versus 186.5 months, p under 0.001, HR 1.93) were significant predictors of poor OS.

Conclusions: Allogeneic transplantation still remains the standard treatment option predominantly for various hematological malignancies. According to our results, type of diagnosis and disease status at allogeneic SCT are the simple, but important prognostic factors for PFS and OS. Patients in remission of disease at allogeneic SCT have got significantly longer PFS and OS.

Conflict of interest: Nothing to disclose for all authors.

P038

Monitoring of chimerism on sorted peripheral CD34+ cells in patients with acute leukemia receiving allogeneic bone marrow transplant

Maria Ida Bonetti ¹ , Benedetta Peruzzi ² , Francesco Mannelli ¹ , Anna Lari ³ , Tommaso Rondelli ² , Sara Bencini ¹ , Ilaria Cutini ¹ , Roberto Caporale ² , Anna Maria Grazia Gelli ² , Francesca Torricelli ³ , Alberto Bosi ¹

¹ Università di Firenze, SOD Ematologia AOU Careggi, Firenze, Italy; ² SOD Laboratorio Generale, Settore Citometria Clinica, AOU Careggi, Firenze, Italy; ³ SOD Diagnostica Genetica, AOU Careggi, Firenze, Italy

Background: An early detection of neoplastic cells after allogeneic hematopoietic stem cells transplantation (HSCT) in adult patients affected by acute leukemia provides the opportunity to manage and control the potential re-emergence of malignant clone. Several groups have reported as rapidly increasing mixed chimerism correlates with higher risk of relapse. Some predictors of impending relapse are Minimal Residual Disease (MRD) monitoring by flow cytometry (FC) and PCR-based chimerism on both Bone Marrow (BM) or Peripheral Blood (PB) samples at specific time-points after HSCT. Restricting chimerism analysis on sorted CD34+ cells could be an early predictor of relapse after HSCT.

Methods: At diagnosis of acute leukemia, the most useful leukemia-associated aberrant immuno-phenotype (LAIP) was established for each patient to be investigated at specific time-points during treatment plan to detect LAIP-positive cells. DNA was extracted from PB of donor and recipient before HSCT and chimerism assessment was performed on whole PB and sorted CD34+ cells of recipient after HSCT at specific time points: day 30, 60, 100, 180, month 12 and 18 from HSCT.

Results: From January 2012 to August 2015, 124 patients affected by acute leukemia received allogeneic transplantation. 47 out of 124 patients have been enrolled in the project. 14 patients were not evaluated due to toxicity after transplantation. Hitherto, data about 33 patients (in whom at least 1 evaluation has been carried out) are available. In most cases, the results were concordant: MRD-negative status and full donor chimerism both in the whole PB and in the CD34+ sorted fraction occurred. We found discordance in 7 patients: 2 of them showed a transitory loss of chimerism which immediately converted to 100% donor at the next detection and always MRD negative; 3 patients showed always full chimerism in CD34+ fraction, whereas MRD was positive at some time-points without relapse; in the other 2 patients, the chimerism analysis on sorted CD34+ cells allowed an early diagnosis of disease recurrence by highlighting the presence of the patient´s pathological cells with concomitant “negative signals” from FC-MRD and whole PB chimerism.

Conclusions: The last 2 patients clearly represent the basic rationale of this study: chimerism assessment may provide the opportunity of an early detection of neoplastic cells after transplantation. Regarding discordant results, the slightly loss of chimerism which immediately converted to 100% mandates for a confirmation with a further detection as soon as possible since one single assessment turns out to be inadequate to predict actual disease relapse. It would be interesting to address the significance of these fluctuations in chimerism both on whole PB and on sorted CD34+ cells on a larger patients' cohort. Chimerism assessment on CD34+ sorted fraction is an early predictor of relapse and might be useful to drive therapeutic strategy in the post transplantation setting.

Conflict of interest: The project was funded by Ministero della Salute and Regione Toscana (CUP D11J09000190003) and by Fondazione Biagioni-Borgogni and Legato Zottola Donation.

The authors declare no competing financial interests.

P039

Outcome of haploidentical haematopoietic stem cell transplant in pediatric acute leukemia - a single center experience from India

Sunil Bhat ¹ , Sohini Chakraborty ¹ , Shobha Badiger ¹ , Pooja Mallya ¹ , Ruchi Chaudhary ¹ , Ravi Joshi ¹ , Archana M V ¹ , Nataraj K S ² , Sharat Damodar ²

¹ Mazumdar Shaw Cancer Center, Narayana Health City, Pediatric Hematology Oncology and Bone Marrow Transplant, Bangalore, India; ² Mazumdar Shaw Cancer Center, Narayana Health City, Bangalore, India

Background: Haematopoietic stem cell transplant (HSCT) remains the potentially curative treatment modality for patients of acute leukemia in first complete remission or beyond. However, fully HLA matched donors are unavailable for majority of them and haploidentical transplant has emerged as a feasible option in such a scenario.

Methods: Retrospective analysis of all pediatric patients with acute leukemia undergoing haploidentical stem cell transplant at our Center from August 2014 to June 2017

Results: Thirty two patients with a median age of 71.5 months (range 26 to 204) formed the study group. 21 (65.6%) cases of acute lymphoblastic leukemia (ALL) (B cell = 16, T cell = 4 and biphenotypic=1) and 11 (34.4%) cases of acute myeloid leukemia (AML) were included in the study. Among the ALL patients 1 was in CR1 while remaining 21 were in CR2, whereas, among the AML patients 4, 6 and 1 patients were in CR1, CR2 and CR3 respectively. Myeloablative conditioning was used in all the cases. TBI based conditioning was used in ALL patients while Fludarabine-Busulphan based conditioning was used in AML patients .GVHD prophylaxis was provided in the form of TCR αβ and CD19 depletion in 12 patients (37.5%) while the remaining 20 (62.5%) received post transplant cyclophosphamide (PTCY) along with calcineurin inhibitors and mycophenolate mofetil. Peripheral blood was the source of stem cells in majority of the patients with a median infused CD34+ cell dose of 19.95 X 10⁶ /kg in TCR depletion group and 6.4 x 10⁶/ kg in the PTCY group. Thirty one patients achieved haematological recovery with neutrophil engraftment at a median of 12 days vs 15 days (P < 0.01) and platelet engraftment at a median of 10 days vs 21.5 days (P < 0.05) in the TCR depletion and PTCY group respectively .The transplant outcomes are described in the table given below (table 1). The cumulative incidences of grade 2-4 GVHD, grade 3-4 GVHD, relapse and non relapse mortality were 20% vs 68.8%, 20% vs 16.7%, 30% vs 43.4% and 25% vs 34.4% in the TCR depletion group vs PTCY group respectively (P > 0.05). The overall survival was 65.6% vs 42.4% in the TCR depletion and PTCY group while the disease free survival was 5 Graph 1 2.5% vs 42.4% in the two groups (figure 1).

Conclusions: Haploidentical HSCT is a feasible approach in paediatric acute leukemia patients lacking a fully matched donor.

Conflict of interest: none

Transplant outcomes	Number	Percentage
Acute GVHD
TCR depletion	4 / 12	33.3%
Grade I-II / III-IV	2 / 2
PTCY	13 / 20	65%
Grade I-II / III-IV	10 / 3
Relapse	6 / 32	18.8%
TCR depletion	2
PTCY	4
CMV reactivation	10/ 32	31.2%

[ [P039 Table] Table 1]

[P039 Figure]

$$$$

P040

Outcomes of allogeneic hematopoietic stem cell transplantation in patients with B-cell acute lymphoblastic leukemia harboring t(4;11)(q21;q23)

Jun Aoki ¹ , Shin Fujisawa ¹ , Shuichi Mizuta ² , Yukiyasu Ozawa ³ , Kazuteru Ohashi ⁴ , Shuichi Ota ⁵ , Hiroyasu Ogawa ⁶ , Takahiro Fukuda ⁷ , Junji Tanaka ⁸ , Yoshiko Atsuta ^9,10 , Shinichi Kako ¹¹

¹ Yokohama City University Medical Center, Yokohama, Japan; ² Toyohashi Medical Center, Toyohashi, Japan; ³ Japanese Red Cross Nagoya First Hospital, Nagoya, Japan; ⁴ Tokyo Metropolitan Cancer and Infectious Diseases Center, Komagome Hospital, Tokyo, Japan; ⁵ Sapporo Hokuyu Hospital, Sapporo, Japan; ⁶ Hyogo College of Medicine, Nishinomiya, Japan; ⁷ National Cancer Center Hospital, Tokyo, Japan; ⁸ Tokyo Women's Medical University, Tokyo, Japan; ⁹ Japanese Data Center for Hematopoietic Cell Transplantation, Nagoya, Japan; ¹⁰ Nagoya University Graduate School of Medicine, Nagoya, Japan; ¹¹ Saitama Medical Center, Jichi Medical University, Omiya, Japan

Background: Acute lymphoblastic leukemia (ALL) with t(4;11)(q21;q23) accounts for approximately 8%-10% of adult B-cell ALL, and is characterized by poor outcomes. Therefore, the presence of this chromosomal abnormality in ALL is an indication for allogeneic hematopoietic stem cell transplantation (allo-HSCT). However, transplantation outcomes of ALL with t(4;11)(q21;q23) remain unclear. We retrospectively examined the impact of t(4;11)(q21;q23) on the outcomes of allo-HSCT in patients with B-cell ALL.

Methods: Clinical data were collected from the registry database of the Japan Society for Hematopoietic Cell Transplantation. We selected B-cell ALL patients who were aged ≥16 years and underwent their first transplantation between January 2000 and December 2015. We compared clinical features and transplantation outcomes between the patients with t(4;11)(q21;q23) and the patients with a normal karyotype. Overall survival (OS) was estimated using the Kaplan-Meier method and compared using the log lank test. In a multivariate analysis, the Cox proportional hazard model was used to analyze OS, using the following covariates: age, sex, disease status, donor source, conditioning regimen, transplantation year, time taken for allo-HSCT from diagnosis and extramedullary disease.

Results: Number of cases with t(4;11)(q21;q23) and a normal karyotype was 107 and 1196 respectively. The median age of the t(4;11)(q21;q23) group was lower than the normal karyotype group (35 years vs. 40 years, P< 0.001). Higher proportion of the t(4;11)(q21;q23) group received allo-HSCT within 180 days of diagnosis (52.3% vs. 25.9%, P< 0.001). The probabilities of OS the t(4;11)(q21;q23) group was worse than the normal karyotype group (4ys, 46.2% vs.52.0%, P=0.024). Subgroup analysis stratified by disease status revealed that the probabilities of OS was comparable between the t(4;11)(q21;q23) group and the normal karyotype group in the CR patients (61.7% vs. 61.1%, P=0.519). In contrast, the t(4;11)(q21;q23) group exhibited poorer OS in the non-CR patients (10.9% vs. 21.8%, P=0.011). After adjusting for covariates, t(4;11)(q21;q23) had no significant impact on OS (HR, 1.27; P=0.100). Subgroup analysis showed t(4;11)(q21;q23) did not affect on OS in the CR patients (HR,1.06; P=0.77). Multivariate analysis in the t(4;11)(q21;q23) group revealed that non-CR at transplantation (HR,4.32, P< 0.001) was significant risk factor for poor OS.

Conclusions: These findings suggested that allo-HSCT in CR may overcome poor prognosis of ALL patients with t(4;11)(q21;q23).

Conflict of interest: All authors: nothing to disclose

P041

Outcomes of high risk myeloid malignancies undergoing Hematopoietic Cell transplant in a tertiary Care facility: real world evidence from India

Vivek S Radhakrishnan, Sanket Shah, Mita Roychowdhury, Jeevan Kumar, Saurabh Jayant Bhave, Reena Nair, Mammen Chandy

Tata Medical Center, Clinical Haematology and Bone Marrow Transplant, Kolkata, India

Background: Hematopoietic Cell transplantation (HCT) remains the cornerstone of treatment of high risk myeloid malignancies. Real world evidence from middle income countries are few.

Methods: We undertook a retrospective chart review of 103 HCT undertaken for 98 patients with high risk myeloid malignancies between Dec 2011 and Oct 2017 (M:69, F:29). Median age: 31 yrs (range: 7-60 yrs). These included Acute Myeloid Leukaemia (64), Acute Pro-myelocytic leukemia (APL: 5), MDS (11), CML (Chronic phase:3, Lymphoid Blast crisis:6, Myeloid blast crisis:5), Myelofibrosis (2), Chronic myelomonocytic Leukemia (1), and Blastic Plasmacytoid dendritic cell neoplasm (1). High risk disease was on account of morphologic criteria (e.g myelodysplasia), high risk cytogenetics, molecular profile or clinical indications (induction failure, relapsed disease, etc.). 34 patients were in CR1, 19 in CR2 or CR3, and 45 patients with active disease (30 primary refractory, and 15 relapsed refractory) at the time of transplant referral. Amongst 34 patients who underwent transplant for relapsed disease: 7 relapsed in < 6 months from CR1, 8 within 6-12 months and 19 relapsed after 12 months. All patients underwent allogeneic HCT, except 5 with relapsed APL (autologous HCT). Allogeneic HCT donor source included 53 Matched sibling donors, 35 haploidentical related donors, and 10 Matched unrelated donors. Conditioning therapy was myeloablative in 83 allogeneic and 5 autologous HCT. Reduced intensity conditioning was undertaken in 15 allogenic HCT. All patients received peripheral blood stem cell transplant (PBSCT). Throat and faecal surveillance culture for Multi-Drug Resistant Organisms (MDRO) was done for all 98 patients. GVHD prophylaxis varied, depending on type of allogeneic transplant, and included either CsA+mMtx±ATG or PTCy+Tac+MMF approach (34 patients).

Results: The median CD34 cell dose was 6.1x10^6 cells/kg. Primary graft failure, 100-day mortality, Acute GVHD, Gr III/IV aGVHD, and cGVHD rates were 7%, 17%, 25%, 10% and 33% patients respectively. At a median follow up of 344 days for the entire cohort, 5 year Overall Survival(OS) and Relapse Free Survival (RFS) were 48% and 40% respectively. Median overall survival: 35.1 months. MDRO colonization was observed in stool surveillance of 92 patients and throat surveillance culture of 30 patients [Among 116 species of MDRO identified in stool samples: 36 were ESBL producing gram negative bacilli(GNB), 38 carbapenem resistant GNB, and 7 vancomycin resistant enterococci]. In this single center retrospective series, the type of transplant, status of disease at the time of transplant, duration of remission in case of relapsed disease, and Blood group or sex mismatch between donor and recipient had no impact on OS or RFS. The only factor which was found to be significantly affecting outcome was gender: females had better outcome compared to males.

Conclusions: In a middle income country with a high prevalence of MDRO colonization, hematopoietic cell transplantation of high risk myeloid malignancies in a tertiary care center yields modest outcomes

Conflict of interest: No relevant conflicts of interest

P042

Patient-Reported Quality of Life (QoL) Following Tisagenlecleucel Infusion in Adult Patients with Relapsed/Refractory (r/r) Diffuse Large B-Cell Lymphoma (DLBCL)

Richard T. Maziarz ¹ , Michael R. Bishop ² , Constantine S. Tam ³ , Peter Borchmann ⁴ , Ulrich Jaeger ⁵ , Joseph McGuirk ⁶ , Edmund K. Waller ⁷ , Samantha Jaglowski ⁸ , Charalambos Andreadis ⁹ , Stephen Ronan Foley ¹⁰ , Isabelle Fleury ¹¹ , Stephan Mielke ¹² , Jason R. Westin ¹³ , Phoebe Joy Ho ¹⁴ , Veronika Bachanova ¹⁵ , Harald Holte ¹⁶ , John M. Magenau ¹⁷ , Oezlem Anak ¹⁸ , Lida Bubuteishvili Pacaud ¹⁹ , Siegbert Guenther ¹⁸ , Jie Zhang ¹⁹ , Lawrence Rasouliyan ²⁰ , Feng Tai ¹⁹ , Gilles A. Salles ²¹ , Stephen J. Schuster ²²

¹ Oregon Health & Science Knight Cancer Institute, Center for Hematologic Malignancies, Portland, OR, United States; ² University of Chicago, Hematopoietic Stem Cell Transplantation Program, Chicago, IL, United States; ³ St Vincent's Hospital and University of Melbourne, Peter MacCallum Cancer Centre, Melbourne, Australia; ⁴ University Hospital of Cologne, Department of Haematology and Oncology, Cologne, Germany; ⁵ Medical University of Vienna, Department of Medicine, Vienna, Austria; ⁶ University of Kansas Medical Center, Department of Blood and Bone Marrow Transplant, Kansas City, KS, United States; ⁷ Winship Cancer Institute of Emory University, Bone Marrow and Stem Cell Transplant Center, Atlanta, GA, United States; ⁸ James Cancer Hospital and Solove Research Institute of Ohio State University Comprehensive Cancer Center, Columbus, OH, United States; ⁹ University of California San Francisco Helen Diller Family Comprehensive Cancer Center, Department of Hematology and Blood and Marrow Transplant, San Francisco, CA, United States; ¹⁰ McMaster University, Juravinski Hospital and Cancer Centre, Hamilton, Canada; ¹¹ University of Montreal, Maisonneuve-Rosemont Hospital, Montreal, Canada; ¹² Würzburg University Medical Center, Center for Allogeneic Stem Cell Transplantation, Würzburg, Germany; ¹³ University of Texas MD Anderson Cancer Center, Department of Lymphoma and Myeloma, Division of Cancer Medicine, Houston, TX, United States; ¹⁴ The University of Sydney, Department of Medicine, Sydney, Australia; ¹⁵ University of Minnesota, Department of Medicine, Division of Hematology, Oncology and Transplantation, Minneapolis, MN, United States; ¹⁶ Oslo University Hospital, Department of Oncology, Oslo, Norway; ¹⁷ University of Michigan Comprehensive Cancer Center, Ann Arbor, MI, United States; ¹⁸ Novartis Pharma AG, Basel, Switzerland; ¹⁹ Novartis Pharmaceuticals Corporation, East Hanover, NJ, United States; ²⁰ RTI Health Solutions, Barcelona, Spain; ²¹ CHU Lyon-Sud Hospital, Hospices Civils de Lyon, Department of Hematology, Lyon, France; ²² University of Pennsylvania, Lymphoma Program, Abramson Cancer Center, Philadelphia, PA, United States

Background: JULIET (NCT02445248) is evaluating a single infusion of tisagenlecleucel (CTL019) in adult patients with r/r DLBCL. Primary analyses showed a best overall response rate of 53% (complete response [CR], 40%).

Methods: Patients were aged ≥18 years with chemorefractory DLBCL after ≥2 lines of therapy and either failed or were otherwise ineligible for autologous hematopoietic stem cell transplant (SCT). At baseline, month-3 and month-6, patients completed the Short Form-36 Health Survey v2 (SF-36) and the Functional Assessment of Cancer Therapy-Lymphoma (FACT-Lym). Higher scores indicate better QoL.

Results: 99/147 patients were infused before data cutoff; 95% had previously received ≥2 chemotherapy treatments and 48% had relapsed after SCT. 43 pts with a best response of CR or partial response (PR) were followed ≥6 months prior to data cutoff or discontinued earlier, and were eligible for analysis. QoL instruments were completed by 76 patients (94%) at baseline and 40 patients who had CR or PR. 26 and 16 of patients with CR or PR completed the assessments at month-3 and -6, respectively. Scores of patients with CR or PR (mean baseline, mean change at month-3, mean change at month-6) indicated sustained improvements over baseline in all categories (Table 1). Similar trends were observed for SF-36.

Conclusions: Improvements in QoL were observed at month-3; scores at month-6 suggested sustained improvements. These results suggest that tisagenlecleucel therapy may improve QoL among responders despite patients experiencing early toxicity.

Clinical Trial Registry: ClinicalTrials.gov NCT02445248

Conflict of interest:

R. Maziarz: Novartis: research funding, honoraria, membership on entity’s board of directors/advisory committees; Incyte: consultancy, honoraria; Juno, Kite: honoraria; Athersys: patents and royalties

M. Bishop: Novartis: consultancy: includes expert testimony; other: steering committee; Kite, Battersea: Consultancy: includes expert testimony; Celgene: honoraria, speakers bureau

C. Tam: Novartis: honoraria.

P. Borchmann: Novartis: Honoraria.

U. Jaeger: Roche, Janssen, Novartis: Honoraria; Membership on Entity's BOD/AC Novartis: RF

J. McGuirk: Novartis: No payment PI CAR-T Study.

E. Waller: Novartis: Consultancy: Includes ET, RF; Cambium Medical: Patents and Royalties, Membership on Entity's BOD/AC; Celldex, PRA, Amgen: Consultancy: Includes ET.

S. Jaglowski: Novartis, Kite: Consultancy, RF; Unum, Pharmacyclics: RF

C. Andreadis: Novartis: RF, Honoraria; Cellerant, Incyte, Pharmacyclics, Amgen: RF;Genentech:Honoraria; Equity Ownership-Spouse; Seattle Genetics, Gilead Astellas:Honoraria.

S. Foley: Novartis: Consultancy

I. Fleury: Merck, Janssen, Roche, Seattle Genetics, Gilead, Amgen, Lundbeck, Novartis:Consultancy.

S. Mielke: KIADIS: Travel Grants; Jazz Pharma:Speakers Bureau; Novartis: Consultancy; Celgene, Cellex GmbH: Speakers Bureau,Travel Grants;MSD: Consultancy,Travel Grants;Gilead:Travel Grants; DGHO: Travel Support; ISCT: Travel Support.

J. Westin: Novartis, Apotex, Kite, Celgene:Membership on Entity's BOD/AC.

P. Ho: Novartis, Amgen, BMS, Celgene, Janssen, Takeda: Honoraria, Membership on Entity's BOD/AC; Janssen,Takeda: Consultancy,.

V. Bachanova: Novartis: RF; Seattle Genetics, Juno, Zymogen: Membership on Entity's BOD/AC; Oxis: RF.

H. Holte: Novartis:Membership on Entity's BOD/AC.

J. Magenau: Nothing to Disclose.

O. Anak, S. Guenther: Novartis Pharma AG: Employment.

L. Pacaud, J. Zhang, F. Tai: Novartis Pharmaceuticals: Employment

L. Rasouliyan: RTI Health Solutions:Employment.

G. Salles: Amgen, BMS, Celgene, Gilead, Janssen, Kite, Merck, Morphosys, Novartis, Servier, Roche: Consultancy, Participation Symposia/AB; Roche: Consultancy, Research Grant;

S. Schuster: Novartis, Celgene, Genentech, Merck, Janssen R&D, HoffmanLaRoche, BMS:RF; Novartis, Celgene, Genentech, Seattle Genetics, Nordic Nanovector: Consultancy: Includes ET

[P042 Figure]

$$$$

P043

Prognostic Effect of Pre-Treatment Cytogenetic Abnormalities In Iranian Adults Acute Myeloid Leukemia After Allogeneic Hematopoietic Stem Cell Transplantation

Marjan Yaghmaie ¹ , Amirabbas Rashidi ¹ , Kamran Alimoghadam ¹ , Amir Kasaeian ¹ , Mohammad Jahani ¹ , Seyed Asadollah Mousavi ¹ , Mohammad Vaezi ¹ , Hossein Kamranzadeh ² , Ardeshir Ghavamzadeh ²

¹ Tehran University of Medical Sciences, Hematology, Oncology and Stem Cell Transplantation Research Center, Tehran, Iran, Islamic Republic of; ² tehran University of Medical Sciences, Hematology, Oncology and Stem Cell Transplantation Research Center, Tehran, Iran, Islamic Republic of

Background: Several studies have revealed the significance of pre-treatment cytogenetic findings in tumor cells as a strong clinical predictor. We aimed to evaluate the effect of pre-treatment cytogenetic abnormality on treatment outcomes in Iranian adults with acute myeloid leukemia (AML) after allogeneic hematopoietic stem cell transplantation (HSCT).

Methods: A total of 323 adult patients with non-M3 AML underwent allogeneic HSCT from March 2011 to March 2015 in Shariati General Hospital, from which 206 patients with pre-treatment cytogenetic data were included in our study with median follow-up time of 47 months. Overall survival (OS) and Disease-free-survival (DFS) rate of different cytogenetic abnormalities were compared with normal karyotype.

Results: Five-year OS rate for all patients were 58.54 %. OS and DFS rates were significantly lower in abn(3q), inv(3)(q21;q26),t(3;3)(q21;q26),del(5q)/add(5q), t(6;9)(p23;q34), -7, abn(7q), +8, 11q23 and abn(11q) (excluding (11;19))(p21;q23) and (11;19)(q23;p13)), abn(17p), complex karyotype (more than 3 unrelated abnormalities) and monosomal karyotype. None of the cytogenetic abnormalities were associated with favorable outcomes.

Conclusions: Our study suggested that age and pre-treatment cytogenetic abnormalities are independent predictors of treatment outcome after HSCT in AML patients. Unlike previous studies, in our study t (8; 21) and Inversion (16) were not associated with favorable outcomes and were classified as intermediate risk groups.

Conflict of interest: nothing to disclose

P044

Prognostic significance of pretransplant minimal residual disease in AML patients with mutated NPM1

Veronika Valkova ¹ , Jan Vydra ¹ , Antonín Vítek ¹ , Marketa Markova ¹ , Jaroslav Polak ² , Radka Petrbokova ² , Ela Cerovska ² , Ludmila Novakova ¹ , Barbora Cemusova ¹ , Michal Kolar ¹ , Petr Cetkovsky ¹

¹ Institute of Haematology and Blood Transfusion, BMT, Prague, Czech Republic; ² Institute of Haematology and Blood Transfusion, Dpt of Genomics, Prague, Czech Republic

Background: There is increasing evidence that in AML, minimal residual disease (MRD) persisting after standard induction therapy is associated with increased risk of relapse and poor survival. Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is considered to be the most effective strategy in the treatment of high risk AML. The aim was to evaluate the prognostic impact of pre-transplant MRD measured by real-time quantitative PCR in patients with NPM1-mutated AML receiving allo-HSCT.

Methods: From 3/2004 to 6/2017, we evaluated 60 patients (pts) with a median age of 52 years (range; 24-66 years), 38 (63%) women. Median follow-up for survival from time of allo-HSCT was 19 months. Twenty-six (43%) pts were FLT3-ITD positive. At allo-HSCT, 44 (73%) pts were in first, 12 (20%) in second complete remission (CR) and 4 had refractory disease (RD). Forty-two (70%) pts underwent myeloablative, 18 reduced-intensity alloHSCT. The donors were: identical sibling in 10 pts, haploidentical related in 4 pts, matched unrelated donor in 31 pts and mismatched UD in 15 pts. In 53 pts were used PBPC, in 7 pts bone marrow. All but one pts had intermediate risk prognostic score. Pre-transplant NPM1 MRD levels were measured in peripheral blood in all 60 pts. Pts with pre-transplant MRD level of ≤10 mutant copies of NPM1 per 10,000 ABL copies were considered as MRD negative. At the time of allo-HSCT 47 pts were MRD negative and 13 positive (4 refractory patients were excluded from further analysis).

Results: The estimated probabilities of 3-year relapse incidence (RI), non-relapse mortality (NRM), event-free survival (EFS), overall survival (OS) and chronic GVHD for the whole cohort were 21%, 32%, 43%, 53% and 84% respectively. Thirty-one (52%) of pts developed acute GVHD. Significantly better results were observed in MRD neg pts with respect to RI and EFS (3-year RI and EFS 19% and 54% and 64% and 13% for MRD neg and MRD positive pts resp, p= 0,001), similar prognostic significance was observed in FLT3 -ITD positivity (3-year RI and EFS 10 % and 53% and 55% and 29% for FLT3-ITD neg and positive pts resp., p= 0,0035 and 0,0005). The difference was not significant in terms of OS, however, there was no difference in OS even in pts with RD or CR. In multivariable analysis, increasing age was the only statistically significant predictor of poor OS (HR 1,8 for every 10 years of age, 95% CI 1,03-3.17; p = 0.04). In terms of EFS we observed inferior outcome in FLT3-ITD positive pts (HR 2,34; 95% CI 1,08- 5,07; p=0.03). FLT3-ITD positivity (HR 7.76 (1.5 - 40),p = 0.014) and NPM1 negativity (HR 0.17 (0.05 - 0.61) p = 0.006) were associated with higher relapse risk.

Conclusions: In AML patients with NPM1 mutation, MRD positivity before allo-HSCT is clearly associated with a higher incidence of relapse. The NRM in our cohort was relatively high, most likely due to GVHD. However, our data suggest that allo-HSCT and its associated development of GVL effect, mostly linked just to GVHD, may overcome the negative prognostic impact of MRD positivity.

Conflict of interest: None

P045

Prophylactic administration of Azacitidine in patients with AML after haploidentical stem cells transplantation increases overall survival

Renat Badaev, Darina Zammoeva, Larisa Girshova, Diana Babenetskaya, Natalia Ilyina, Yulia Alexeeva, Dmitriy Motorin

Federal Almazov North-West Medical Research Centre, St. Petersburg, Russian Federation

Background: Haplo-SCT is an appropriate alternative, when match donor unavailable.

Although, GvHD and relapses could potentially decrease the survival. Prophylactic administration of Azacitidine, performing antileukemic and immunomodulating activity, can help to overcome these obstacles.

Methods: 18 patients with AML after haplo-SCT were included in the study. All patients achieved MRD-negative remission on day 30 after SCT. MRD was assessed by measurement of WT1-gene expression. Administration of Azacitidine began not earlier than 2 months after SCT in case of full engraftment and absence of active GVHD. Dose of Azacitidine was 100 mg D1-D5 every 28 days during 1 year after SCT. DLI were added in case of molecular relapse.

Results: 11 patients received prophylactic administration of Azacitidine and 7 were in control group. Median time of beginning of Azacitidine - 4 months (2-10). Median number of courses - 3,5 (1-9). The most frequent toxicity was hematological: anemia 1-2 - 27,8% (N=5), thrombocytopenia 3-4 - 11,1% (N=2), neutropenia 3-4 - 27,8% (N=5). No case of secondary graft failure occurred.

9 patients (81,8%) had acute GVHD before treatment with Azacitidine and only 5 (45,4%) - after Aza-treatment. Four patients had relapse of aGVHD (3 - skin, 1 - gut) and only one case of gut-GVHD occurred de novo after treatment. Two patients required systemic immunosuppression with methylprednisolone. In control group aGVHD occurred in 5 patients (71,4%) and 2 required systemic immunosuppression. Difference was not statistically significant due to small number of patients.

Rate of chronic GVHD in Aza-group and control group was comparable 9,1% (N=1) и 14,3% (N=1). All cases were skin-chGVHD accompanying with hepatic-GVHD in one patient.

Molecular relapse didn`t observe in Aza-group. In control group two patients (28,6%) had molecular relapse and were treated by Azacitidine with DLI. The first patient died because of severe aGVHD and another one achieved MRD-negative remission and still alive. We didn`t see any case of bone marrow relapse, probably because of intense GVL after haplo-SCT and short period of observation.

Currently, 10 patients in Aza-group (90,9%) and 3 patients in control group (42,9%) are alive with median observation period 9,5 and 7 months respectively. In Aza-group one patient died of aGVHD and in control group - 2 of infection and 2 of aGVHD. We saw significantly better overall survival in Aza-group (p=0,021).

Conclusions: Prophylactic administration of Aza after haplo-SCT in patient with AML is safe and well tolerated. Aza prophylaxis significantly improves overall survival. This improvement could be due to the decrease of aGvHD and relapses, although, the differences appeared to be statistically nonsignificant in our cohort of patients.

Conflict of interest: None

Number of Haplo-SCT	18
Age	18-57 (median 42)
Sex : M/F	7/11
Status before SCT:
CR MRD-negative	11
CR MRD-positive	5
No CR	4
Graft source:
Bone marrow	2
Peripheral blood	16
CD34+ mln/kg	1,9 - 10,4 (median 4,18)
Conditioning regimens:
Flu/Cy/Mel	13
Bu/Flu/Cy	3
Conditioning with previous cytoreduction	2
GVHD prophylaxis	PT-Cy, MMF and CsA

[ [P045 Table] Characteristics of Patients]

[P045 Figure]

$$$$

P046

Prospective single center experience of allogeneic stem cell transplantation in adult patients with NPM1 acute myeloid leukemia

Federico Lussana ¹ , Chiara Caprioli ¹ , Paola Stefanoni ¹ , Chiara Pavoni ¹ , Orietta Spinelli ¹ , Ksenija Buklijas ¹ , Pamela Zanghì ¹ , Alessandra Algarotti ¹ , Caterina Micò ¹ , Anna Grassi ¹ , Tamara Intermesoli ¹ , Alessandro Rambaldi ^1,2

¹ Ospedale Papa Giovanni XXIII, Hematology and Bone Marrow Transplant Unit, Bergamo, Italy; ² Università degli Studi di Milano, Department of Oncology and Hematology, Milan, Italy

Background: Nucleophosmin (NPM1) is one of the most commonly mutated genes in acute

myeloid leukemia (AML), present in 20-30% of cases. Mutations in NPM1 represent a distinct entity in the World Health Organization (WHO) classification and commonly the prognosis of patients with NPM1 mutation is considered favorable. Nevertheless, recent studies showed a correlation between the persistence of NPM1 MRD level and patient adverse outcome. This association has generated substantial interest in using results of MRD-testing for the decision to transplant stratifying patients, but the benefit associated with allogeneic stem cell transplantation (alloHSCT) in first remission remains to be investigated.

Methods: From year 2006, 72 AML patients bearing the NPM1 mutation, with a median age of 53 (range 16-82) were consecutively treated according to the Northern Italy Leukaemia Group (NILG) protocol 02/06 [ClinicalTrials.gov Identifier: NCT00495287]. Patients were considered eligible to allogeneic transplant in first remission if, at diagnosis, they were FLT3 ITD positive, or had a high white blood cells (WBC) count (>50X10⁹/L), or they showed a persistent minimal residual disease (MRD) as molecularly detected after consolidation with high dose cytarabine.

Results: The main clinical findings of patients analyzed are summarized in Table 1.

Patients receiving alloHSCT were younger, had a higher WBC count with a FLT-ITD mutation at diagnosis, and showed more frequently a persistent MRD postivity after consolidation compared to patients not undergoing tranplantation (Table 1). After induction course, 66 patients (92%) achieved a complete hematological remission (CR) and 41 (62%) underwent an alloHSCT in first complete remission. MRD status after consolidation chemotherapy was available for 61 patients (85%). Forty-one patients received an alloHSCT, the donors being 8 siblings, 27 unrelated, 1 haploidentical and 5 cord blood units. The 5-years disease free survival (DFS) was significantly better for patients receiving an alloHSCT 69% (95%CI:52-81%) vs 29% (95%CI 11-50%), p=0.046 (Figure 1).

However, the 5 years OS was not significantly different being 67% (95%CI: 50%-80%) in patients receiving alloHSCT compared to 55% (95%CI:30-75%) in patients not undergoing alloHSCT (p=0.705) in first remission. For the whole patients' cohort (n=72), the 5 years overall survival (OS) was 59% (95%CI: 46%-71%). By multivariate analysis, the presense of FLT3-ITD mutation and the persistence of MRD after consolidation chemotherapy were associated with a shorter survival indipendently from the transplant consolidation.

Conclusions: Our results shows that alloHSCT improve DFS compared to standard chemotherapy also in patients with NPM1 acute myeloid leukemia at higher risk of leukemia relapse for the presence of FLT-ITD mutation and persistence of MRD postivity before transplant.

Conflict of interest: None of the authors has anything to disclose.

	Allo HSCT (N=41)	No allo HSCT (N=25)
Age years, median (range)	50 (16-75)	60 (21-82)
Sex, N (%) Male Female	17 (41%) 24 (59%)	13 (52%) 12 (48%)
LDH, U/L median (range)	1157 (725-2294)	1089 (757-1566)
WBC, X 109/L, median (range)	36 (16-86)	15 (3-56)
Molecular abnormalities None FLT3ITD Others	16 (39%) 17 (41%) 8 (20%)	13 (52%) 5 (20%) 7 (28%)
MRD post-consolidation, N (%) Positive Negative	22 (55%) 18 (45%)	14 (67%) 7 (33%)
Donor type, N(%) Sibling Unrelated Cord blood HLA Haploidentical	8 (20%) 27 (66%) 5 (12%) 1 (2%)	- - - -

[ [P046 Table] Patients' characteristics]

[P046 Figure]

$$$$

P047

Reduced-intensity transplantation in patients with acute myeloid leukemia beyond complete remission - single centre experience

Michal Karas, Katerina Steinerova, Pavel Jindra, Daniel Lysak, Marcela Hrabetova, Jiri Sramek

Charles University Hospital Pilsen, Hemato-Oncology Dpt., Pilsen, Czech Republic

Background: reduced-intensity transplantation (RIT) for acute myeloid leukemia (AML) beyond complete remission (CR) is associated with higher relapse incidence, lower progression-free survival (PFS) and overall survival (OS). But elderly patients (pts) or pts with comorbidities can rarely benefit from myeloablative conditioning or intensive chemotherapy (CHT) for AML due to high risk of treatment-related mortality (TRM). To evaluate the role of RIT and potential effect of pre-transplantation characteristics in treatment of pts with AML beyond CR we retrospectively analysed the outcome of such pts undergoing RIT in our centre.

Methods: since 2002 68 pts with median of age 60 years (22-74 years) with AML beyond CR underwent RIT. HCT-CI was ≥3 in 26 (38%) pts. 29 pts were in the 1^st or 2^nd untreated relapse of AML, 20 pts had chemotherapy-resistant disease and 19 pts underwent up-front RIT without previous AML therapy, all pts had more than 5% blasts in the bone marrow (BM). Donors were in 28% HLA identical related, 10% HLA haploidentical related, 40% HLA matched unrelated and 22% HLA mismatched unrelated. The source of stem cells was peripheral blood and the median of infused CD 34+ cells was 6,55x10^6/kg (1,74-14,34x10^6/kg). The conditioning regimen consisted of fludarabine and melphalan in 65% pts and 35% pts underwent RIT after sequential regimen of CHT and RIC.

Results: 65 (96%) pts engrafted and achieved complete remission (CR) with full donor chimerism (FDC) at day +30 after RIT. 45 pts (69%) developed acute GVHD (20% grade III-IV) and among 56 evaluable pts 22 (39%) of them developed chronic GVHD (9 mild, 5 moderate, 8 severe). With median follow-up 86 months (3-179 months) 21 pts (31%) are alive (20 pts in continuing CR with FDC). 24 pts (35%) relapsed with median of time from RIT to relapse 6 months (1-27 months) and 23 of them died. 24 pts (35%) died due to TRM and 9 (13%) of them till day +100 after RIT. The estimated probabilities of 3-years PFS and OS are 31% and 34%. Only presence of blasts in PB (p = 0.0001), more than 20% of BM blasts (p= 0.0285) and HCT-CI ≥ 3 (p= 0.039) statistically significantly negatively affected the outcome of RIT.

Conclusions: in spite of relatively small number of evaluated pts and retrospective type of analysis our data suggest that RIT can achieved continuing long-term CR in about third of pts with AML beyond CR not eligible for myeloablative conditioning or intensive CHT of AML. The most important pre-transplantation factors significantly influencing RIT outcome were finding of circulating blasts in PB, percentage of BM blasts and HCT-CI. Status of disease, diagnostic cytogenetic findings, age, type of donor and type of conditioning regimen did not significantly affect the outcome of RIT.

Conflict of interest: The authors declare no conflict of interest.

P048

Relapse after allogeneic hematopoietic stem cell transplantation for acute leukemia in children: a survey by the Turkish Pediatric Bone Marrow Study Group of 255 cases

Volkan Hazar ¹ , Gülsün Tezcan Karasu ² , Gülyüz Öztürk ³ , Alphan Küpesiz ⁴ , Savaş Kansoy ⁵ , Namık Özbek ⁶ , Vedat Uygun ⁷ , Talia İleri ⁸ , Fatma Visal Okur ⁹ , Ülker Koçak ¹⁰ , Suar Çakı Kilic ² , Arzu Akçay ³ , Elif Güler ⁴ , Salih Gözmen ⁵ , Musa Karakükcü ¹¹ , Ibrahim Bayram ¹² , Tekin Aksu ⁶ , Akif Yeşilipek ⁷ , Barboros Şahin Karagün ¹³ , Sebnem Yılmaz Bengoa ¹⁴ , Mehmet Ertem ⁸ , Duygu Uçkan ⁹ , Zühre Kaya ¹⁰ , Tunç Fışkın ¹⁵ , Didem Atay ³ , Funda Tayfun Küpesiz ⁴ , Orhan Gürsel ¹⁶ , Yöntem Yaman ¹⁷ , Ceyhun Bozkurt ¹⁸ , Müge Gökçe ¹⁹ , Serap Aksoylar ^5,20

¹ Medical Park Göztepe Hospital, Pediatric Hematology & Oncology and BMT Unit, Istanbul, Turkey; ² Medical Park Göztepe Hospital, Istanbul, Turkey; ³ Acıbadem University Faculty of Medicine, Atakent Hospital, Istanbul, Turkey; ⁴ Akdeniz University Faculty of Medicine, Antalya, Turkey; ⁵ Ege University Faculty of Medicine, Izmir, Turkey; ⁶ Ankara Child Health and Diseases Hematology Oncology Training and Research Hospital, Ankara, Turkey; ⁷ Medical Park Antalya Hospital, Antalya, Turkey; ⁸ Ankara University Faculty of Medicine, Ankara, Turkey; ⁹ Hacettepe University Faculty of Medicine, Ankara, Turkey; ¹⁰ Gazi University Faculty of Medicine, Ankara, Turkey; ¹¹ Erciyes University Faculty of Medicine, Kayseri, Turkey; ¹² Çukurova University Faculty of Medicine, Adana, Turkey; ¹³ Acıbadem University Faculty of Medicine, Adana Hospital, Adana, Turkey; ¹⁴ Dokuz Eylül University Faculty of Medicine, Izmir, Turkey; ¹⁵ Medical Park Samsun Hospital, Samsun, Turkey; ¹⁶ Health Science University Gülhane Medical Faculty, Ankara, Turkey; ¹⁷ Istanbul Medipol University Faculty of Medicine, Istanbul, Turkey; ¹⁸ Medical Park Bahçelievler Hospital, Istanbul, Turkey; ¹⁹ Yeni Yüzyıl University Faculty of Medicine, Gaziosmanpaşa Hospital, Istanbul, Turkey; ²⁰ Chief of the Turkish Pediatric BMT Group, Istanbul, Turkey

Background: Relapse is an important cause of death after hematopoietic stem cell transplantation (HSCT) for acute leukemia which represents a challenging management problem.

Methods: A questionnaire was sent to all Turkish pediatric transplant centers reporting leukemic relapse in a patient cohort of 938 HSCT for acute leukemia registered by the Turkish Pediatric Bone Marrow Registry between 1^st of January, 2004 and 31^st of December, 2015. Evaluable data was obtained in 255 patients (154 ALL, 94 AML, 5 bilineage leukemia and 2 indifferentiated leukemia) who underwent allogeneic HSCT and then relapsed.

Results: Median interval from transplantation and relapse was 180 days (range, 0-1740). Two hundred and twenty-eight patients (139 ALL, 82 AML, 5 bilineage leukemia and 2 indifferentiated leukemia) received further treatment after relapse. Of these patients, 105 achieved a complete remission and had a higher cumulative overall survival of 12 months compare with patients not achieving complete remission [49.3% (95% CI 39.5-59.1) vs 7.4% (95% CI 2.6-11.2), p< 0.001]. Factors predictive for successful remission induction after relapse were age (≤10 years at first transplantation) [OR: 2.6, 95% CI 1.5-4.6, p=0.001], isolated extramedullary relapse (OR: 3.5, 95% CI 1.6-7.7, p=0.002) and interval (>6 months) between first HSCT and relapse (OR: 1.8, 95% CI 1.0-3.1, p=0.043). Fifty-five patients, 51 of whom were in complete remission, received a second HSCT. Of these patients, 14 were alive without disease with a median follow-up time of 990 days (range, 528-2982).

Conclusions: The results show that further treatment can prolong survival in pediatric patients who relapsed after HSCT for acute leukemia. In selected patients, long-term leukemia-free survival could be achieved by second HSCT.

Conflict of interest: The authors declare no conflict of interest.

[P048 Figure]

$$$$

P049

Safety and Efficacy of Blinatumomab in Nine Month Old ALL infantile with Persistent MRD Following Chemotherapy

Suleimman Al-Sweedan

King Faisal Specialist Hospital & Research Center, Riyadh, Saudi Arabia

Background: Blinatumomab (Blincyto®, Amgen Inc.), a bispecific CD19-directed CD-3 T-cell engager has produced remarkable responses in various B-cell malignancies, mainly for patients with refractory/relapsed acute lymphoblastic leukemia (ALL) but also in patients with refractory and heavily pretreated B-cell lymphomas. Though it is well established that Blinatumomab is an effective and tolerable treatment option for pediatric ALL, there is an insufficient evidence on its safety and efficacy in patients who are younger a year.

Methods: We herein present a 14 months-old female patient, diagnosed with infantile acute lymphoblastic leukemia (ALL) at the age of 6 months. She presented with a white blood cell count of 1 million, platelets of 15,000, and a hemoglobin of 3.5 g/dL with mild hepatosplenomegaly. She underwent leukapheresis, and peripheral flow cytometry confirmed the diagnosis of ALL. Cytogenetics is positive for MLL. CNS 1.

Results: BM immunophenotype results (03/7/2017) are consistent with positive B-ALL Minimal Residual Disease (0.02% of Mononuclear Cells) after the first cycle, 2nd BM immunophenotype after second cycle was negative.

Conclusions: In conclusion, in our patient, Blinatumomab found to be effective in terms of disease control before allo-SCT and showed an extremely safe profile.

Conflict of interest: NA

P050

Salvage chemotherapy and donor lymphocyte infusion induces durable remissions in AML relapsing following allogeneic stem cell transplantation

Marcin Lubowiecki ¹ , Katalin Balassa ^1,2 , Edmund Watson ¹ , Rebecca Cash ¹ , Daja Barton ¹ , Mandy Ellis ¹ , Lara Rowley ¹ , Denise Wareham ¹ , Claudia Costa ¹ , Rachel Miller ¹ , Rachel Pawson ^1,2 , Angela Hamblin ¹ , Lynn Quek ¹ , Paresh Vyas ¹ , Vanderson Rocha ^1,2 , Andy Peniket ¹ , Robert Danby ¹

¹ Oxford University Hospitals NHS Foundation Trust, Oxford, United Kingdom; ² NHS Blood and Transplant, Oxford, United Kingdom

Background: Allogeneic haematopoietic stem cell transplantation (HSCT) is an established curative treatment for myeloid malignancies. Nonetheless, relapse occurs in up to 40% of patients and outcomes post-allograft remain poor, with overall survival at 1 year reported around 20% (Bejanyan, 2015; Devillier, 2013). There is evidence showing that the use of a cytotoxic approach in selected patients together with donor lymphocyte infusions (DLI) or second allograft can achieve meaningful response rates of around 20-30% (Lim, 2017; Vrhovac, 2016; Schroeder, 2013;).

Methods: We analysed the outcome of AML patients who developed morphological relapse after allogeneic HSCT performed between April 2011 and August 2017 at our centre. Patients underwent 3 monthly bone marrow assessments in the first year post transplant and thereafter when relapse was suspected. Patients with relapsed AML were offered treatment with either intensive AML chemotherapy (CT) or Azacitidine followed by DLI. In appropriate cases palliative chemotherapy or best supportive care was given. Overall survival (OS) data were analysed by the log-rank test and Kaplan-Meier estimates were computed.

Results: Median age at transplantation was 54 years (range 18-69). The majority of patients received reduced intensity conditioning (84%). The stem cells originated from MUD in 67%, MRD in 28%, UCB source in 4% and a haploidentical family donor in 1%. Median follow-up time was 632 days (range from 8 to 2427). 109 patients were transplanted for AML of whom 25 (23%) relapsed at time of follow-up. Post-transplant relapse occurred at a median of 238 days (range 48-825). Following relapse we were able to treat 19 patients (76%) with salvage chemotherapy: 15 patients (60% of relapsed patients) were treated with intensive AML CT (usually FLAG-Ida or FLA). Patients who were considered not to be fit for intensive therapy were treated with Azacitidine (n=4; 16%), although this therapy was not available throughout the study period. Two patients (8%) received palliative chemotherapy (8%) and four (16%) best supportive care. Only one patient underwent a second allograft following salvage chemotherapy. The majority of patients in the salvage CT/Azacitidine group (79%) received DLI following count recovery.

Overall survival for all relapsed patients was 34% at 1 year and 28% at 2 years post-relapse. For those treated with intensive chemotherapy 1-year OS was 50% compared to 10% if another treatment approach was used (p=0.005)

(Figure 1). Of the 19 patients treated with high dose CT or Azacitidine, 53% achieved complete remission and of those, 67% were alive in ongoing complete remission by the time of last follow-up.

Conclusions: We were able to offer salvage chemotherapy to the majority (76%) of patients relapsing following HSCT for AML. Our data suggest that this approach is worthwhile and leads to durable remissions without the use of second stem cell transplant.

References:

Bejanyan, N., et al. (2015) Biol Blood Marrow Transplant, March (21(3)), 454-459.

Devillier, R., et al. (2013) Leukemia & Lymphoma, 54:6, 1228-1234

Lim, A.B.M., et al. (2017 [Epub ahead-of-print]) Internal Medicine Journal, June 19, doi: 10.1111/imj.13522.

Schroeder, T et al. (2013) Leukemia, 27, 1229-1235.

Vrhovac, R., et al. (2016) Bone Marrow Transplantation, 51, 186-193.

Conflict of interest: [All authors]: nothing to disclose

[P050 Figure]

$$$$

P051

Sequential conditioning in unfavorable AML: a single center experience

Magalie Joris ¹ , Amandine Charbonnier ¹ , Bérengère Gruson ¹ , Caroline Delette ¹ , Marie-Noëlle Lacassagne ² , Nicolas Guillaume ³ , Loïc Garçon ⁴ , Jean-Pierre Marolleau ¹ , Delphine Lebon ¹

¹ CHU Amiens, Hématologie Clinique, Amiens, France; ² CHU Amiens, Laboratoire de Thérapie Cellulaire, Amiens, France; ³ CHU Amiens, Laboratoire HLA, Amiens, France; ⁴ CHU Amiens, Laboratoire Hématologie, Amiens, France

Background: Response rates to common salvage regimens reach up to 10-15% in patients with refractory acute myeloid leukemia (AML) or early relapse. Without allogeneic HSCT, the long-term overall survival (OS) for patients with refractory AML hardly reaches 5%. Several studies have demonstrated that salvage chemotherapy with sequential conditioning could reduce leukemia relapse risk with an acceptable toxicity profile for unfavorable acute myeloid leukemias (AML) and myelodysplastic syndromes (MDS). Therefore, we decided to assess this procedure in our center at Amiens University Hospital.

Methods: We conducted a monocentric retrospective study, including 34 patients aged over 18 years undergoing a hematopoietic stem cell transplant (HSCT) with sequential conditioning between January 2012 and December 2016, for refractory or high risk AML or MDS. 32/34 patients received sequential conditioning containing clofarabine (SET-RIC), the 2 others received Amsacrine based conditioning (FLAMSA).

Results: The median age was 48 years (range 27-70). Before conditioning, 19 patients (56%) had active disease and 15 (44%) were in complete remission (CR) including 6 with positive MRD. 18 patients had no or 1 prior treatment and 14 had at least 2 previous courses of chemotherapy. 11 (32%) patients had HLA-identical sibling donors, 14 (41%) match unrelated donors, 4 (12%) mismatch unrelated donors and 5 (15%) haploidentical donors. Majority of patients (85%) received peripheral blood stem cell (PBSC) PBSC with median CD34+ count of 7,71.10⁶/kg (1,84-8,44). Acute GvHD prophylaxis with Ciclosporin A, in combination with Mycophenolate mofetil for unrelated donors, was withdrawal with a median time of 90 days.

With a median follow-up of 32 months, overall survival (OS) at 1 and 2 years was 70,6% and 46%. Median OS was 18 months (0-48 months) and median disease free survival (DFS) was 16 months (0-48 months). 11 patients (32%) experienced relapse after HSCT with a median time from HSCT to relapse of 5 months (1-28 months). 56% of patients presented with grade I-II acute graft versus host disease (GVHD)and 12,5% with grade III IV aGVHD . GVHD free relapse free survival at 1 and 2 years was 44% and 29%. One-year cumulative incidence of disease related death and non-relapse mortality was 17,6% and 11,8% respectively. 14 patients (44%) received immunomodulation with 5 Azacitidine and donor lymphocyte infusion (DLI) if no GVHD occurred within day 120. OS was 86 % in the 14 patients receiving DLI. In univariate analysis, relapse after HSCT and immunomodulation post HSCT (Figure) were significantly associated with overall survival (p< 0,0001 and p=0,016 respectively) but not the disease status before HSCT (p=0,47) neither aGVHD occurrence (p=0,14). In a multivariate analysis, only relapse after HSCT remained predictive of outcome (p=0,034).

Conclusions: According to previous publications, our study show interesting results of sequential conditioning regimen on refractory AML with an acceptable toxicity profile. In addition with post HSCT immunomodulation, we achieve very good OS and DFS whatever cytogenetic and disease status before HSCT. Unfortunately, we failed to identify some predictive factors for outcome like GVHD occurrence which was describe elsewhere probably because of the small size of our cohort.

Conflict of interest: JORIS: nothing to disclose

[P051 Figure]

$$$$

P052

Significance of minimal residual disease before allogeneic hematopoietic cell transplantation for acute myeloid leukemia

Sergey Bondarenko, Boris Afanasyev

Pavlov First Saint-Petersburg State Medical University, R.M.Gorbacheva Memorial Institute of Oncology, Haematology and Transplantation, Saint-Petersburg, Russian Federation

Background: Allogeneic hematopoietic stem cell transplantation (alloHSCT) is an effective therapy for many patients with acute myeloid leukemia (AML). The relapse after alloHSCT is the most common cause of treatment failure. However, multiparameter flow cytometry, cytogenetic and molecular studies in the onset of the disease makes it possible to detect the minimal residual disease (MRD) in AML remission in order to determine the prognosis and the therapy strategy.

Methods: We analysed 195 consecutive patients with AML since 2012 to 2016. Median age was 36(18-70) years. The median follow-up was 23(23-467) months. Disease status before alloHSCT: CR1 - 148 (76%), CR2 - 47(24%) patients. Molecular and genetic low-risk group was defined in 31 (17%), standard risk in 119(59%) and 45(24%) high-risk patients. MRD was not determined (MRD (-)) in 155(80%) patients. A positive result of MRD (MRD (+)) before alloHSCT was in 40 (20%) patients, including mutations, gene overexpression and leukemia associated immunophenotypes.

Results: Disease-free survival (DFS) after alloHSCT in CR1MRD(-) was 67% (95%CI 57-77), and in MRD(+) 54% (95%CI 36-72) (p=.03), whereas in patients with CR2MRD(-) 57% (95%CI 41-73) and CR2MRD(+) - 33% (95%CI 1-88) (p=ns). The relapse incidence (RI) in the CR1MRD(-) was 22% (95%CI 13-32), and in MRD(+) - 36% (95%CI 19-53), whereas in the CR2MRD(-) 17% (95%CI 7-32), and in MRD(+) 66% (95%CI 15-91). With myeloablative conditioning (MAC), the DFS and RI after alloHSCT in CR1 did not depended on MRD: patients with MRD(-) 70% (95%CI 59-81) vs MRD(+) 64% (95%CI 36-92) and 17% (95%CI 7-39) vs 23% (95%CI 12-35) respectively. Whereas after reduced intensity conditioning (RIC), the DFC was higher, and the RI was lower in patients with MRD(-): 67% (95%CI 54-80) vs 50% (95%CI 27-77) (p =.03) and 50% (95%CI 25-70) vs 23% (95%CI 12-35) (p < .000). The cytogenetic risk group (intermediate and high) and the donor type (related and unrelated) had no effect on DFS and RI: 74% vs 63% and 47% vs 38%, 16% vs 19% and 59% vs 63 %; 68% vs 57% and 65% vs 42%, 20% vs 29% and 29% vs 58%. In the absence of chronic graft-versus-host disease(cGVHD), DFS and RI were better in patients with MRD(-) 64% (95%CI 45-83) vs 52% (95%CI 32-72)(p = 0.01) and 34% (95%CI 15-54) vs 44% (95%CI 23-63)(p =.03). In patients with cGVHD, the difference was insignificant: 70% vs 80% and 17% vs 20%. In multivariate analysis, the impact on the RI has had the status of the disease (MRD(+)/MRD(-))(HR2.3, 95%CI, 1.1-4.9, p=.03) and the presence of cGVHD(yes/no) (HR0.3, 95%CI, 0.2-1.0, =.05). DFS was lower with MRD(+) (HR1.8, 95%CI, 1.0-3.5, p =.05).

Conclusions: Immunophenotypic, cytogenetic and molecular data in the onset of the disease is necessary for monitoring the MRD in AML remission. The presence of MRD has a negative effect on DFS and RI after alloHSCT. MRD(+) status in front of the alloHSCT is an indication for the intensification of the conditioning regimen.

Conflict of interest: The authors declare no competing financial interests.

P053

Standard-NORMOXIC versus physiological HYPOXIC culture of AML-patients´ (pts) whole blood (WB) samples with immunemodulatory Kits yields comparable proportions of Dendritic cells and functional results

Fatemeh Doraneh Gard

University Hospital of LMU, Munich, Germany

Background: For a clinical treatment strategy DC_leu could be generated ex vivo and transferred to the pts or alternatively blasts modulated to leukemia-derived DC (DC_leu) in vivo-resulting in T-cell activation. Combinations of GM-CSF+PICIBANIL (I), or +PGE2 (K) or + PGE1 (M) are kits that convert blasts into a 'DC_leu-vaccine' and induce antileukemic T-cell (memory) without induction of blast proliferation in a WB-culture model, in the presence of soluble and cellular components of the pts. To simulate physiological conditions most we compared normoxic (N*; 21%0₂) vs physiological hypoxic (H*; 10% O₂) culture conditions.

Methods: 15 AML- and 8 healthy WB-samples were cultured in parallel in N* and H* conditions treated with “I”, “K” and “M” kits which followed by MLC with the probands' T-cells and blast-cytotoxicity-assays in AML-samples. Cellular compositions (DC-, blast-subtypes and immunestatus) and functional results were compared.

Results: Generation of DC/DC_leu from WB in normoxia vs hypoxia with different kits: Under N* vs H* conditions comparable proportions of DC were generated with Kit I, K, M(9-13%) from 8 healthy- and 10-11% from 15 AML WB-samples. %DC_leu were also comparable (7-10%) as well as %DC_mature (9-11%). Success of DC-generation (>10% DC, >5%DC_leu) was comparable under N* v s H* in healthy and AML-WB. Controls without added kits (“w/o”) failed to produce DC/DC_leu. Furthermore, treatments with Kits did not induce blasts' proliferation under N* or H* conditions.

After mixed lymphocyte culture (MLC) of T-cells with kit-treated AML-WB under N* vs H* with I,K,M comparable % were generated for T_prolif (17-40%) and T_em-_eff (30-48%), however I and M, but not K-treated WB increased % of T_non-naïve (I,M:55-62, K:44%), (I,M: 30-48, K:37%) and T_cm (I,M:7-24, K:3%), under N* and H* significantly.

Antileukemic activity of T-cells (with or w/o previous kit-treatment of WB) in normoxia vs hypoxia: Cytotoxicity-assays after T-cells' MLC with kit-treated (vs untreated) WB showed in 8 of 10 cases with I- and in 6 of 8 cases with M treated WB (in N* vs H*) an improved blast-lysis (by 75-80%). Results obtained with parallel analysed cases under N* vs H* were comparable in 11of18 (61%) cases. In some cases (especially after I-treatment) differences in antileukemic functionality in N* vs H* settings were found.

Conclusions: Cultivation of PB-cells with kits under N* vs H* conditions were comparable (DC-types, T-cell subtypes after MLC, mediation of antileukemic reactivity). This means, that standard N* culture conditions (of WB) simulate reactions in the body well. Therefore production of cellular products (e.g for adoptive cell transfer) under H* is not necessary. Our data show, that pts' treatment with M and I might improve antileukemic reactivity in vivo by a DC_leu-mediated mechanism.

Conflict of interest: [F. D. Gard]: nothing to disclose

P054

Strategy for monitoring minimal residual disease using immunoglobulin clonality in patients with B-cell lymphoblastic leukemia

Nack-Gyun Chung ¹ , Seok Lee ¹ , Irine Jo ² , Seongkoo Kim ¹ , Jae Wook Lee ¹ , Jae-Ho Yoon ¹ , Bin Cho ¹ , Kyungja Han ² , Yonggoo Kim ^2,3 , Myungshin Kim ^2,3

¹ The Catholic Blood and Marrow Transplantation Center, St. Mary's Hospital, College of Medicine, The Catholic University of Korea, Seoul, Korea, Republic of; ² St. Mary's Hospital, College of Medicine, The Catholic University of Korea, Department of Laboratory Medicine, Seoul, Korea, Republic of; ³ Catholic Genetic Laboratory Center, College of Medicine, The Catholic University of Korea, Seoul, Korea, Republic of

Background: Minimal residual disease (MRD) monitoring is one of the most powerful tools to predict the risk of relapse. In this study we validate a feasible MRD monitoring strategy for B-lymphoblastic leukemia (B-ALL) using high-throughput sequencing (HTS) of immunoglobulin (Ig) clonality before implementation into routine clinical tests.

Methods: We selected 60 BM specimens from 34 patients who were diagnosed with B-ALL and followed after treatment at the Catholic Blood and Marrow Transplantation Center, Seoul St. Mary's Hospital, The Catholic University of Korea. Twelve B-ALL with t(12;21)(p13.2;q22.1); ETV6-RUNX1, 11 B-ALL with t(9;22)(q34.1;q11.2); BCR-ABL1 and 11 B-ALL not otherwise specified were included. To quantification of leukemia-specific fusion transcripts, the fusion transcript of ETV6-RUNX1 and BCR-ABL1 was measured via reverse transcription quantitative-PCR (RT-qPCR) using the Real-Q ETV6-RUNX1 and Real-Q BCR-ABL1 quantification kit (Biosewoom Inc., Seoul, Korea), respectively. We used the LymphoTrack^® IGH FR1/2/3 assay panel (InVivoScribe Technologies) to assess the clonal IGH rearrangement via HTS. When clonal IGH rearrangement was negative in the initial specimen after using the primers targeting the conserved framework region (FR) 1 within the variable and joining region, a test using FR2, FR3 and the IGK assay panel (InVivoScribe Technologies) was performed.

Results: All diagnostic BM specimens in the study contained more than 70% leukemic blasts, except for 2 cases (median 92.5%; range 27-99%). IGH clonality was observed in all included B-ALL(n=34) after using primers targeting three framework regions, and the sum of the IGH clonal burden varied(median 79.47%; range 9.47-96.77%). IGK clonality was identified in 70% of patients and availed in cases with low IGH clonal burden. The total IGH clonal burden was significantly correlated with the proportion of leukemic blasts in bone marrow and the quantity of leukemia-specific fusion transcripts including ETV6-RUNX1 and BCR-ABL1. We could recognize the different response of each clone and the emerging clones originating from the trace of Ig rearrangement present in the initial specimen. The IGH clonal burden after induction chemotherapy represented patient outcomes well. The IGH assay also provided information of repertoire diversity of IGH rearrangement for which the low initial value increased after treatment.

Conclusions: The Ig clonality assay via HTS will be a promising tool for MRD monitoring of B-ALL through an adequate strategy to identify all significant clones, monitor individual clones, and determine repertoire diversity. Further studies to accumulate prospective data with clinical implication will help refine the strategy.

Conflict of interest: None

P055

Systematic review and meta-analysis on the role of transplant in infant leukemia

Joy Mburu, Muthoni Mburu, Tony Truong, Arun Narendran

University of Calgary, Hematology Oncology, Calgary, Canada

Background: Infants with leukemia have a poor prognosis despite advances in therapy in childhood leukemia. This is especially so for those with acute lymphoblastic leukemia associated with rearrangements of the KMT2A gene, and are less than 6 months of age at diagnosis. Role of transplant in high risk childhood leukemia's is well defined, while in infants with leukemia impact of transplant has not been clearly defined.

Methods: A review of the published literature was carried out in CINAHL, EMBASE, MEDLINE, PubMed, Cochrane databases and the grey literature, unto the period of September 2016., with inclusion of both retrospective and prospective studies. A Librarian did a search of various databases. Information retrieval from the articles was done by 2 reviewers. Inclusion criteria, also key words were based on PICOS. Population - infants (less than a year old at diagnosis of leukemia), leukemia (including both myeloid and lymphoid). Intervention - chemotherapy or hematopoietic stem cell transplant. Outcome - Event free survival and Overall survival. We excluded studies not in English. We utilized the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) for the study methodology. High risk of bias in articles assessed, this was graded according to Newcastle-Ottawa Scale nonrandomized studies.

Results: There were 1848 articles identified from the database searched. Following duplicate removal, 1202 titles were evaluated for appropriateness. 492 abstracts were screened, and a total of 163 articles were assessed for eligibility. Studies included for analysis of infant leukemia were 31. 20 studies on acute lymphoblastic leukemia and 11 on acute myeloid leukemia. For acute Lymphoblastic leukemia, 8 prospective trials, failed to outline the benefit in both event free survival and overall survival, that is similar to results from 12 retrospective studies. This was considered in the background of the differences in characteristics of the infants, time to transplant and choice of conditioning regimes. For acute myeloid leukemia, only prospective studies were used as they included outcome data beyond 3 years. Infants with AML have good survival data though similar to ALL, there was a small number of patients in the studies like in lymphoblastic leukemia. The event free survival of infants following transplant was no different from those who had continuous chemotherapy. Meta analysis was done using Revman software. In both myeloid and lymphoid leukemia's in infant and comparison of the chemotherapy and transplant group did not show any statistical signifance and there was a high rate of incontinency of the data ranging from 50-95%.

Conclusions: High variation in studies leading to heterogeneity of data, that is including differences in time to transplant, either in complete remission one or two, small sample sizes and choice of chemotherapy and condition regimens make it difficult to carry out a meta- analysis and draw any conclusions. There is a need for randomized control trials to be done for a clear role to be drawn for both myeloid and lymphoblastic leukemia in infants.

No conflict of interest to declare

Conflict of interest: No conflict of interest to disclose

P056

The clinical implication of cytogenetic clonal evolution pattern in relapsed adult acute lymphoblastic leukemia patients

Ji Hyun Lee ¹ , Yoo Jin Lee ²

¹ Dong-A University, Department of Internal Medicine, Busan, Korea, Republic of; ² Kyungpook National University School of Medicine, Department of Internal Medicine, Daegu, Korea, Republic of

Background: Adult acute lymphoblastic leukemia (ALL) patients commonly relapse after allogeneic stem cell transplantation (Allo-SCT) and chemotherapy. ALL relapse has been related to clonal cytogenetic evolution, but there is no study which focuses on the cytogenetic pattern or its clinical implications after treatment.

Methods: Two hundred and twenty-four ALL was diagnosed as ALL in two centers in South Korea between January, 2000 and December, 2016. Among these patients, 181 patients were able to proceed to treatment (80.8%) and among 104 patients (57.5%) achieved complete remission (CR) after allo-SCT or chemotherapy only. 63 patients (60.6%) relapsed after CR and finally 51 patients who had cytogenetic results both at initial diagnosis and at relapse were analyzed.

Results: Eighteen and 33 patients relapsed after allo-HSCT and conventional chemotherapy using chromosome banding analysis combined with fluorescent in situ hybridization. The male to female ratio was 31:20 and the median age in allo-SCT group and conventional chemotherapy only group was 34.0 years (20-54 years) and 46 years (20-76 years) respectively. At diagnosis, aberrant karyotypes were more frequent

in the HSCT than in the chemotherapy cohort (14 of 18; 77.8% versus 16 of 33; 48.5%; P = 0.042).

Clonal changes from diagnosis and relapse were more frequent in the allo-SCT group

(15 of 18; 83.3% versus 20 of 31; 64.5%) compared with the conventional chemotherapy only group (not significant), mostly due to the clonal evolution. Appearance of new ≥3 cytogenetic alterations was more frequent in the allo-SCT group (4 of 18; 22.2% versus 3 of 33; 9.1%, not significant). The mean number of cytogenetic alterations was increased from 1.94 (standard deviation, SD ± 1.47) at diagnosis to 4.39 (SD ± 1.47) at relapse in the allo-SCT group, but in the conventional chemotherapy only group, 2.06 (SD ± 5.12) to 2.35 (SD ± 3.21) (at diagnosis, P = 0.047, at relapse, not significant). Clonal change did not correlate to overall survival.

Conclusions: Clonal change and cytogenetic complexity was more frequently observed in the hematopoietic stem cell recipient group. The change of karyotype did not correlate to the overall survival of adult ALL patients, which might be affected the innate poor prognosis in this group of patients.

Conflict of interest: No conflict of interest

P057

Abstract previously published

P058

The outcome of allogeneic stem cell transplantation in patients with high risk MDS and AML: single center experience

Khalid Halahleh ¹ , Sarah Odaily ² , Abdallah Al Tell ² , Tala Al-Awabdeh ² , Mohammad Ma'kosa ³ , Rula Najjar ⁴ , Rawad Rihani ⁵ , Mayada Abu Shanab ⁶ , Nilly Hussein ⁷ , Eman Khattab ⁸ , Shanta Sharma ⁹ , Ahmad Abu Khader ⁹ , Dana Yousef ¹⁰ , Husam Abu-Jazar ¹ , Abdelghani Tbakhi ¹¹ , Mahmoud Sarhan ¹²

¹ King Hussein Cancer Center, Adult Medical Oncology Hematology, Bone Marrow Transplantation Program, Amman, Jordan; ² King Hussein Cancer Center, Adult Medical Oncology Hematology, Amman, Jordan; ³ King Hussein Cancer Center, Adult Medical Oncology Hematology, Bone Marrow Transplantation, Amman, Jordan; ⁴ King Hussein Cancer Center, Clinical Pharmacy Department, Bone Marrow Transplantation Service, Amman, Jordan; ⁵ King Hussein Cancer Center, Pediatric Medical Oncology Hematology, Bone Marrow Transplantation Program, Amman, Jordan; ⁶ King Hussein Cancer Center, P. Bone Marrow Transplantation Program, Amman, Jordan; ⁷ King Hussein Cancer Center, Amman, Jordan; ⁸ King Hussein Cancer Center, Pediatric Onolcogy Hematology, Bone Marrow Transplantation Program, Amman, Jordan; ⁹ King Hussein Cancer Center, Department of Cell Therapy and Applied Genomics, Amman, Jordan; ¹⁰ King Hussein Cancer Center, Nursing Department, Amman, Jordan; ¹¹ King Hussein Cancer Center, Department of Cell Theray and Applied Genomics, Amman, Jordan; ¹² King Hussein Cancer Center, Bone Marrow Transplantation, Amman, Jordan

Background: Allogeneic hematopoietic cell transplantation (HCT) is an effective, potentially curative treatment for advanced or high-risk AML and myelodysplastic syndrome (MDS) using different intensity conditioning regimens (1). Myeloablative conditioning is associated with a reduced risk of relapse,but does not translate to improved survival due to increased transplant-related mortality(TRM).Reduced-intensity conditioning (RIC) allowed extension of HCT to a much wider patient population by reducing the toxicity without compromising HCTefficacy. In this retrospective study, we report our experience on 64 patients with high risk MDS and AML using different conditioning regimens and report their long term outcome.

Methods: We performed a single center retrospective analysis on 64 adult patients with high risk myelodysplastic syndrome (MDS) or acute myeloid leukemia (AML), who underwent allogeneic HCT at King Hussein Cancer Center between January-2008 till December-2016.Using Kaplan-Meier method, we reported the outcome of patients in regard to overall and progression free survival, cumulative incidence of relapse, TRM, GvHD and identification of predictive factors for poor outcome. 16 patients with MDS, fourty-eight with de novo and secondary AML.41/19/4 were in CR1/CR2,CR3/active MDS respectively. Seventeen were males and 47 were females. Sixty-one received mached related donor, one MUD, one with one allele mismatched related donor and one umbilical cord blood transplantation.Fourty-one and twenty-three received combination of calcineurin inhibitor with either methotrexate and MMF respectively.Fourty-three cases received MAC and twenty-one reduced intensity conditioning before transplantation.

Results: Total of sixty-four patients were included. The median age at time of HCT was 42 years (range: 17-69). All patient received PBSC as graft sourse.90% of patients engrafted on time and 88% had sustained full donor chimerism. 38.2% developed aGvHD grade II-IV, 11% grade III- IV.51.5% developed cGvHD.After a median follow up time of 27.9 months, the probability of 3-year and 5-year progression free survival and overall survival was 49.8 % + 8.2% and 41.5 % + 10.2% and 41% ±6.6% and 34% ±8.3% respectively. The probability of 3-year and 5-year cumulative incidence of relapse 39.1% and 42.8%. The probability of 5-year TRM 55.1%±11.7%.

After analyzing the data, there was significant improved survival outcome in those who had cGvHD (P-0.0001), but there was only trend toward improved progression free survival and overall survival in patients transplanted in CR1 with de novo AML and who had an intermediate-good risk cytogenetics compared with poor risk group (P-0.2).

Conclusions: Allogeneic HCT is potentially curative approach in patients with high risk AML and MDS with improved survival in patients, who developed cGvHD and only trend toward improved survival outcome in patients transplanted in CR1 and had intermediate risk.

Conflict of interest: No conflict of interest

P059

The Role and Mechanism of PTPN21 Mutation in Acute Lymphoblastic Leukemia Relapse after allo-HSCT

Huafang Wang, Ni Zhu, Binsheng Wang, Limengmeng Wang, He Huang

Zhejiang University School of Medicine, Bone Marrow Transplantation Center, The First Affiliated Hospital, Hangzhou, China

Background: Disease relapse is one of the leading causes of death in patients with acute lymphoblastic leukemia (ALL) after allogeneic hematopoietic stem cell transplantation (allo-HSCT). The mechanism of leukemia relapse after allo-HSCT involve in the biological characteristics of leukemia cells .In our previous study, we performed exome sequencing analyses on matched diagnosis and relapse sample to explore genetic basis of relapse. We found that cells responsible for ALL relapse after allo-HSCT showed different gene mutations to the primary leukemia cells at diagnosis and acquired new gene mutations, including PTPN21 gene mutation happened in exon13(exon13: c. 1514 C>A: p. P505Q).

Methods: We designed several small guide RNA(also called sgRNA) located in the exon 13 of PTPN21 near the mutation site, synthesizing the sgRNA-CAS9 plasmids and donor template plasmid by molecule cloning technology and Screening sgRNA-CAS9 plasmid with the highest shear efficiency by T7E1 method. NALM6, a type of PH^-ALL cell line, was used to construct PTPN21 mutation model in vitro. Using Crispr-Cas9 technology, we successfully gained the NALM6 cell line with PTPN21 mutation in exon13(exon13: c. 1514 C>A: p. P505Q). The proliferation of NALM6 cell lines was detected by CCK8 method. Cell cycle, apoptosis and drug resistance of NALM6 cell lines with mutation or not were detected through flow cytometry, and signal pathway associated proteins were detected by Western Blot and Immunofluorescence.

Results:

• 1. Compared with negative control group, the proliferation of NALM6 cell line with PTPN21 above mutation was increased significantly.

• 2. PTPN21 mutation in NALM6 cell line can promote cells' cell cycle transition from G0 phase to S phase.

• 3. There is no significant effect of PTPN21 mutation on the apoptosis and drug resistance of NALM6 cells.

• 4. The mutation of PTPN21 gene in NALM6 cell line enhances the activation level of PTPN21-SRC-ERK pathway, and acts directly or indirectly on downstream P21-CDK2 pathway, thereby regulating cell cycle. See graph below.

Conclusions: PTPN21 gene mutation (exon13: c. 1514 C>A : p. P505Q) can significantly affect the proliferation and the progress of cell cycle in ALL. On the mechanism, PTPN21 mutation on above site may potentially activate downstream SRC-ERK signalling pathway and the P21-CDK2 signalling pathway. which may be closely associated with the recurrence of ALL after allo-HSCT.

Conflict of interest: None of the authors has anything to disclose.

[P059 Figure]

$$$$

Figure 1: A. The specific site of PTPN21 gene mutation (exon13: c. 1514 C>A : p. P505Q).B-D. Flow chart of PTPN21 mutation cell line construction. Figure 2: The comparision in proliferation, cell cycle, apoptosis between negative control and PTPN21 mutation groups. Figure 3. The signal pathway associated proteins of PTPN21 downstream were detected by Western Blot and Immunofluorescence.

P060

The role of high-dose cytarabine as consolidation therapy before allogeneic hematopoietic cell transplantation for acute myeloid leukaemia without core-binding factor in the first complete remission

Yoo Jin Lee ¹ , Sang Kyun Sohn ¹ , Joon Ho Moon ¹ , Hyeoung-Joon Kim ² , Jae Sook Ahn ² , Seo-Yeon Ahn ² , Young Rok Do ³ , Ji Hyun Lee ⁴

¹ Kyungpook National University Hospital, Daegu, Korea, Republic of; ² Chonnam National University Hwasun Hospital, Hwasun, Jeollanamdo, Korea, Republic of; ³ Keimyung University Dongsan Medical Center, Daegu, Korea, Republic of; ⁴ Dong-A Medical Center, Pusan, Korea, Republic of

Background: The optimal number of high-dose cytarabine (HDAraC) consolidation cycles before allogeneic hematopoietic cell transplantation (allo-HCT) for acute myeloid leukaemia (AML) is not fully standardised. This study evaluated the therapeutic value of the number of HDAraC consolidation cycles before transplantation for patients with AML in the first complete remission (CR).

Methods: We retrospectively reviewed the medical records of 241 patients who received allo-HCT for AML without the core-binding factor (CBF) between 1998 and 2014 in Korea. All the patients received induction chemotherapy consisting of idarubicin (12 mg/m²/day on days 1-3) and cytarabine (100 mg/m² every 12 hours on days 1-7). After 186 patients (77.2%) achieved first CR; they were reclassified into three subgroups: the C0 (0 cycle of HDAraC), C1 (1 cycle), and C2 (2 cycles) groups, according to the number of HDAraC consolidation cycles before transplantation.

Results: The median age at diagnosis was 43 years (range, 15-67 years). Thirty-two patients (17.2%) had high-risk cytogenetics and 154 (82.8%) had intermediate-risk cytogenetics by ELN. Among 186 patients in first CR, 21 (11.1%), 116 (61.4%), and 52 (27.5%) patients were in the C0, C1, and C2 groups, respectively. The cumulative HDAraC dose was 3 g/m² (range, 3-9 g/m²) in the C1 and 12 g/m² (6-18 g/m²) in the C2 groups. HDAraC-related toxicity included febrile neutropenia (19.6% vs. 41.7%) and documented infection (19.6% vs. 29.2%, C1 vs. C2, p = 0.041). Ten patients (5.4%) had a relapse before transplantation: 3 (14.3%) in the C0, 3 (2.6%) in the C1, and 4 (7.8%) in the C2 groups (p=0.064).

The median time from CR to allo-HCT was 126 days (range, 60-270 days), 118 days (range, 60-174) in C0, 113 days (range, 84-214) in C1, and 158 days (range, 75-270) in C2. The pre-transplant status was CR1 in 176 (94.6%), er CR2 in 4 (2.2%), and relapse in 6 (3.2%). Ninety-nine patients (53.2%) received myeloablative conditioning regimens. The median dose of CD34+ cells was 3.00×10⁶/kg (range, 0.46-20.60×10⁶/kg). Ninety-nine percent of the patients achieved neutrophil engraftment at a median time of 12 days (range, 9-30 days). The incidence rates of acute and chronic graft-versus-host disease (GVHD) were 48.9% and 41.9%, respectively. The median follow-up duration was 1135 days (range, 140-6774 days). The 2-year relapse-free survival (RFS) was 47.2% ± 11.7%, 71.6% ± 4.4%, and 79.9% ± 5.7% in the C0, C1, and C2 groups, respectively (p=0.032). The 2-year overall survival and non-relapse mortality did not significantly differ according to the number of HDAraC consolidation cycles. The adverse cytogenetic risk group had a benefit from HDAraC consolidation (hazard ratio [HR] = 0.467; 95% confidence interval [CI], 0.221-0.988; p=0.015). In the multivariate analysis, the following factors were associated with RFS: ENL adverse risk group (HR, 2.407; 95% CI, 1.372-5.930; p=0.002), no HDAraC consolidation (HR, 3.008; 95% CI, 1.381-6.550; p=0.006), and chronic GVHD (HR, 0.361; 95% CI, 0.166-0.783; p=0.010).

Conclusions: The use of HDAraC consolidation as a pre-HCT strategy had a positive role in AML without CBF. One or two cycle of HDAraC seemed to maximize the following HCT outcome in AML patients.

Conflict of interest: The authors have declared no conflicts of interest.

[P060 Figure]

$$$$

P061

Treosulfan, fludarabine and cytarabine (FLAT) myeloablative conditioning for autologous stem cell transplantation in elderly patients with acute myeloid leukemia in first complete remission

Massimo Bernardi, Francesca Pavesi, Carlo Messina, Francesca Lorentino, Jacopo Peccatori, Andrea Assanelli, Luca Vago, Bernhard Gentner, Raffaella Milani, Gabriele Casirati, Elisa Sala, Sara Mastaglio, Fabio Ciceri

San Raffaele Scientific Institute, Milano, Italy

Background: Consolidation of complete remission (CR) with autoHSCT is an option for patients (pts) with AML, in particular for elderly ones, as an alternative to alloHSCT. Up to now, no conditioning regimen for autoHSCT has proved preferable in terms of toxicities and improvement of leukemia free survival. Treosulfan is an alkylating agent which demonstrated to achieve maximum disease control with minimal toxicity in combination with fludarabine prior to alloHSCT in pts with high risk or relapsed/refractory AML. We designed a new conditioning regimen including treosulfan combined with fludarabine and cytarabine (FLAT) and tested prior to autoHSCT in AML pts. Data on feasibility and outcome are here reported.

Methods: Period 7/2006-10/2017, 27 pts with de novo AML, median age 67 (18-76). Cytogenetics: favourable 1, normal 22, intermediate 1, complex 1, not evaluable 2. Molecular (18 pts evaluated): CBFB-MYH11 1, NPM1mut or CEBPAmut alone 8, FLT3ITD 5, negative 4. Prognostic risk (ELN 2017): favourable 8, intermediate 7, adverse 3, not evaluable 9. All pts were in CR1 after a median number of 2 chemo cycles (2-4) before autoHSCT. FLAT regimen: treosulfan 10 gr/sqm for 3 days, fludarabine 30 mg/sqm for 5 days, cytarabine 2 gr/sqm for 5 days, PEG filgrastim 1 s.c. vial after autoHSCT. Graft: PBSC, median CD34+ 6x10⁶/kg of BW (3.8-8.5). Time from CR1 to autoHSCT: median 92 days (44-135).

Results: 26 pts were alive in CR1 at day +30, 1 patient died, in G4 pancytopenia, at day +31 because of invasive fungal infection (IFI). Median time to hematopoietic recovery: neutrophils (> 500) day +11 (10-38), platelets (> 20.000) day +21 (13-103). Extra-hematologic toxicities (CTCAE v4.0): median grade 2 (0-4), see also table 1. At last follow up (FU) 13 (48%) pts are alive in CR1, with a median FU from autoHSCT of 641 days (33-4083). Thirteen pts have relapsed at a median of 547 days (31-1808) after transplant. Median EFS from autoHSCT is 368 days (33-4083), median OS from diagnosis is 1004 days (165-4258). Fourteen pts were > 65 yrs old (10 older than 70 yrs): 5 (36%) are alive in CR1, with a median FU from autoSCT of 454 days (33-4064). Overall non-relapse mortality was 4%.

Conclusions: Consolidation of CR1 in AML pts with autoHSCT after conditioning with the new myeloablative FLAT regimen is feasibile and well tolerated. Prolonged OS from diagnosis and EFS from transplant have been obtained, also in elderly pts. A phase II study in pts older than 65 is ongoing in our Center to confirm these preliminary data.

Conflict of interest: None of the authors has anything to disclose.

	All pts	Pts ≤ 65 yrs	Pts > 65 yrs
All pts	27	12	15
FUO (N°)	7 (26%)	4 (33%)	3 (20%)
Sepsis (N°)	11 (41%)	4 (33%)	7 (47%)
Gram negative/positive/molds	3/7/1	1/3/0	2/4/1
Extra-Hematologic toxicities (N°)	15 (55%)	7 (58%)	8 (53%)
hepatic/cardiac/neurologic/other	5/5/1/4	2/2/0/3	3/3/1/1
Neutrophils < 500 (N° days, median/range)	11/9-38	11/10-38	10/9-34
Platelets < 20.000 (N° days, median/range)	21/9-103	24/9-100	16/12-103

[ [P061 Table] Table 1]

[P061 Figure]

$$$$

P062

Umbilical Cord Blood Transplantation for Secondary AML: A Retrospective Study from the Acute Leukemia Working Party of the EBMT and Eurocord

Frederic Baron ¹ , Myriam Labopin ² , Annalisa Ruggeri ³ , Mohamad Mohty ⁴ , Didier Blaise ⁵ , Patrice Chevallier ⁶ , Jaime Sanz ⁷ , Nathalie Fegueux ⁸ , Jan Cornelissen ⁹ , Alessandro Rambaldi ¹⁰ , Bipin Savani ¹¹ , Eliane Gluckman ³ , Arnon Nagler ¹²

¹ University of Liege, Liege, Belgium; ² EBMT Paris Office, CEREST-TC, Paris, France; ³ Eurocord, Hopital Saint Louis, Paris, France; ⁴ Hopital Saint Antoine, Universite Pierre & Marie Curie, INSERM, UMRs 938, Paris, France; ⁵ Institut Paoli Calmettes, Marseille, France; ⁶ CHU of Nantes, Nantes, France; ⁷ Hospital Universitario la Fe, Valencia, Spain; ⁸ CHU Lapeyronie, Montpellier, France; ⁹ Erasmus Medical Center Rotterdam, Rotterdam, Netherlands; ¹⁰ Azienda Ospedaliera Papa Giovanni XXIII, Bergamo, Italy; ¹¹ bipin.savani@Vanderbilt.Edu, Nashville, TN, United States; ¹² The Chaim Sheba Medical Center, Tel-Hashomer, Israel

Background: Allogeneic hematopoietic stem cell transplantation is the only curative option for most patients with secondary acute myeloid leukemia (sAML). Umbilical cord blood transplantation (UCBT) is a treatment option for patients (pts) with primary AML who lack an HLA-matched donor. No systematic large analysis has evaluated risk factors and transplantation outcomes of UCBT for sAML. Therefore, the Acute Leukemia Working Party of the EBMT and Eurocord performed a retrospective registry study on sAML pts given UCBT.

Methods: Inclusion criteria included age > or = to 18 years, secondary AML, first UCBT between 2002-2016, first complete remission (CR1) or active disease at transplantation, and no ex vivo manipulation of the UCB.

Results: 146 pts (58% female) met the study inclusion criteria. Previous diagnoses included myelodysplastic syndrome/myeloproliferative disorder (n=97), solid tumor (n=29, including 23 breast cancer pts), lymphoma (n=13), acute leukemia (n=5), chronic lymphocytic leukemia (n=1) and multiple myeloma (n=1). Status at transplantation was CR1 in 97 pts, primary refractory in 30 pts and relapsed sAML in the remaining 19 pts. The proportion of pts with good-, intermediate- and high-risk cytogenetics was 3%, 38%, and 17%, respectively (missing 42%). Fifty-eight pts were given a single CBT (sCBT) and the remaining 88 pts a double CBT (dCBT). Conditioning was myeloablative (MAC) in 51 (35%) pts and reduced-intensity (RIC) in the remaining 95 (65%) pts. Forty percent of the pts received ATG. One hundred and eighteen pts (83%) achieved neutrophil engraftment while 25 (17%) pts failed to engraft (including 13% of patients transplanted in CR1 versus 30% of patients transplanted with active disease (P=0.008). The 100-day cumulative incidences of grade II-IV and grade III-IV acute GVHD were 35% and 19%, respectively. Two- year cumulative incidences of chronic and extensive chronic GVHD were 25% and 12%, respectively. Two-year incidences of relapse and of non-relapse mortality were 25% and 35%, respectively, in pts in CR1 versus 36% (P=0.06) and 49% (P=0.03) in pts with active disease at transplantation. Two-year OS, LFS and GVHD-free relapse free survival (GRFS) were 42%, 40% and 26%, respectively, in pts in CR1 versus 19% (P< 0.001), 16% (P< 0.001) and 12% (P=0.001), respectively, in pts with active disease at transplantation. Factors associated with OS in multivariate analyses included better survival in female recipients (HR=0.55, P=0.004) and worse survival in pts with active disease at transplantation (HR=2.4, P< 0.001).

Conclusions: This is the first analysis of CBT in rather large group of pts with sAML .Our results suggest that CBT may rescue about 40% of patients with secondary AML in CR1 in the absence of an HLA-identical sibling donor. Future attempts should focus on reducing NRM and improving engraftment. Post-transplant strategies aiming in reducing relapse rates in these very high risk group of patients are also warranted.

Clinical Trial Registry: NA

Conflict of interest: No conflicts of interest

P063

Veno-Occlusive Disease (VOD) Characteristics in Pediatric Patients With Acute Myeloid Leukemia (AML) Receiving Gemtuzumab Ozogamicin (GO) Before Allogeneic Stem Cell Transplant (SCT)

Christine Duncan ¹ , Andrew St. Martin ² , Waleska Perez ² , Patricia Steinert ² , Mei-Jie Zhang ² , Deborah Chirnomas ³ , Caroline J. Hoang ³ , Fausto R. Loberiza, Jr ³ , Wael Saber ²

¹ Dana-Farber Cancer Institute, Boston, MA, United States; ² Center for International Blood & Marrow Transplant Research, Milwaukee, WI, United States; ³ Pfizer Inc, New York, NY, United States

Background: VOD has been reported in AML patients receiving GO, an anti-CD33 antibody-drug conjugate. This analysis characterized VOD risk in pediatric patients receiving GO before SCT.

Methods: Data were from a subset of patients randomly selected for research level reporting to the Center for International Blood & Marrow Transplant Research (CIBMTR). Comparisons of the incidence of VOD at 100 days and overall survival were assessed using logistic regression and Cox regression models, respectively. A stepwise model building approach was used, with variables attaining P≤0.05 retained in the final model.

Results: This analysis included 152 pediatric patients who previously received GO and 348 control patients without GO exposure who underwent first allogeneic SCT between 2008-2011. Median (range) age was 8 (< 1-18) years in patients with and without GO exposure. Most patients with/without GO exposure had achieved first complete remission (with, 43%; without, 57%) or ≥second complete remission (with, 38%; without, 27%). Donor source was balanced between patients with/without GO exposure. Overall incidence of VOD and severe VOD at 100 days was higher in patients with (16% and 8%, respectively) than without GO exposure (10% and 3%). Overall survival probabilities were generally comparable in patients with and without GO exposure at 100 days (84% vs 87%), 6 months (78% vs 80%), and 1 year (69% vs 69%) from SCT and at 100 days (64% vs 68%), 6 months (54% vs 62%), and 1 year (39% vs 54%) from VOD onset. Based on multivariate analyses, GO exposure was associated with a significant risk of VOD (OR, 2.15; 95% CI, 1.25-3.70; P=0.01) but not death (HR, 1.13; 95% CI, 0.86-1.49; P=0.39).

Conclusions: In pediatric AML patients, GO exposure before SCT was associated with a higher incidence of VOD. However, GO exposure was not associated with an increased risk of death. Limitations of the study include selection bias associated with the retrospective design. Future analyses will evaluate the impact of dose on VOD risk and survival outcomes in this cohort.

Encore of BMT Tandem 2018. Funding: Pfizer. Data from the Coordinating Center of the CIBMTR are preliminary.

Conflict of interest:

C. Duncan, A. St. Martin, W. Perez, P. Steinert, M. Zhang, W. Saber: nothing to disclose

D. Chirnomas, C. Hoang, F. Loberiza: employees of and own stocks in Pfizer Inc.

P064

With cGVHD but without intervention therapy had better survival in salvage allo-HSCT for AML in NR status

Yue Lu, Yan LI Zhao, Xing Yu Cao, De Yan Liu, Ming Xiong, Jia Rui Zhou, Rui Juan Sun, Zhi Jie Wei, Jian Ping Zhang, Dao Pei Lu

Ludaopei Hospital, Langfang, China

Background: Salvage allo-HSCT appears to be the only possible cure way for refractory or relapsed AMLin NR status. However, leukemia relapse remains a major cause of transplant failure . the outcome still disappointed .In order to improved the outcome, we optimization our transplantation protocol, now we reported the results.

Methods: we retrospectively analyzed 136 patients with refractory or relapsed AML who underwent allo-HSCT from April 2012 to April 2015. The median percent of blasts was 25 (5-90)%, 55 cases of them ≥ 5%~<20%, 47 cases ≥ 20%~<50%, 22 cases ≥ 50%~<80%,12 cases ≥80%.Male to female 85 /51.the median age were 27 (2-60) years. 57 of them are primary refractory and 79 failured in remission after relapsed. 23 patients received MSD-HSCT, 14 received MUD-HSCT and 99 received Haplo-HSCT.the conditioning regimen based with busulfan (3.2 mg/kg per day) for 4 days/ Cy (1.8g/m² per day) for 2 days/ and ATG,19 of them combinated with HD-Ara-c (2~3g/m² per day) for 3 days; 7 combinated with IDA(10~12mg/m² per days) and Ara-c (3g/m²q12h per days) for 3 days respectively; 22 cases combinated with Decitabine (20mg/m² per days and Ara-c (2g/m² per days) for 5 days recpectively ± IDA (10mg/m² per day)for 3 days; 77 cased combinated with Fludarabine (30mg/m² per days) and Ara-c (2g/m² per days) for 5 days recpectively ±IDA (10mg/m² per day) for 3 days; 11 cases combinated with Amsacrine (100mg/m² per day) and Ara-c (2g/m² per day) for 4 days.136 of 29 patients received pre-emptive intervention,15 patients received MRD-directed intervention; 32 patients received chemotherapy and immunotherapy, while other 60 patients without intervention therapy . intervention therapy including donor lymphocyte infused and donor lymphocyte natural kill cell or cytotoxicity lymphocyte cell infused.

Results: The median follow up time were 24 (3-57) months; the 3-years-OS,DFS, Relapsed related mortality, TRM were44.7±4.4%, 43.1±4.4%, 44.4±4.7%and 16.3±3.8% respectively.Univariate analysis of factors with survival showed: lower percent of blasts pre-HSCT has better OS [≥5%~<20% (55.9±6.8%) vs ≥20%~<50% (43.9±7.8%) vs≥50%~<80% (31.8±6.8%) vs≥80% (25.0±4.3%), P=0.038),conditioning regimen including IDA has better OS compared with other [(HD-Ara-c (26.3±10.1%) vs IDA/Ara-c (85.7±13.2%) vs Dec/Ara-c±IDA (59.1±10.5%) vs FLAG±IDA (50.0±5.8%) vs Amsa/Ara-c (18.2±4.2%), P=0.026)], post-HSCT without intervention therapy had best OS [(pre-emptive intervention vs MRD-directed intervention vs intervention after relapsed vs without intervention,P=0.000)], with II~IV aGVHD has lower OS compared to 0~IaGVHD (34.2±7.2% vs 53.6±5.4%,P=0.034), with cGVHD has better OS (53.6±5.6% vs 18.2±4.2%, P=0.000), but there were no significant difference with age (≤14y vs >14y, P=0.752), sex (male vs female, P=0.981), HCT-CI (0-1 vs ≥2, P=0.549), type of disease (Primary vs Secondary, P=0.564), with extramedullary lesions (Yes vs No, P=0.444), Donor-patient sex match (male - male vs male - female vs female - female vs female- male, P=0.823), MNC (×108/kg) (<8.76 vs ≥8.76, P=0.229), CD34 cell counts (×106/kg) (<4 vs ≥4, P=0.512), CD3 cell counts (×108/kg) (<1.67 vs ≥1.67,P=0.748), Donor type (MSD vs MUD vs Haplo-D,P=0.313). Multivariate analysis factor showed: post-HSCT without intervention has higher OS and DFS, with cGVHD not only has higher OS and DFS, but also had lower incidence of relapsed mortality.to analysed the impacted of with cGVHD on the patients who post-HSCT without intervention showed,the patients with cGVHD who has higher OS (86.8±5.5% vs 59.1±10.5%,P=0.01), but the patients with II~IVaGVHD has higher TRM (P=0.007) and lower OS (P=0.002).

Conclusions: the patients With cGVHD but without intervention therapy had better survival in salvage allo-HSCT for AML in NR status.

Conflict of interest: nothing to disclose

P065

WT1-specific T-cells in healthy donors and leukemia patients and the possibility of improved detection of the immune response using WT1 peptide-pulsed dendritic cells

Julia Panchenko ^1,2 , Viсtoriia Kiseleva ¹ , Vladimir Zhogov ¹ , Natalya Khripkova ¹ , Pavel Trakhtman ¹ , Nikolay Starostin ¹ , Michael Maschan ¹ , Elena Osipova ¹

¹ Dmitry Rogachev National Center of Pediatric Hematology, Oncology and Immunology, Moscow, Russian Federation; ² Moscow Institute of Physics and Technology, Moscow, Russian Federation

Background: WT1 antigen is a promising target for cancer immunotherapy because it has a high immunogenicity and is hyperexpressed in various types of leukemia and solid tumors. The complexity of determining WT1-specific T-lymphocytes is associated with a low frequency of these cells in the peripheral blood. In this regard, the study of the possibility of improved detection of the immune response against WT1 antigen may assist in development of immunotherapeutic approach to cancer.

Methods: The frequency of WT1-specific T cells in peripheral blood of healthy donors (n = 68) and leukemia patients (AML, n=4, ALL, n=6) after allogenic stem cells transplantation were monitored by dextramer staining and enzyme-linked immunospot (ELISPOT) assays with the WT1-overlapping peptide pool. Samples with more than 5 specific lymphocytes per 500 000 mononuclear cells considered to be positive.

Monocyte-derived dendritic cells (DC) were generated from mononuclear (MNC) adherent cells from healthy donors (n=10) in the presence of IL-4 and granulocyte-macrophage colony-stimulating factor. On day 7, immature DC were matured with a cytokine cocktail consisting of IL-4, granulocyte-macrophage colony-stimulating factor, tumor necrosis factor (TNF)α, harvested after 48 h, pulsed with WT1 peptide pool and washed from peptide. Control DC were generated using the same method without incubation with WT1. The mature phenotype of generated DC were determined by Flow Cytometry with HLA-DR, CD83, HLA-ABC, CD86 antibodies. Donor MNC were cultivated alone or with autologous DC (WT1 specific and control) to analyze the number of WT1 specific T-lymphocytes by ELISPOT-assay.

Results: WT1 specific T-lymphocytes above 0,001% from mononuclear cells were detected in 22 of 68 (32,4%) donors. The frequency of anti-WT1 lymphocytes in healthy donors ranged from 6 to 328 per 500 000 MNC (median 16,5). In 6% of donors WT1 specific lymphocytes were detected in more than 0,01% of MNC. Specific anti-WT1 T-cells were also detected in periferical blood of 2 of 6 patients with ALL (12 and 20 WT1 specific lymphocytes per 500 000 MNC), and 2 of 4 of AML (8 and 45 WT1 specific lymphocytes per 500 000 MNC) patients after allo HSCT.

After 24-hour incubation of MNC with WT1-pulsed dendritic cells (10MNC: 1DC) the number of anti-WT1 T-lymphocytes (as detected by IFN gamma ELISPOT) increased by 2.8 to 20.3 times. Control cultures with DC only (pulsed or not pulsed) were negative in ELISPOT-assay. Such a large range of values reflects the heterogeneity of the population of donors in relation to the immune response to the WT1 antigen, which emphasizes the importance of personalized approach when planning immunotherapy of tumors. Using cultural inserts (Carrier Plate System, Nunc) for 12-well plate we have demonstrated that WT1-specific DC influence the T-lymphocytes reactivity to antigen by direct contact.

Conclusions: The incubation of MNC with WT1-specific DC leads to a significant increase of T-cell immune response against the WT1 antigen. These results can be used to develop new approaches of dendritic cell vaccination against WT1 antigen.

Conflict of interest: Nothing to disclose

Aplastic anaemia

P066

Allogeneic bone marrow transplantation for children and adolescents with severe aplastic anemia in Brazil: A multicenter study on behalf of the Brazil-Seattle Consortium Study Group

Luiz Darrigo Jr ¹ , Vergilio Coulturato ² , Mair Souza ² , Ederson Mattos ² , Gisele Loth ³ , Rodolfo Calixto ⁴ , Adriana Seber ⁵ , Victor Zecchin ⁶ , Liane Daudt ⁷ , Alessandra Paz ⁷ , Rita Barbosa Tavares ⁸ , Leonardo Arcuri ⁸ , Antonio Macedo ⁹ , Ana Karine Vieira ⁹ , Cilmara Kuwahara ¹⁰ , Roseane Gouveia ⁵ , Lisandro Ribeiro ¹¹ , Juliana Fernandes ¹² , Mary Flowers ¹³ , Ricardo Pasquini ³ , Carmem Bonfim ³

¹ Ribeirão Preto Medical School, University of São Paulo, Ribeirão Preto, Brazil; ² Hospital Amaral de Carvalho, Jahu, Brazil; ³ Universidade Federal do Paraná, Curitiba, Brazil; ⁴ Real Hospital Português de Beneficência, Recife, Brazil; ⁵ Hospital Samaritano de São Paulo, São Paulo, Brazil; ⁶ Graacc, São Paulo, Brazil; ⁷ Hospital das Clínicas de Porto Alegre, Porto Alegre, Brazil; ⁸ INCA, Rio de Janeiro, Brazil; ⁹ Universidade Federal de Minas Gerais, Belo Horizonte, Brazil; ¹⁰ Hospital Pequeno Príncipe, Curitiba, Brazil; ¹¹ Hospital Nossa Senhora das Graças, Curitiba, Brazil; ¹² ITACI, São Paulo, Brazil; ¹³ Fred Hutchinson Cancer Research Center, Seattle, WA, United States

Background: Idiopathic severe aplastic anemia (SAA) is a rare life-threatening disease and bone marrow transplantation (BMT) from a matched sibling donor is the treatment of choice for pediatric patients. The aim of this study is to determine major transplants outcomes after BMT for children with SAA transplanted in Brazil.

Methods: This is a retrospective study that analyzed 115 consecutive patients under the age of 18 years who underwent a first allogeneic BMT for SAA between 2010 and 2014 in 12 Brazilian BMT centers. The median age was 10 years and 69% were male. Graft was an unmanipulated marrow from a matched related (MRD; sibling n=66 and other relative n=3) or unrelated (URD; HLA-matched n=37, and mismatched n=9) donor. Conditioning regimen widely varied among institutions. In MRD transplants, 44% received Cyclophosphamide ± r-ATG, 33% Busulfan-based, and 20% Fludarabine-based regimen. Most URD BMT (63%) were Fludarabine-based and 34% Busulfan-based. In vivo T-cell depletion included r-ATG (n=31) or Campath (n=4) in 50% of the transplantswith MRD and all but one from URD. Calcineurin inhibitor and methotrexate were used in 89% of BMT. Acute and chronic graft-versus-host disease (GVHD) were graded using Glucksberg and NIH scales, respectively.

Results: With a median follow-up of 4.5 years, 87 of 115 patients are alive. The 4-year overall survival (OS) was 77% (95% CI: 70-86%), not significantly different from MRD (81%) or URD (71%; p=0.11). Seven patients died before Day+21 and were not evaluable for engraftment. No MRD had primary graft failure (GF), in contrast with 7/46 with URD (cumulative incidence 16%; 95% CI 4-26%). Out of the 101 patients who engrafted, 11 had graft rejection, after a median of 228 days (cumulative incidence: 11% (95% CI: 5-17%), not significantly different between donor types. Thirteen patients had a 2nd BMT and 7 were alive at last follow-up. The cumulative incidence of grades III-IV acute GVHD was 7% (95% CI:2-12%) and of chronic GVHD at 4 years was 17% (95% CI:10-24%), with no significant difference between donor types. Twenty-seven patients died after a median of 2.9 months after BMT. Major causes of death included infections (n=16), bleeding (n=3), secondary neoplasia (n=2), and GVHD (n=2). The 6-month transplant related mortality was higher after URD than MRD (OR=2.4, 95%CI: 0.9-6.4) although it did not reach significance (p=0.09).

Conclusions: The OS of pediatric patients transplanted for SAA in Brazil is 77% (95% CI: 70-86%), similar among MRD or URD, but lower than expected in other international reports. Our group is working to standardize a national pediatric BMT protocol to improve results of allogeneic BMT for pediatric SAA in Brazil.

Conflict of interest: There is no conflict of interest to report.

P067

ALLOGENEIC HEMATOPOEITIC STEM CELL TRANSPLANTATION (HSCT) AFTER FLUDARABINE BASED CONDITIONING IN 61 PATIENTS WITH FANCONI ANEMIA

Sara Zerkout, Malek Benakli, Nadia Rahmoune, Hanane Bouarab, Dina Ait Ouali, Mounira Baazizi, Farih Mehdid, Rachida Belhadj, Chahrazed Benchouk, Kahina Tarhi, Rose-Marie Hamladji, Redhouane Ahmed Nacer

Hematology and Bone Marrow Transplantation Department, Algiers, Algeria

Background: Fanconi anemia (FA) is the most common constitutional aplasia with haematological and extra-haematological abnormalities, and high risk of transformation into myelodysplasia (MDS) or acute leukemia (AL). The only curative treatment is HSCT. We report the results of allogeneic HSCT in 61 patients (pts) with FA performed during a period of 15 years.

Methods: From January 2002 to December 2016, 61 pts with FA underwent allogeneic HSCT including 55 sibling HLA identical, 05 phenoidentical and one from an unrelated cord blood. The median age is 11 years (4.5 to 25) of whom 34 (55%) have less than 10 years. The sex ratio (M / F) is 1.1. The average diagnosis-transplant time is 26 months (4 to 102). Fourty-eight pts (78.7%) had a transfusion history of which 06 (9.8%) had more than 20 red blood units transfusions. Thirteen (22.4%) pts received previous treatment with androgens, corticosteroids or a combination of both. Two types of conditioning regimen were used successively in time: fludarabine-cyclophosphamide-antithymocyte globulin (FLU-CY-ATG: 44 pts) and fludarabine-cyclophosphamide (FLU-CY: 17 pts). GVHD prophylaxis included association of ciclosporin and methotrexate.

The grafts used were peripheral blood stem cells in 56 pts (91,8%) with an average CD34 + cell count: 9.42 106/kg (2.06-36.31), bone marrow transplant in 4 pts with an average level of nucleated cells 3.67 ×108/ kg (1.83-4.96) and cord blood with CD34 + cells: 1.33 ×105/ kg and NC: 7.73 ×107/ kg. At July 2017, the minimal follow up is 7 months and maximal is 187 months.

Results: The median time to graft (PNC > 0.5 x 10⁹/l) was 14 days (7- 36). One patient had severe veino-occlusive disease (VOD). Acute GVHD was observed in 9 pts (14.7%), of which 03 (4.9%) grade III-IV. Chronic GVHD in 13 pts (25.5%), including 06 of extensive form. CMV reactivation is noted in 9 pts (16%). Six cases (10,9%) of rejection were observed of which 2 are alive, one after a boost and the other after a second allograft. After a median follow-up of 98 months (7-186), 45 pts (74%) are alive with a strictly normal blood count and 16 pts (26%) died, of which 11 (18%) due to TRM (infection 5, GVHA 2, capillary leak syndrome 1, hemorrhage 2, VOD 1). The 1-year TRM according to the type of conditioning is estimated at 21% and 11.3% respectively. One pt showed malignant transformation (acute leukemia at 5 months after rejection). After 15,5 years, Overall survival (OS) and event-free survival (EFS) are 72.5% and 62.8% respectively. The OS according to the type of conditioning at the same time (72 months) is 68,2% (FLU-CY-ATG) and 87% (FLU-CY) without significant difference (0,2 difference (p:10-5).

Conclusions: Allogeneic HSCT remains the only therapeutic alternative in FA. It permits to restore hematopoietic and to prevent progression to MDS or AL. The FLU-CY conditioning, adopted from 5 years in our center, is well tolerated and has given better results in terms of EFS and TRM.

Conflict of interest: nothing to disclose

P068

Effect of stem cells on Allogeneic Hematopoietic Stem Cell Transplantation in Adult Patients with Idiopathic Aplastic Anemia

Hawk Kim ¹ , Kyoo-Hyung Lee ² , Sang Kyun Sohn ³ , Seok Jin Kim ⁴ , Inho Kim ⁵ , Sung Hyun Kim ⁶ , Yong Park ⁷ , Jae Yong Kwak ³ , Min Kyoung Kim ⁸ , Sung Hwa Bae ⁹ , Ho Jin Shin ¹⁰ , Jong Ho Won ¹¹ , Won Sik Lee ¹² , Yunsuk Choi ¹³

¹ Gachon University Gil Medical Center, Gachon University College of Medicine, Division of Hematology, Incheon, Korea, Republic of; ² Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea, Republic of; ³ Kyungpook National University Hospital, Daegu, Korea, Republic of; ⁴ Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea, Republic of; ⁵ Seoul National University Hospital, Seoul, Korea, Republic of; ⁶ Dong-A University Medical Center, Busan, Korea, Republic of; ⁷ Korea University Anam Hospital, Seoul, Korea, Republic of; ⁸ Yeungnam University Medical Center, Daegu, Korea, Republic of; ⁹ Daegu Catholic University Hospital, Daegu, Korea, Republic of; ¹⁰ Pusan National University Hospital, Busan, Kosovo, Republic of; ¹¹ Soon Chun Hyang University Hospital, Seoul, Korea, Republic of; ¹² Inje University Busan Paik Hospital, Busan, Korea, Republic of; ¹³ Ulsan University Hospital, University of Ulsan College of Medicine, Division of Hematology and Cellular Therapy, Ulsan, Korea, Republic of

Background: Many sAA patients receive hematopoietic stem cells from PBSC even though they receive stem cell from MSD. We need to know which additional factors are affecting on transplantation outcomes when BM or PBSC is infused.

Methods: We retrospectively investigated the effect of stem cells on allogeneic hematopoietic cell transplantation (alloHCT) in idiopathic adult aplastic anemia (AA). Mismatched donors (MMD) included haplo-identical family donor, mismatched unrelated donor or partially matched family donor. Alternative donor (AD) referred either MUD or MMD.

Results: Total 267 patients were included in this analysis. BM was associated with low incidence of acute graft versus host disease (GvHD; p< 0.001) but not in G3/4 acute GvHD (p=0.427). BM, however, had no impact on other transplantation outcomes including chronic GvHD (p=0.673) and primary/secondary graft failure (p=0.774). Higher stem cell dose had no impact on other transplantation outcomes except for low incidence of extensive chronic GvHD in MSD (p=0.025). Multivariate analysis on overall survival in MSD revealed that only Age at alloHSCT< 31 years old (p=0.010) and prior platelet transfusion less than 86U (p=0.046) were the favorable prognostic factors. On the other hand, multivariate analysis in AD revealed that higher stem cell dose (HR=2.596; 95% CI 1.020-6.609; p=0.045) was the only significant favorable factors on overall.

Conclusions: PBSCs can be preferable in AD because higher stem cell dose can be easily achieved for longer overall survival in expense of acute GvHD. However, BM stem cells are preferred in MSD.

Conflict of interest: There is no conflict of interest to disclose

P069

Haploidentical hematopoietic stem cell transplantation with selection of CD34+ and CD45RA+ cells depletion as novel approach for Severe Aplastic Anaemia in pediatric patients

Mercedes Gasior Kabat ¹ , Laura Sissini ² , David Bueno ³ , Raquel De Paz Arias ¹ , Isabel Badell ⁴ , Montserrat Torrent ⁵ , Sergi Querol ² , Sonsoles Sanroman ¹ , Diego Plaza ⁶ , Antonio Marcos ¹ , Ana Belen Romero ¹ , Ana Sastre ⁶ , Jaime Valentin ⁷ , Antonio Pérez-Martínez ⁶

¹ Hospital Universitario La Paz, Hematology, Madrid, Spain; ² Hospital Universitario Sant Pau, Hematology, Barcelona, Spain; ³ Hospital Universitario La Paz, Pediatric Hemato-oncology, Madrid, Spain; ⁴ Hospital Universitario Sant Pau, Pediatric Onco-Hematology, Barcelona, Spain; ⁵ Hospital Universitario Sant Pau, Pediatric Hemato-Oncology, Barcelona, Spain; ⁶ Hospital Universitario La Paz, Pediatric Onco-Hematology, Madrid, Spain; ⁷ Hospital Universitario La Paz, IDIPAZ, Madrid, Spain

Background: Haploidentical stem cell transplantation (HSCT) has been occasionally carried out in the setting of aplastic anaemia (AA) for patients lacking proper-matched donor. Haplo-HSCT in the paediatric setting requires T-cell depletion, which carries profound immunosuppression and high risk of graft failure. Naïve T-cells identified by CD45RA expression are believed to cause graft-versus-host-disease (GvHD), while CD45RA- T-cells are memory cells that provide anti-infection and anti-tumoral effects. Depleting CD45RA+ naïve cells and retaining memory T-cells in the graft is a novel approach to haploidentical HSCT for children with severe AA.

Methods: Three children with severe AA received CD45RA-depleted haplo-HSCT following non-myeloablative conditioning. Cell-selection performed on G-CSF-mobilized peripheral blood. Two cellular products obtained using CliniMACS device. Patients (detailed in Table 1) have received 2 haploidentical cellular products. First product was result of a CD34-positive selection; second product was the result of CD45RA depletion.

Results: Three patients received haplo-HSCT from February to July 2017. Two of 3 patients were male. Median age at transplantation was 11 years (range 9-12). Graft composition detailed in Table 1. Neutrophil and platelet engraftment achieved at median of 10 days (range 9-11) and 12 days (range 10-13) post-transplant respectively. No graft failure was seen. Two patients developed aGvHD>grade II with gastrointestinal tract and skin involvement, all steroids responsive. One patient presented clinical features of probable cGvHD. All patients (55.5%) remain alive with mean follow-up 129 (range 70-228) days post-transplant. Chimerism at day +30 was full donor. All patients presented cytomegalovirus reactivation, none progressed to CMV disease. No severe infectious complications seen.

Conclusions: CD45RA-depleted haplo-HCT showed a rapid and sustained neutrophil and platelet engraftment as well as a full donor chimerism. Acute GvHD was esteroid-responsive. No severe infections were seen. CD45RA-depleted haplo-HCT can be a novel approach for transplantation in patients with SAA lacking proper-matched family or unrelated donor.

Conflict of interest: None

	Patient 1	Patient 2	Patient 3
Sex	Male	Female	Male
Age at transplantation (years)	9	12	11
Donor	Mother	Mother	Brother
Graf composition
CD34+ (x106/Kg)	6,26	7,73	6,41
CD45RA+/Kg	1,09x102	0,5 x102	1,8 x103
CD45RO+/Kg	5 x107	5 x107	8,34 x106
Log CD45RA+ depletion	>5	4,69	3,83
CD20+/Kg	9,3 x104	3,3 x104	1,40 x105

[ [P069 Table] Table 1. Patient details and grafts composition]

P070

Hematopoietic Stem cell transplant in Aplastic anemia in patients older than 40 years: The experience of the Spanish Group of Hematopoietic Stem Cell Transplant (GETH)

Lucrecia Yañez ¹ , Jaime Sanz ² , Irene Garcia Cadenas ³ , Lourdes Vazquez ⁴ , Ildefonso Espigado ⁵ , Isabel Sanchez Ortega ⁶ , Mi Kwon ⁷ , Carlos Solano ⁸

¹ Hospital Universitario Marqués de Valdecilla, Santander, Spain; ² Hospital Universitario la Fe, Valencia, Spain; ³ Hospital Santa Creu I Sant Pau, Barcelona, Spain; ⁴ Hospital Universitario de Salamanca, Salamanca, Spain; ⁵ Hospital Universitario Virgen del Rocío, Sevilla, Spain; ⁶ ICO - Hospital Duran i Reynals, Barcelona, Spain; ⁷ Hospital Gregorio Marañón, Madrid, Spain; ⁸ Hospital Clinico de Valencia, Hematology, Valencia, Spain

Background: Although survival after HLA identical bone marrow transplant or immunosuppressive therapy (IST) is similar in patients older than 40 years (55% vs. 58%), current guidelines recommend the latter for these group of patients (Bacigalupo, Blood 2017).

The aim of this retrospective study is to evaluate the role of hematopoietic stem cell transplant (SCT) in older patients with Aplastic anemia and to know if previous IST affects negatively SCT ouctomes.

Methods: Seven Spanish centers reported a total of 28 patients who underwent a SCT between 1998 and 2015. Patients with 2^nd transplant or further were excluded. Median age at diagnosis was 48 years (43-59) and 16 (57.1%) were male. Seventeen patients had information at diagnosis about etiology (idiopathic 13, secondary to toxic or drug 1, secondary to other disease 1), severity (moderate 3, severe 6, very severe 8), chromosome analysis (normal 8, abnormal 1, not done or failed 8) and clinical features (hemorrhage 5, systemic infection 5). Two patients had a PNH clone ≥ 5%.

Results: Previous IST: At least one course of cyclosporine and ATG was given to 23 patients (82.1%). Time from IST to SCT was 6 months (4-11). Transplant characteristics: Median age at SCT was 50 years (45-60). Despite it was not statistically significant, time from diagnosis to SCT was delayed more than 3 months in those patients who received IST (6.8 vs. 3.1). The conditioning regimen mainly consisted of cyclophosphamide alone (16) or combined with fludarabine (10). Three patients (10.7%) did not receive ATG. The GVHD prophylaxis was quite different between centers being the most frequent the combination of cyclosporine with methotrexate (5), mycophenolate (4) or corticosteroids (7). Stem cell source was bone marrow in 15 cases (53.6%) being the donor an identical sibling in 21 (75.0%). One patient underwent a haploidentical transplant. Transplant complications: Two patients (7.1%) did not engraft and 2 patients presented an early graft loss (7.7%). Nine patients (34.6%) developed acute GVHD (3 grade I, 5 grade II, 1 grade III) and six (26.1%) developed chronic GVHD (4 limited, 2 extensive). Outcome: Five patients (17.8%) died before day +100 due to infections in 4. With a median follow up of 59 moths (8-89), 16 patients (57.1%) are alive, 12 in complete remission and 4 in partial remission. The probability of survival was higher in those who did not receive previous IST (p=0.079), Figure 1.

Conclusions: In line with prior reports, our study recognizes SCT as a suitable option for older patients diagnosed with aplastic anemia. The better outcomes observed in patients who received an upfront SCT from an identical sibling need further confirmation in larger series.

Conflict of interest: The authors have nothing to disclose.

[P070 Figure]

$$$$

P071

Abstract previously published

P072

Rescue Haploidentical Transplantation for Aplastic Anaemia Following Primary Graft Failure

Santanu Sen ¹ , Sameer Tulpule ²

¹ Kokilaben Dhirubhai Ambani Hospital, Department of Paediatric Oncology & Stem Cell Transplantation, Mumbai, India; ² Kokilaben Dhirubhai Ambani Hospital, Department of Stem Cell Transplantation, Mumbai, India

Background: Primary graft failure remains one the important causes of mortality and morbidity in cases of stem cell transplantation. We report on a case of Aplastic Anaemia where an immediate rescue haploidentical stem cell transplantation from a different donor resulted in a successful engraftment. We suggest that an immediate transplantation with repeat conditioning with chemo-radiation therapy can result in a successful engraftment in these cases of primary graft failure.

Methods: A 3yr old male child was diagnosed to have Severe Aplastic Anemia. Unfortunately there were no matched sibling donors and matched unrelated donor search did not yield any acceptable matches. We elected to undertake a haploidentical transplant using his fathers stem cells. Father was 5/10 HLA matched by high resolution typing and was DSA negative and had the same blood group. He was conditioned with ATG (10mg/kg/day x 3 days) Fludarabine (30 mg/kg x 5 days), Cyclophosphamide (14.5 mg/kg x 2 days), and low dose TBI at 200cGy and GVHD prophylaxis with Post Transplant Cyclophosphamide. CD34 dose used was 8.9x10^6/kg.

Unfortunately he failed to engraft with counts remaining very low at Day + 21 and a marrow examination was severely hypocellular. Hence at day +21 we undertook a second rescue haploidentical transplant for the primary graft failure. As he had rejected his fathers stem cells, we decided not to use the father the second time around as we surmised that the patient might have developed effective antibodies again him. Hence this time we decided to re-transplant him the second time using his mother stem cells. Though he was profoundly cytopenic & neutropenic, we decided to give him full repeat conditioning with the same protocol as we surmised that he still had a active immune system that had rejected the primary graft. Repeat conditioning using the same protocol was started from Day + 22 and mother's stem cells was infused on Day +28.

Results: He engrafted successfully by Day +12 of the second transplant and though the transplant was complicated by a cutaneous fungal infection and CMV reactivation, child was discharged home with 100% donor engraftment. Currently the child is Day + 182 and though he had a mild gut GVHD, this was easily managed with oral budesonide.

Conclusions: We suggest that a urgent second haploidentical transplant using repeat conditioning can be used successfully in cases of primary graft failure even in the face of severe cytopenia.

Conflict of interest: S. Sen: Nothing to disclose

P073

Results of allogeneic stem cell transplantation in patients affected by Shwachman-Diamond syndrome. A retrospective study of SAA-EBMT

Simone Cesaro ¹ , Gloria Tridello ¹ , Christian Kratz ² , Marta Pillon ³ , S Halkes ⁴ , Edoardo Lanino ⁵ , Regis Peffault De Latour ⁶ , C Diaz De Heredia ⁷ , Robert Wynn ⁸ , Johann Greil ⁹ , Franco Locatelli ¹⁰ , Paul Veys ¹¹ , Anne Uyttebroeck ¹² , Per Ljungman ¹³ , Patrice Chevalier ¹⁴ , Marc Ansari ¹⁵ , Isabel Badell ¹⁶ , Tayfun Gungor ¹⁷ , Salim Rahuman ¹⁸ , Johanna Tischer ¹⁹ , Cristina Tecchio ²⁰ , Nigel Russel ²¹ , Alicja Chybicka ²² , K Kallay ²³ , Owen Smith ²⁴ , B Afanasyev ²⁵ , J Stein ²⁶ , Jan Styczynski ²⁷ , Paul Bosman ²⁸ , Carlo Dufour ⁵

¹ Azienda Ospedaliera Universitaria Integrata, Verona, Italy; ² Hannover Medical University, Hannover, Germany; ³ Clinica di Oncoematologia Pediatrica, Padova, Italy; ⁴ Leiden University Medical Hospital, Leiden, Netherlands; ⁵ Istituto G. Gaslini, Genova, Italy; ⁶ Hopital St. Louis, Paris, France; ⁷ Hospital Vall d'Hebron, Barcelona, Spain; ⁸ Central Manchester NHS Trust, Manchester, Manchester, United Kingdom; ⁹ University of Heidelberg, Heidelberg, Germany; ¹⁰ IRRCS Ospedale Pediatrico Bambino Gesu, Rome, Italy; ¹¹ Great Ormond Street Hospital, London, United Kingdom; ¹² University Hospital Leuven,, Leuven, Belgium; ¹³ Karolinska University Hospital, Stockholm, Sweden; ¹⁴ CHU Nantes, Nantes, France; ¹⁵ Hopitaux Universitaires de Geneve, Geneva, Switzerland; ¹⁶ Hospital de la Santa Creu i Sant Pau, Barcelona, Spain; ¹⁷ University Children's Hospital, Zurich, Switzerland; ¹⁸ Royal Liverpool University Hospital, Liverpool, United Kingdom; ¹⁹ Klinikum Grosshadem, Munich, Germany; ²⁰ University Hematology, Verona, Italy; ²¹ Nottingham University Hospitals NHS Trust, Nottingham, United Kingdom; ²² Wroclaw Medical University, Wroclaw, Poland; ²³ United St. Istvan and St. Laszlo Hospital, Budapest, Hungary; ²⁴ Our Ladys Childrens Hospital, Dubllin, Ireland; ²⁵ Pavlov Medical University of St. Petersburg, St. Petersburg, Russian Federation; ²⁶ Schneider Children's Medical Center, Petach-Tikva, Israel; ²⁷ University Hospital, Collegium Medicum UMK, Bydgoszcz, Poland; ²⁸ EBMT Data Office, Leiden, Italy

Background: Shwachman-Diamond syndrome (SDS) is a rare multisystemic disease that affects 1/77.000 children. SDS is the third cause of congenital bone marrow dysfunction, after Fanconi anemia and Blackfan-Diamond anemia, and the second cause of pancreatic insufficiency after cystic fibrosis. The transmission of SDS is autosomal recessive. The long-term survival of SDS patients is threatened by the occurrence of bone marrow failure that can evolve to severe aplastic anemia, myelodysplastic syndrome, and acute leukemia. In these cases, allogeneic haematopoietic stem cell transplant (HSCT) is the only valuable therapeutic option. In this study, we analyzed the outcome of SDS patients given HSCT and registered in the EBMT registry.

Methods: Data on HSCT in SDS patients were retrieved retrospectively from the EBMT-Promise database. Data collection was completed with a specific case report form sent to the centers.

Results: 74 SDS patients, F/M 39/35, were transplanted in 43 EBMT centers over the period 1988-2016 for SAA (60) or MDS/AML (14), after a median time from diagnosis of 3.4 yrs (0.2-30.6). Median age at HSCT was 8.7 yrs (0.8-42.9). Donor was a sibling or a family donor in 32% of cases (sibling 18, matched family donor 4, haploidentical parent 2), or an unrelated donor in the remaining 68% (21 matched, 14 mismatched, 15 matching not known). Donor source was BM (70% of cases), PB (19%) and CB (11%). The D/R gender pair was 22% F/M, 78% other combination; the CMV R/D serostatus was: 39% neg/neg, 20% neg/pos, 20% pos/neg, and 20% pos/pos. The conditioning regimen was myeloablative and of reduced intensity in 73% and 27% of patients, respectively. TBI and TLI were used in 8 and 2 patients, respectively. GVHD prophylaxis was based on ex-vivo T-depletion in 9 patients and on in-vivo T-depletion in 49 patients (36 ATG, 13 alemtuzumab). Moreover, 64 patients (86%) received a pharmacological prophylaxis for GVHD based mainly on CSA and short methotrexate or on CSA plus other drugs. The 60-day C.I. of PMN engraftment was 85% after a median time of 17.5 days (10-102), the rate of graft failure being 15%. Grade I-II and II-IV acute GVHD occurred in 28% and 24% of patients, respectively, while chronic GVHD was diagnosed in 21%. After a median f-up of 7.3 years, the 5-year OS was 63% (CI, 5 1-73). The 5-year OS was 71% for SAA and 29% for MSD/AML, p 0.0004. The mortality rate was 38% (10% relapse of MDS/AML or graft failure, 28% toxicity or other causes not specified).

Conclusions: This is the largest series reported so far of SDS patients given HSCT . HSCT showed to be a valuable therapeutic option in SDS patients, especially when transplanted for SAA. The safety and tolerability of HSCT is improved compared to previous historical series although the type of donor selection, intensity of conditioning regimen, and type of GVHD prophylaxis are still a matter of investigation.

Conflict of interest: Nothing to disclose for all authors

P074

Results of haploidentical hematopoietic stem cell transplantation in patients with aplastic anemia: A single center experience

Zafer Gulbas, Demet Çekdemir, İmran Dora, Eda Er, Çiğdem Eren

Anadolu Medical Center Hospital, Bone Marrow Transplantation, Kocaeli, Turkey

Background: New transplant approaches are necessary for patients with refractory severe aplastic anemia (SAA) who lack a matched sibling or unrelated donor (UD) or who have failed UD. We report our results with haploidentical hematopoietic stem cell transplantation (haploHSCT) using with posttransplant high-dose cyclophosphamide.

Methods: We retrospectively summarized 11patients who failed to respond to previous therapies with antithymocyte immunoglobulin and CsA and underwent haplo-HSCT from 2010 to 2017. The conditioning regimen before HSCT included cyclophosphamide, fludarabine, rabbit ATG. The recipients for graft versus host disease (GVHD) prophylaxis received posttransplant high-dose cyclophosphamide. Subsequent to successful allo-HSCT, the hematopoietic reconstitution was observed, a median interval from diagnosis to transplantation of 1.5 years (male, 6 patients; female, 5 patients; median age range, 18-61.

Results: We found that hematopoietic reconstruction was achieved in 9 patients . Graft failure occured in 2 patients and those patients underwent second transplantation from other family haplo donor and both engrafted. Two patients died in 100 days due to sepsis. We did not observed severe (grade 3 -4)acute or chronic graft versus host disease any of the patients.

Conclusions: Our findings show that haploHSCT can successfully rescue refractory SAA patients who lack donor. Delayed for the haploidentic transplantation increase the sepsis related death rate.

Conflict of interest: No conflicts of interest

P075

Screening for telomere disease is crucial to improve the outcome of HSCT for pediatric patients suffering from severe aplastic anemia

Iris Nederlof ¹ , Caroline Lindemans ¹ , Dorine Bresters ² , Marije Bartels ¹

¹ University Medical Center Utrecht, Pediatric Hematology and Stem Cell Transplantation, Utrecht, Netherlands; ² Leiden University Medical Center, Pediatric Hematology/Oncology, Leiden, Netherlands

Background: Telomeropathies compromise a complex spectrum of diseases caused by unprotected and shortened telomeres, resulting in bone marrow failure, and other organ disease, including lung fibrosis and liver cirrhosis. Patients with telomeropathies, of which dyskeratosis congenita (DC) is the best-known example, are at increased risk of hematological stem cell transplantation (HSCT)- associated complications, predominantly significant organ damage following myeloablative conditioning regimen. In patients suffering from severe aplastic anemia (SAA), the preexistence of an underlying congenital bone marrow failure (BMF) disorders, including Fanconi anemia (FA) and DC, should be considered. However, telomeropathies encompass a wide spectrum and clinical presentation can therefore be subtle. Determining telomere length in patients with SAA can be used to predict the likelihood of an underlying telomeropathy (characterized by telomeres < p1) and to adjust treatment regimen if needed. The objective of this study is to investigate clinical characteristics of patients suffering from SAA or severe BMF with respect to a proven or suspected telomeropathy and outcome of allogeneic HSCT.

Methods: We investigated 46 patients (0-18 years) undergoing allogeneic HSCT at two Dutch pediatric HSCT departments between 2009 and September 2017. Patients suffering from SAA (n=29), refractory cytopenia of childhood (RCC, n=11), or BMF and a proven telomeropathy (n=6) were included. Two patients suffering from telomeropathy-associated BMF that were not yet scheduled for SCT were included in the analysis on characteristics pre-SCT. A telomeropathy was diagnosed based on telomere length < p1 in lymphocytes and granulocytes determined by Flow FISH (www.repeatdiagnostics.com). A retrospective analysis on general clinical characteristics, radiological and laboratory parameters associated with BMF, liver dysfunction or pulmonary disorder, complications and toxicity was performed for all patients. Mann-Whitney and Fisher's exact tests were used to analyze the differences between groups.

Results: In the telomeropathy group, congenital malformations were reported more frequently (4/8 versus 4/43, p=0.0147). We also observed elevated levels of ASAT and ALAT (both >1x ULN), and γGT (>2x ULN) in the telomeropathy group (p< 0.0001, p=0.0245). Abnormal radiological appearances of the liver were more frequently reported pre-HSCT in the telomeropathy group (6/8) compared to the control group (5/43, p=0.0006). Our data suggest a trend for macrocytic anemia in the telomeropathy group (4/8) compared to the non-tested group (4/43) (p=0.076). There was no difference in reported radiological abnormalities of the lungs before HSCT. Four patients with a telomeropathy (4/6, 66%) died following HSCT complications, compared to 11 (26%) in the non-tested group. Interestingly, based on our findings in the telomeropathy group, we have identified 12 patients in the non-tested group suggestive for an underlying telomeropathy, of which 5/12 (42%) died from severe HSCT-associated complications.

Conclusions: SAA/severe BMF can be the result of an underlying telomeropathy other than classical DC. In patients that will undergo allogeneic HSCT, determining telomere length can help to identify patients that have an increased risk for HSCT-associated complications. This will contribute to a better outcome of HSCT for all patients with SAA/ severe BMF and to a better understanding of telomere disease in general.

Conflict of interest: None of the authors has anything to disclose.

P076

Very long-term follow up of aplastic anemia treated with allogeneic stem cell transplantation and immunosuppressive treatment

Felicitas Zurbriggen, Jakob R. Passweg, Beatrice Drexler

University and University Hospital of Basel, Hematology, Basel, Switzerland

Background: Aplastic anemia (AA) historically was a fatal condition. Since the 1970s patients benefited from treatments consisting mainly of immunosuppression (IST, originally horse ATG) and hematopoietic stem cell transplantation (HSCT). Although the response rate and survival after treatment improved, problems remain, including incomplete response, relapse, disease- and treatment-related complications, late effects and clonal evolution. In order to assess very long-term outcome in aplastic anemia we analysed our patients in a single - center study over 44 years.

Methods: We analysed long term outcome of 302 patients treated between 1973 and 2017 at the University Hospital Basel.

Results: Median age at diagnosis was 29.9 years (range 0.71-80). Eight patients were diagnosed with moderate AA, 208 with severe AA and 23 with very severe AA. Horse ATG was first-line treatment in 224 patients, 74 patients received allogeneic HSCT and 4 had other first line therapy. With a median follow up of 17 years, overall survival at 30 years after initial treatment was 49 + 12% in transplanted patients and 41 + 9% in patients with initial ATG treatment. There were more early deaths in the HSCT group and more late deaths in the ATG group. Death rate was particular high in the first 10 years posttransplant, whereas transplant patients afterwards mostly survived with stable hematopoiesis (see figure). Patients with immunosuppressive treatment showed less early deaths but had a lower overall survival after 30 years. At last follow up most of long-term survivors are in complete or partial remission. More details will be presented at the conference.

Conclusions: Most of aplastic anemia survivors show long-term hematopoesis, whereas death occurs mostly early in the HSCT group and more late in the ATG group.

Conflict of interest: None of the authors has anything to disclose.

[P076 Figure]

$$$$

Autoimmune diseases

P077

A retrospective analysis of false positive antibodies in patients receiving intravenous immunoglobulin for immune mediated neuropathies

Allison Clendenan ¹ , Jacquelyn Henry ² , Carri Alldredge ¹ , Richard Burt ²

¹ Northwestern University, Division of Immunotherapy, Chicago, IL, United States; ² Northwestern University, Division of Immunotherapy for Autoimmune Diseases, Chicago, IL, United States

Background: Intravenous immunoglobulin (IVIg) is pooled preparation of IgG from thousands of donors and can possibly cause false positive antibodies in patients receiving IVIg potentially leading to improper diagnoses.

Methods: Records of 115 patients with immune mediated neuropathies who underwent evaluation at Northwestern Medicine in Chicago, IL were reviewed. Data collected included whether or not the patient was receiving IVIg at the time of evaluation, antibodies for hepatitis B core (anti-HBc), human T-lymphotropic virus (HTLV), human immunodeficiency virus (HIV), syphilis, and anti-glutamic acid decarboxylase (anti-GAD).

Results: The graph below depicts the percentage of patients with false positive antibodies divided into patients who were receiving IVIg versus those who were not receiving IVIg at the time of evaluation. Twenty-seven patients were not receiving IVIg. In the group not receiving IVIg there were two false positives (7%, 2/27) identified (one anti-HBc, one syphilis). Eighty-eight patients were receiving IVIg. One hundred and forty-seven false positives were identified in this group. Of these false positives 73% (64/88) were anti-HBc, 1% (1/88) was HIV, 40% (35/88) were HTLV, 82% (23/28) were syphilis, and 92% (23/26) were anti-GAD.

Conclusions: IVIg increases the incidence of false positive antibodies for infections, specifically, anti-HBc, HTLV, and syphilis. Confirmatory testing should be completed if the antibodies are positive and a patient is receiving IVIg. Anti-GAD can also be falsely elevated due to IVIg. Diagnoses supported by high anti-GAD in the blood, such as type 1 diabetes and stiff person syndrome, should be used with caution in patients receiving IVIg due to the likelihood of it being falsely positive.

Conflict of interest: None of the authors has anything to disclose.

[P077 Figure]

$$$$

P078

Analysis of Immune Reconstitution by Multicolour Flow Cytometry in Multiple Sclerosis and Non-Autoimmune Patients Following Autologous Haematopoietic Stem Cell Transplantation

Carole D Ford ¹ , Melissa Khoo ¹ , John Zaunders ² , Jennifer Massey ³ , Kevin Hendrawan ¹ , Ian Sutton ³ , David D F Ma ^1,3 , John J Moore ³

¹ St Vincent's Centre for Applied Medical Research, Sydney, Australia; ² The Kirby Institute, Sydney, Australia; ³ St Vincent's Hospital, Darlinghurst, Australia

Background: Autologous haematopoietic stem cell transplantation (ASCT) is a promising strategy for Multiple Sclerosis (MS) patients that do not respond to conventional treatments but the mechanisms able to achieve clinical improvement of MS in transplant recipients are not fully understood.

Methods: We hypothesised that ASCT eliminates the pro-inflammatory disease-causing T cells to halt disease progress, and promotes tolerance via multiple mechanisms, including reconstitution of immune-regulatory suppressive cell subsets, to achieve sustained disease control. We designed polychromatic flow cytometry panels with 9-14 colours to characterise over 30 subpopulations of peripheral blood cells (PBMC) of MS patients, undergoing ASCT in a Phase II study at our institute. Various differentiation markers, chemokine receptors, and trafficking markers were used to identify populations unique to autoimmune disease and compared these findings with those obtained from non-autoimmune Lymphoma patients treated with ASCT. Analysis included t-test, Wilcoxon signed rank test and tSNE.

Results: Immune reconstitution was assessed by multicolour flow cytometry in eighteen MS patients at pre-ASCT and 3, 6, and 12 months post-ASCT time points. Pro-inflammatory disease-causing T cells, such as CCR6+CD161hi mucosal-associated invariant T (MAIT) cells (a Th17 subset with brain-homing abilities) and CD49d+CCR6+ CNS-homing cells, significantly decreased following treatment in both MS (p < 0.05) and NHL patients (p < 0.05). Interestingly in MS patients, regulatory T cells increased significantly at 3 months (p < 0.05) post ASCT while in Lymphoma patients, regulatory T cells decreased at 6 months post ASCT (p < 0.05). Expression of cytotoxic CD8+CD57+ T cells tripled at 3 months post ASCT (p < 0.0001) in MS patients while no significant change occurred in Lymphoma patients. Natural killer (CD56hi) expression doubled (p < 0.05) by 3 months post ASCT in MS patients but in Lymphoma patients there was no significant change post transplantation.

Conclusions: ASCT depleted pro-inflammatory T cells in both MS and Lymphoma patients. However, we found the pattern of expression in other immune-regulatory cell subsets differed in these patient cohorts post transplantation. These distinct patterns of immune reconstitution between MS and Lymphoma post-ASCT provide tantalising indication that these changes may relate to the different underlying pathogenic process and aid to pinpoint changes specific to immune deregulation in MS. These findings have the potential to allow us to track disease progression and treatment outcomes by monitoring the presence of specific cell types, which may help to improve the management of patients living with MS.

Conflict of interest: Nothing to disclose

P079

Autologous Haematopoietic Stem Cell Transplantation (AHSCT) for Anti-Neutrophil Cytoplasmic Antibody (ANCA)-associated Vasculitis (AAV): Results of a EBMT Retrospective Survey

Clare Samuelson ¹ , Tobias Alexander ² , Thomas Daikeler ³ , Manuela Badoglio ⁴ , Jörg Henes ⁵ , Mohamed Akil ¹ , Lars Skagerlind ⁶ , Gerhard Ehninger ⁷ , Esa Jantunen ⁸ , Dominique Farge ⁹ , John A Snowden ¹

¹ Sheffield Teaching Hospitals NHS Foundation Trust, Sheffield, United Kingdom; ² Charite - University Medicine Berlin, Department of Rheumatology and Clinical Immunology, Berlin, Germany; ³ University Hospital Basel, Department of Rheumatology, Basel, Switzerland; ⁴ Saint Antoine Hospital, EBMT Study Office, Paris, France; ⁵ Medizinische Klinik II, Universitätsklinikum Tübingen, Tübingen, Germany; ⁶ University Hospital Umea, Department of Hematology, Umea, Sweden; ⁷ Medizinische Klinik und Poliklinik I, Universitätsklinikum der Technischen Universität Dresden, Dresden, Germany; ⁸ University Hospital Kuopio, Department of Internal Medicine, Kuopio, Finland; ⁹ St. Louis Hospital, Internal Medicine, Paris, France

Background: ANCA-associated vasculitides (AAV) are chronic autoimmune diseases, which can present with life-threatening multi-system involvement. Current treatment options are still suboptimal with risks of disease progression and treatment-related toxicity. Autologous haematopoietic stem cell transplantation (AHSCT) has been used in a number of such cases but patient outcomes following such management are unclear.

Methods: Adults who received AHSCT for AAV and whose data were registered with the EBMT were identified retrospectively through the EBMT database. Treating physicians were surveyed to produce a retrospective evaluation of outcomes.

Results: 7 patients underwent AHSCT primarily for AAV between 1999-2014 in 6 centres across Europe. 5 females and 2 males were transplanted; 6 had a diagnosis of granulomatosis with polyangiitis and 1 eosoniphilic granulomatosis with polyangiitis. Median age was 39 years (range 32-55 years). Patients had received 4-6 prior lines of therapy, including cyclophosphamide (CYC, with median cumulative dose of 80g) and steroids in every case, and rituximab in 4 cases. Stem cell source was peripheral blood in every case; CD34-selection was performed in 4 cases, mean CD34⁺ cell dose was 4.2 x 10⁶/kg (range 0.6-7.9x10⁶/kg). Conditioning regimen was CYC/ATG in 5 patients and CYC in 2 patients. Median follow-up was 86 months (range 1-204 months). Transplant-related mortality (TRM) occurred in 2 causes. All but one patient went into remission but 3 later relapsed at 6, 12 and 36 months, respectively, and required further treatment for disease control. At time of last clinical follow-up, 3 patients had drug-dependent partial response; 1 had drug-dependent complete remission and 1 had drug-free complete remission.

Conclusions: Outcomes of AHSCT for these heavily pre-treated AAV patients were variable. Only 1 patient achieved drug-free complete remission and TRM was observed in a quarter. Nevertheless, AHSCT had the potential to stabilize AAV in patients who initially failed to respond to conventional therapies. These data do not support AHSCT for advanced stage ANCA-positive vasculitis, although it may have a place as salvage treatment in otherwise refractory patients. As for other autoimmune diseases, AHSCT may provide better outcomes when performed at early stage of disease. Overall, AHSCT should only performed in clinical trial settings in experienced centers.

Conflict of interest:

J. Snowden declares honoraria for speaking from Sanofi and Jazz.

T. Alexander declares honoria and/or traval support from Abbvie, Amgen, Neovii and Pfizer.

The other authors declare no conflicts of interest.

P080

Autologous Haematopoietic Stem Cell Transplantation (Auto-HSCT) in Stiff Person Syndrome (SPS): The Sheffield Experience

Alice Elizabeth Thorpe, Lewis Kass-Iliyya, Helen Jessop, Andrew Chantry, Marios Hadjivassiliou, Basil Sharrack, John A Snowden

Sheffield Teaching Hospitals NHS Foundation Trust, Sheffield, United Kingdom

Background: Stiff Person Syndrome (SPS) is a rare (1/1,000,000) immune-mediated neurological disorder characterised by muscle rigidity, spasms and high glutamic acid decarboxylase (GAD) antibodies. Auto-HSCT has successfully been used to treat limited numbers of SPS patients (Sanders S et al, JAMA Neurology, 2014). We describe our single centre experience of assessing referrals of SPS for potential auto-HSCT.

Methods: Between 2015-2017 we reviewed six patients with SPS from across the UK as summarised in table 1. One patient (Patient 6) was declined auto-HSCT as they were found to be stable on Mycophenolate and five were deemed suitable for autograft (Patients 1-5). Two patients have been treated. Patient 1 had the Progressive Encephalomyelitis with Rigidity and Myoclonus (PERM) variant of SPS with ataxia. Patient 2 had Classical SPS. Both patients were significantly disabled and had failed multiple immunotherapies including intravenous immunoglobulin (IVIG) and plasmapheresis. The patients were mobilised with Cyclophosphamide (Cy) 2g/m² + G-CSF and conditioned with Cy 200mg/kg + ATG followed by auto-HSCT, as per EBMT Guidelines (Snowden JA et al, Bone Marrow Transplantation, 2012). A third patient will be scheduled for auto-HSCT pending funding authorisation. Two patients have been declined funding.

Results: Despite significantly reduced performance status, both patients tolerated the procedure with routine toxicities. Patient 1 engrafted on day +10 and was hospitalised until day +16. He developed EBV reactivation but did not require treatment. Subsequent neurological assessment showed disease stabilisation with normalisation of anti-GAD titres and fortnightly IVIG was discontinued. At 12 months post-transplant, MR spectroscopy of the cerebellum showed significant increase in the NAA/Cr ratio reflecting improved metabolic activity. Patient 2 engrafted on day +10 and was discharged on day +16. She discontinued fortnightly IVIG and assessment at 6-months showed resolution of neurological symptoms and normalisation of clinical examination. Her anti-GAD titre remained high.

Conclusions: Auto-HSCT may be a viable treatment option in patients with severe SPS, supporting previous work (Sanders S et al, JAMA Neurology, 2014). It may also prove to be a more cost effective treatment in patients requiring regular treatment with expensive modalities, such as IVIG. Further work is warranted to establish long-term safety, efficacy and cost-effectiveness of auto-HSCT in SPS, along with optimising patient selection and transplant technique.

Conflict of interest: J. A. Snowden declares honoraria for speaking from Sanofi and Jazz.

Patient	Age	Gender	SPS Phenotype	Disease duration prior to auto-HSCT assessment (years)	Baseline anti-GAD titre (U/ml)	Baseline Neurophysiology	Prior Immunomodulatory Therapy	Outcome
1	52	Male	PERM & ataxia	5	372	Abnormal blink reflex.	Plasmapheresis, IVIG	Auto-HSCT (NHS funded)
2	36	Female	Classical SPS	8	>2000	Brainstem hyper-excitability, exaggerated startle reflex.	Plasmapheresis, IVIG	Auto-HSCT (Self-funded)
3	37	Female	Classical SPS	9	>2000	Continuous motor unit activity, brainstem hyper-excitability.	IVIG, Rituximab, Azathioprine	Awaiting NHS funding
4	53	Female	Classical SPS	16	>2000	Not available.	IVIG	NHS funding declined
5	52	Male	PERM	8	<5	Exaggerated startle reflex with variable delays.	IVIG, Plasmapheresis, Mycophenolate, Azathioprine, Steroids	NHS funding declined
6	42	Male	Classical SPS	7	>2000	Failure of suppression of 2nd R2 response in paired blink reflex. Continuous motor unit activity.	IVIG, Plasmapheresis, Mycophenolate	Did not progress to auto-HSCT as stable on Mycophenolate

[ [P080 Table] Table 1]

P081

High Dose Chemotherapy and Autologous Stem Cell Rescue For Multiple Sclerosis: Preliminary Data From India

Gaurav Kharya ¹ , Silky Jain ¹ , Divya Bansal ¹ , Anil Khetrapal ² , Vikrant Bhar ³ , Manish Mahajan ⁴ , Sumit Singh ⁴

¹ Artemis Hospital, Pediatric Hematology Oncology Immunology and Bone Marrow Transplant, Gurugram, India; ² Artemis Hospital, Department of Transfusion Medicine, Gurugram, India; ³ Artemis Hospital, Department of Lab Hematology, Gurugram, India; ⁴ Artemis Hospital, Agrim Institute of Neurosciences, Gurugram, India

Background: Multiple sclerosis is a debilitating autoimmune disease leading to progressive neurological deterioration. It is categorized into relapsing remitting (RRMS), secondary progressive (SPMS) or primary progressive (PPMS). Extended disability status scale (EDSS) decides the extent of disability ranging from 0-10. Disease modifying agents (DMA) have been used to treat symptoms with variable results but they are limited by cost and adverse effects. There is new found interest in high dose chemotherapy and autologous stem cell rescue (HDC+ASCR) for treatment of MS. We report our initial experience from India.

Methods: Retrospective data from 37 patients suffering from MS who underwent HDC+ASCR between January 2017 till date was analyzed. Initial 10 patients received cyclophosphamide + GCSF based mobilization but subsequent patients underwent GCSF primed peripheral blood stem cells collected on day 5 with a target pre CD34 count of >20%. One patient required plerixifor as GCSF mobilization was suboptimal. Stem cell product was cryopreserved with DMSO while patient received conditioning chemotherapy. All 37 were conditioned using cyclophosphamide@50mg/kg/day x 3 or 4 days (day -5 to -2) and rATG (Thymoglobulin)@2.5mg/kg/day x 3 days (day -4 to -2). Median CD34 count infused was 7.3x10⁶/kg (range 3.82 to 13.84).

Results: Median age at transplant was 47 years (range 33 to 67). 23 were < 50 yrs whereas 14> 50 yrs of age. 22 patients had EDSS< 6 whereas 15 had an EDSS of >6. 13 were RRMS, 18 SPMS and 5 PPMS. 4 were drug naïve whereas 33 received at least 2 DMA before entering into transplant. At a median follow up of 200 days (range 16 to 357) overall survival is 92%. 3 patients died due to transplant related complications (2 sepsis, 1 drug toxicity). Two patients developed DMSO induced kidney damage whereas one developed neurotoxicity. At the last follow up none of the patients were on any DMA with no progression of symptoms. A small subset of patients reported symptom reversal which needs to be substantiated by a neurologist.

Conclusions: High dose chemotherapy with autologous stem cell rescue (HDC+ASCR) is a useful modality to treat multiple sclerosis both in drug naïve and drug refractory cases. Larger studies and longer follow up is required to further consolidate this claim.

Clinical Trial Registry: None.

Conflict of interest: None.

P082

Impact of Autologous Haematopoietic Stem Cell Transplantation (AHSCT) in Multiple Sclerosis on Treatment Costs - Real World Analysis in a Nationwide Study

Katarzyna Orlewska ¹ , Krzysztof Bogusz ¹ , Monika Nojszewska ² , Mirosław Markiewicz ³ , Robert Liwoch ³ , Grzegorz Helbig ³ , Andrzej Śliwczyński ⁴ , Emilian Snarski ¹

¹ Medical University of Warsaw, Hematology, Oncology and Internal Diseases, Warszawa, Poland; ² Medical University of Warsaw, Department of Neurology, Warszawa, Poland; ³ Medical University of Silesia, Department of Hematology and Bone Marrow Transplantation, School of Medicine in Katowice, Katowice, Poland; ⁴ Central Office of the National Health Fund, Warszawa, Poland

Background: Autologous haematopoietic stem cell transplantation (AHSCT) is an extremely effective therapeutic option for patients with severe multiple sclerosis (MS). The AHSCT is expensive and there is limited data on impact on real cost of treatment of MS patients outside of clinical trials. We collaborated with the National Health Fund (NHF) which collects all data on costs associated with treatment to assess the nationwide impact of AHSCT on costs of MS treatment among over 40 000 MS patients.

Methods: This is a retrospective observational study based on healthcare administrative data from the National Health Fund (NHF), covering an entire Polish population of about 38 million inhabitants and all treatment costs within the system. Statistical analysis was performed using paired t-test in MedCalc. The costs were calculated from public payer's perspective in Polish złoty and are presented in Euro (1 EUR = 4.2 PLN, 2017).

Results: According to data from the NHF, the prevalence of MS in Poland was 103.06 and 113.06 per a population of 100,000 in 2008 and 2016, respectively. During the years 2008-2016, 102 patients suffering from MS underwent AHSCT in Poland and the treatment was covered by NHF. We have analysed all 102 patients comprehensively, taking into consideration all available data before and after AHSCT. According to the NHF data, the average treatment-related costs were about 4 520 EUR per year before the transplantation, and were reduced to an average of 810 EUR per year after the transplantation (p< 0.0001). The cost of AHSCT from NHF perspective is about 12 380 EUR. During the studied period only 3 out of 102 patients required treatment with disease modifying drugs after AHSCT - two patients were treated with dimethyl fumarate approximately 2 years after the procedure (one dose and two doses within a month, respectively), one patient was treated with interferon beta-1b within a month after the procedure. Moreover, there were no AHSCT related deaths among patients in the studied group.

Conclusions: There is limited data on the impact of AHSCT on costs of treatment of MS. In a real world analysis of nationwide medical data, we present that AHSCT significantly reduces the use of disease modifying drugs among MS patients after AHSCT and leads to rapid reduction of treatment costs of these patients for the public payer. If the amount of treatment cost remains stable for the 3 years after AHSCT the expense of transplantation may be compensated within that time. The NHF data does not allow the assessment of clinical progression of the disease.

Clinical Trial Registry: Not applicable

Conflict of interest: Nothing to disclose

P083

Impact of homeostatic proliferation, telomere attrition, PD-1 expression and CMV-reactivity on TCR diversity and relapse after AHSCT in systemic sclerosis patients

Lucas C M Arruda ^1,2 , Emmanuel Clave ^3,4 , João R. Lima-Júnior ² , Kelen C. R. Malmegrim ^2,5 , Corinne Douay ⁴ , Antônio José Alberdi ^3,6 , Maria Carolina Oliveira ^1,2,7 , Antoine Toubert ^3,4

¹ Ribeirão Preto Medical School, University of São Paulo, Department of Biochemistry and Immunology, Ribeirão Preto, Brazil; ² Center for Cell-based Therapy, Regional Hemotherapy Center of the Ribeirão Preto Medical School, University of São Paulo, Ribeirão Preto, Brazil; ³ Université Paris Diderot, Sorbonne Paris Cité, Paris, France; ⁴ INSERM UMR-1160, Institut Universitaire d'Hématologie, Hôpital Saint-Louis - APHP, Paris, France; ⁵ School of Pharmaceutical Sciences of Ribeirão Preto, University of São Paulo, Department of Clinical, Toxicological and Bromatological Analysis, Ribeirão Preto, Brazil; ⁶ Plateforme Technologique, Institut Universitaire d'Hématologie, Paris, France; ⁷ Division of Clinical Immunology, Ribeirão Preto Medical School, University of São Paulo, Department of Internal Medicine, Ribeirão Preto, Brazil

Background: Autologous hematopoietic stem cell transplantation (AHSCT) has been used to treat severe and progressive systemic sclerosis (SSc) patients, however, exactly how AHSCT reshapes the immune system still need to be completely clarified. We assessed the role of replicative senescent and exhausted cells reconstitution and TCR characteristics on clinical outcome post-AHSCT in SSc patients.

Methods: Twenty-five SSc patients were clinically monitored for skin thickness (modified Rodnan's skin score, mRSS) and lung function (Forced Vital Capacity, FVC) before and semiannually until 3-years post-AHSCT. Patients were retrospectively classified as responders (n=19) and non-responders (n=6) according to clinical outcomes. Following EBMT guidelines, blood samples were collected for quantification of telomere length by RT-qPCR, evaluation of TCR characteristics by New Generation Sequencing, reconstitution of senescent CD8⁺CD28^-CD57⁺ and exhausted PD1⁺ cells by FACS, and serum cytokines level measurement by CBA-Flex. VDJTools was used to assess the TCR diversity (Shannon index), the relative TCR usage frequency (%) of each Vβ and Jβ family, and the changes in Vβ-Jβ segment pairing in CDR3 junctions. VDJviz and VDJdb databases were screened to annotate the frequency of CMV-specific TCR sequence motifs.

Results: From 6 months post-AHSCT until the end of follow-up, mRSS decreased (P≤0.001) and the pulmonary function stabilized when compared to pre-AHSCT. In parallel, inflammatory cytokine levels (IL-6, IL-1β) decreased transiently at 6 months (P< 0.05). Relative T/S telomere length ratio decreased by 41% at 6 months when compared to baseline (P=0.015), correlating with the increase of CD8⁺CD28^- cells expressing CD57 and FoxP3 (r_s=-0.76, P=0.036). Consequently, there was an increase of CD8⁺CD28^-CD57⁺ (P=0.004) and CD8⁺CD28^-FoxP3⁺ (P=0.001) percentages, with the CD57 expression by CD8⁺CD28^- cells correlating negatively with the telomere length (r_s=-0.34, P=0.015). PD-1 expression on both CD4⁺ and CD8⁺ T-cells increased at 6 months (P≤0.01). After AHSCT, responder patients presented lower TGF-β, IL-6, G-CSF (P≤0.01) and IL-1, IL-17A, MIP-1 and IL-12 (P< 0.05) levels than non-responders, as well as higher PD-1 expression on T- (P< 0.05) and B- (P≤0.01) cells. A strong shift in Vβ-segments usage was observed for some families exclusively in responder patients. The TRBV13 segment usage increased post-AHSCT when compared to baseline (P=0.031), while there was a significant decrease (P< 0.05) of the families TRBV18, TRBV24-1, TRBV29-1, TRBV4-1, TRBV4-3 and TRBV6-1 percentage usage. No such changes are present for Jβ-segments. For non-responding patients, no significant shift on Vβ- neither Jβ-segments was observed following AHSCT. V-J pairing changes in responder patients resulted in increase of TCR diversity post-AHSCT as compared to baseline (P=0.030), while decreased in non-responder patients. Among responders, only one patient presented CMV-specific sequences, that remained stable post-AHSCT (0.18% to 0.19%). The non-responder patients experienced an increase of CMV-specific sequences from 0.11±0.19% to 0.32±0.41% (P=0.031), in parallel, all these patients presented a decrease of TCR diversity.

Conclusions: Homeostatic proliferation post-AHSCT results in transient telomere attrition and increased numbers of senescent and exhausted cells, leading to high PD-1 expression in patients with better clinical outcomes after AHSCT. Additionally, our results indicate a possible relationship between CMV infection and reduced TCR diversity and disease relapse after AHSCT in SSc patients.

Conflict of interest: All the authors have nothing to disclose

P084

Peripheral blood stem cell mobilisation with Cyclophosphamide and G-CSF in patients with Multiple Sclerosis: A safe and predictable procedure

Sara Lozano, Philippa Woolley, Ian Gabriel, Renuka Palanicawandar, Maria Teresa Cencioni, Paolo Muraro, Richard Nicholas, Malik Omar, Sinju Thomas, Tsembayena Dlamini, Rosalinda Elio, Eduardo Olavarria

Imperial College London, Haematology, London, United Kingdom

Background: Autologous hematopoietic stem cell transplantation (HSCT) has been utilised for the treatment of severe multiple sclerosis (MS). It results in significant improvement of neurological function, although patients could experience exacerbations of MS-related symptoms during the procedure. However, there is a lack of precise information regarding complications of the stem cell mobilisation and collection procedure in this patient population.

Methods: We analysed 45 patients diagnosed with MS that underwent a stem cell mobilization and collection from March 2016 to November 2017. The median age was 38 years (17-625). 21 patients (47%) were male. The interval from diagnosis to HSCT was 114.3 months (range 11.6-128.3). 24 patients (53%) had relapsing-remitting (RRMS), 16 patients (35%) secondary-progressive (SPMS) and 6 patients (12%) primary-progressive (PPMS) Multiple Sclerosis. The median number of previous therapies was 2 (range 0-4). Only 2 patients (12%) had not received any prior treatment. The median Expanded Disability Status Scale (EDSS) score was 6 (range 2-7). Peripheral blood stem cells were mobilized with cyclophosphamide (CY) 2g/m2 in day +1 and daily GCSF (5 µg/kg subcutaneously) starting from day +3 from CY until the completion of the harvest.

Results: Of the total cohort, 35 patients (79%) underwent mobilization with CY+GCSF uneventfully. Only 2 patients (4%) had an exacerbation of MS (one with fatigue and increase of spasticity, other with worsening weakness) during the mobilisation period. Five patients required hospital admission before or during the mobilization procedure: four patients developed neutropenic fever (including one septic episode) and one patient developed intractable G-CSF associated back pain. Two additional patients were hospitalised because of non-neutropenic fevers and flare-up of MS symptoms after the stem cell collection.

The median CD34+ cell dose was 7.75 ×10⁶/kg (range 2.2-24.3). The median number of apheresis was 1 (range 1-2), with only 2 patients (4%) needing a second day of collection. All but one patient mobilised very predictably sufficient stem cells on day +10 after Cyclophosphamide allowing for a scheduling of Cyclophosphamide on a Friday and stem cell collection on a Monday. The peripheral blood CD34+ cells concentration on day +10 correlated with the final CD34+ cells yield of the collection (figure 1).

Since February 2017, all Cyclophosphamide infusions were given as an out-patients procedure. There were no differences in post mobilisation complications or stem cell collection efficiency between 20 (44%) in-patient and 25 (56%) out-patient mobilisations.

Conclusions: In our hands, after CY and GCSF mobilization there is a very low rate of exacerbations of MS symptoms and an acceptable rate of post mobilisation complications. Exacerbations of MS symptoms could occur after the stem cell collection procedure has been completed, raising the need for a prolonged follow up.

Conflict of interest: Nothing to disclose

[P084 Figure]

$$$$

P085

Primary Immunodeficiency Disorders behind haematological cytopenia: early recognition for a prompt treatment strategy

Sara Ciullini Mannurita ¹ , Maddalena Bagni ¹ , Marina Vignoli ¹ , Fabio Tucci ² , Veronica Tintori ³ , Stefano Frenos ³ , Maria Chiara Sanvito ³ , Marinella Veltroni ² , Claudio Favre ² , Eleonora Gambineri ^1,3

¹ University of Florence, Department of 'NEUROFARBA': Section of Child's Health, Firenze, Italy; ² 'Anna Meyer' Children's Hospital, Department of Haematology-Oncology, Florence, Italy; ³ 'Anna Meyer' Children's Hospital, Department of Haematology-Oncology: BMT Unit, Florence, Italy

Background: Cytopenia, defined as the reduction of one or more mature blood cell types (eg. neutropenia, anemia, or thrombocytopenia) in the peripheral blood, may be a typical first symptom of primary immunodeficiency disorders (PIDDs). Possible causes of cytopenia in PIDDs comprise mainly immune dysregulation, bone marrow failure (BMF) and myelodysplasia. Our goal is to investigate possible immune mediated mechanisms underlying chronic cytopenia in children in order to to achieve an early diagnosis and consequently offer timely and appropriate therapy.

Methods: Among the patients referred to our Paediatric Haematology/Oncology Unit, we selected thirty-five patients affected by chronic cytopenia. These patients were characterized throughout clinical, immunological, haematological and genetic investigations. Nineteen patients were further evaluated with immunophenotyping by flow cytometry and Next Generation Sequencing (NGS) analysis of a gene panel including genes frequently implicated in immunodeficiency or immunodysregulation.

Results: Among thirty-five patients, seven patients showed a distinctive early onset clinical and haematological phenotypes and were promptly diagnosed with Fanconi Anemia, Shwachman-Diamond syndrome and severe congenital neutropenia; four patients were defined as affected by uncharacterized BMF and twenty-four affected by immune-mediated cytopenia (fourteen cases of isolated autoimmune neutropenia, one case of neutropenia associated with thrombocytopenia, two cases of autoimmune pancytopenia and seven patients with different combinations of cytopenia and multi-organ autoimmunity).

Immune phenotype analysis was performed on nineteen patients with immune mediated cytopenia. A specific pattern was found with abnormalities affecting T CD8 and B subtypes. We observed a decrease of CD8 effector memory (EM) and terminally differentiated (TEMRA) cell populations, while an increase of CD8 naive and central memory (CM) cells levels. B lymphocytic populations showed a decrease in total B cells, B switched cells and B memory cells and an increase in CD21low cells. Absence of plasmacytoid and monocytoid dendritic cells was observed in one case.

Preliminary results from NGS studies in 4 patients revealed mutations in the following genes: STAT3 (gain-of-function mutation), IKAROS, GATA2 and a novel gene not previously associated with autoimmunity and immune dysregulation. Further validation studies are ongoing.

Based on our immunological and genetic studies four patients were successfully treated with haematopoietic stem cell transplantation (HSCT) in haploidentical, matched-related and unrelated setting using reduce intensity conditioning.

Conclusions: Our findings show how the presence of specific lymphocyte subpopulation patterns can be an important indicator of an immune-mediated pathogenesis of cytopenia and a helpful warning sign of PIDDs that should promptly be investigated with genetic analysis. Based on our experience an early diagnosis offers timely and appropriate therapy. HSCT reverses most of the inherited conditions with good long-term outcomes.

Conflict of interest: None of the authors has anything to disclose.

P086

Prognostic factors for clinical response in systemic lupus erythematosus patients treated by allogeneic mesenchymal stem cell

Lihui Wen ¹ , Myriam Labopin ² , Manuela Badoglio ³ , Dandan Wang ¹ , Dominique Farge ⁴ , Lingyun Sun ¹

¹ The Affiliated Drum Tower Hospital of Nanjing University Medical School, Department of Rheumatology and Immunology, Nanjing, China; ² iEBMT Data Office, Sant Antoine Hospital, Department of Haematology, INSERM UMR 938, Université Pierre et Marie Curie, Paris, France; ³ Hôpital Saint-Antoine, EBMT Data Office, INSERM UMR 938, Paris, France; ⁴ Hôpital St-Louis, AP-HP Assistance Publique des Hôpitaux de Paris, CRMR Centre de Référence des Maladies Auto-Immunes Systémiques Rares d'Iles-de-France, Filière 'FAI2R', Paris Denis Diderot University, Unité Clinique de Médecine Interne, Maladies Auto-Immunes et Pathologie Vasculaire, UF 04, Paris, France

Background: Using a single-center cohort of Chinese adult patients with systemic lupus erythematosus (SLE),defined by ACR criteria, we analyzed the predictors of the clinical response observed during the year after bone marrow-derived (BM) and/or umbilical cord-derived (UC) allogeneic mesenchymal stem cell (MSC) treatment.

Methods: Inclusion criteria were : adult active SLE patients (age>18years) with a SLE Disease Activity Index (SLEDAI) score ≥ 8, despite prior treatment (oral cyclophosphamide (CYC) (500-700 mg/m²) for ≥ 6 months or mycophenolate mofetil (MMF) (≥ 1000 mg/d) or leflunomide (LEF) (20 mg/d) for ≥ 3 months, or prednisone ≥ 20 mg/d with or without hydroxychloroquine (HCQ)) who received at least one intravenous injection of allogeneic BM and/or UC-MSC (1 million/kilo bodyweight). SLE clinical and biological symptoms, SLEDAI score were assessed at baseline, 1, 3, 6, and 12 months during follow-up. Clinical remission (CR) was defined as SLEDAI < 3 without major organ activity, and ≤ 5 mg/day prednisone with or without maintenance oral CYC, MMF, LEF. Low disease activity (LDA) was defined by SLEDAI score ≤ 4 without major organ activity, and prednisone ≤ 7.5 mg/day, with or without maintenance oral CYC, MMF, LEF. Prognostic factors for CR and LDA were identified by univariate and multivariate logistic regression. All variables associated with response in univariate analysis (p ≤ 0.10) were included in the model, then a stepwise selection was used.

Results: Among 69 (61 female) patients, median aged 32y (range: 20-54), baseline SLEDAI was 13 (8-34), 39% were anti-dsDNA positive, 43.5% had low serum complement levels and SLE duration before MSC injection was 62 months (range: 2-264). Prior treatment consisted of CYC (68.1%), MMF (27.5%), LEF (20.3 %), prednisone (100%) or HCQ (44.9%), 42 CYC induction (60.9%). All patients received at least one (18 BM/51 UC) MSC injections or two within 1 month, using the same MSC dose and source in 24 patients (9 BM/15 UC). Median (range) of SLEDAI at 1, 3, 6, 12 months after MSC injection were respectively 9 (4-20), 8 (0-16), 6 (0-18) and 5 (0-18) (figure 1). CR was achieved for 16 (23.2%) patients and LDA in 40 (58%). Using multivariate analysis, older age (OR = 0.90, p = 0.025), no arthralgia/arthritis (OR = 0.06, p = 0.002), no prior oral CYC (OR = 0.08, p = 0.005) or HCQ (OR = 0.08, p = 0.008) were associated with higher CR rate. No association was found between CR and anti-dsDNA, cardiorespiratory disorder, prior prednisone, CYC induction or a second MSC injection. By univariate analysis, older age and no arthralgia/arthritis were associated with higher rate of LDA (p=0.006 and p=0.03). All patients with neurological disorder (n=5) at baseline achieved LDA.

Conclusions: In active SLE patients, one intravenous injection of allogeneic MSC allowed 23% CR and 58% LDA within a year. Further studies, using unique immunosuppressive regimen and MSC sources in collaboration with EBMT will allow refined analysis for the determinants of clinical response to allogenic MSC, a promising therapeutic option in multi-systemic SLE.

Conflict of interest: The authors declare no conflict of interest.

[P086 Figure]

$$$$

P087

Reconstitution of natural or vaccination-driven immunity after aHSCT in multiple sclerosis

Chiara Innocenti ¹ , Maria Pia Amato ² , Alessandro Barilaro ² , Gaetano Bianchini ³ , Teresa Capobianco ³ , Maria Antonia Calia ³ , Romina Ceschini ³ , Arianna Fani ¹ , Marta Giannini ² , Antonella Gozzini ¹ , Stefano Guidi ¹ , Alice Mariottini ² , Luca Massacesi ² , Chiara Nozzoli ¹ , Anna Repice ² , Gianmaria Rossolini ^3,4 , Riccardo Saccardi ¹

¹ Careggi University Hospital of Florence, Cell Therapy and Blood Bank, Florence, Italy; ² Careggi University Hospital of Florence, Neurology, Florence, Italy; ³ Careggi University Hospital of Florence, Microbiology and Virology Unit, Florence, Italy; ⁴ Careggi University Hospital of Florence, Department of Experimental and Clinical Medicine, Florence, Italy

Background: In the last 20 years, high dose chemotherapy with autologous hematopoietic stem cell transplantation (aHSCT) has emerged as an effective and safe treatment for aggressive forms of multiple sclerosis (MS). Nevertheless, no data have been published about reconstitution of natural or vaccination-driven immunity after aHSCT and the indication to re-vaccinate patients. We report an analysis of antigen-specific immune recovery in a series of MS patients after aHSCT.

Methods: Blood samples from 22 MS patients who underwent aHSCT in our Center between 2006 and 2017 were analyzed. Each patient underwent mobilization with cyclophosphamide (CTX) 4gr/sqm + GCS- F and was conditioned with BEAM/ATG regimen. Antibody titres for varicella-zoster, measles, rubella, polio, hepatitis B viruses, and tetanus and diphtheria toxins were analysed before mobilization (baseline) and then 2 years after aHSCT. Chemiluminescent Microparticle Immunoassays (CMIA) were performed to determine Rubella and hepatitis B surface antigen antibodies (anti-HBsAg). Enzyme immunoassays were performed to assess Varicella Zoster, measles, diphtheria, tetanus IgG antibodies and IgG against polioviruses. All patients received prophylaxis with acyclovir and Thrimethoprim-Sulphametoxazole for six months after aHSCT. Patients did not receive re-vaccinations in the follow-up period.

Results: All patients showed a complete and sustained engraftment: median (range) time to PMN≥0.5x10⁹/L and platelets≥20x10⁹/L were 12(9-15) and 11(7-14) days, respectively. Patients showed a complete return to a normal lymphocyte subsets counts within two years from transplant. No case of measles, rubella, or chickenpox occurred after transplant, with median follow-up of 40(17-124) months. Table 1 reports antibodies titres at baseline and 2 years after aHSCT. A loss of serum protective immunity was observed in a variable rate of patients with all the tested agents, ranging between 11% in measles and 44% in diphtheria.

Conclusions: Re-vaccination in patients who underwent an autologous HSCT for a severe autoimmune disease is a common practice in many centers and is also recommended in the EBMT guidelines. However, some concerns about this practice was raised for the putative role of vaccinations as a autoimmunity trigger; indeed, no evidences about the loss of protective antibody titer in either natural or vaccination-driven immunity were reported in patients transplanted for autoimmune diseaseas. In our series, a high variability in antibodies titers at baseline was shown; however, most patients showing a protective titre against the tested agents showed a trend to maintain it. A possible approach might be to revaccinate the patients with unprotective titre at 2 years after HSCT; a larger series of data is necessary to provide conclusive evidences on this topic.

Conflict of interest: None of the authors has anything to disclose.

	pre aHSCT	+2 years	pre aHSCT	+2 years	pre aHSCT	+2 years	pre aHSCT	+2 years
	VZV IgG		RUBELLA IgG		MEASLES IgG		antiHBS
negative	1	3	2	3	18	16	11	13
positive	20	16	20	17	3	6	11	9
uncertain	1	3	0	2	1	0
	POLIO IgG		TETANUS IgG		DIPHTHERIA IgG
negative	1	4	0	0	8	11
positive	17	10	15	9	9	5
uncertain	4	8	7	13	5	6

[ [P087 Table] Antibody titres at baseline and 2 years]

P088

Successful unrelated umbilical cord blood transplantation with reduce intensity conditioning for very early onset - Inflammatory bowel disease: an innovative report from China

Xiaowen Zhai, Xiaowen Qian

Children's Hospital of Fudan University, Department of Hematology/Oncology and Bone Marrow Transplant Unit, Shanghai, China

Background: Inflammatory bowel disease (IBD) is a chronic disorder of the gastrointestinal tract. Very early onset IBD (VEO-IBD) represents those severe children with disease onset occurring before 6 years old. Interleukin-10 receptors (IL-10RA, IL-10RB) mutation are considered to be one of the very important genes for VEO-IBD. Currently variant treatment, such as steroid medication, immunosuppressive agents, biological agents could not get complete remission. Allogeneic hematopoietic stem cell transplantation (allo-HSCT) was reported to induce remission in those with VEO-IBD.

Methods: We performed unrelated umbilical cord blood transplantation (UCBT) in 13 consecutive children with VEO-IBD due to IL-10 receptor mutation between 2015 and 2017. Median age of 13 children was 11 months (range, 4 to 46 months), and median body weight was 7 kg (range, 3.2 to 10 kg). All patients received reduced intensity conditioning (RIC) regimen consisting of busulfan, fludarabine and cytarabine. Prophylaxis for graft-versus-host disease (GVHD) was tacrolimus. All patients received a ≤2 HLA alleles-mismatched cord unit, 2 were HLA fully matched, 3 were 9/10 matched, 8 were 8/10 matched. Median nucleated cells of the cord blood were 14.29x10⁷/kg (range, 7.06 to 51.5 x10⁷/kg), and median CD34+ cells were 4.99 x10⁵/kg (range, 1.55 to 18.39 x10⁵/kg).

Results: 10/13 (76.9%) patients engrafted, median time of neutrophil engraftment was 25.5 days, and median time of platelet engraftment was 27 days. Median follow-up time was 6 months (range, 2 to 29 months), 9/13 (69.2%) patients were alive with continuous donor engraftment, and achieved complete clinical remissions. Colonoscopy at 6-12 months after transplantation in 6 children revealed the mucosa healing. 2 patients had grade III/IV acute graft-versus-host disease (GVHD). 4/13 (30.8%) patients died, and the causes of death in 3 patients was infection leading to multi-organ failure after graft failure that occurred at 1-2 months after transplant, 1 patient died of HHV6 virus infection in the lungs and bronchiolitis obliterans.

Conclusions: It is the first clinical trial that unrelated UCBT was performed in VEO-IBD children in China. Our data should unrelated UCBT with RIC should be considered as a potentially curative therapeutic option in children with VEO-IBD.

Conflict of interest: The author declare that they have no conflict of interest.

Cell therapy / cellular therapy

P089

Adoptive immunotherapy with virus specific T cell: Experience of 2 Spanish cell-therapy laboratories

Laura Alonso ¹ , Alba Perez ² , Maria Isabel Benítez ¹ , Luisa Sisinni ³ , Eduardo Lopez Granados ^4,5 , Guillermo Orti ⁶ , Pere Soler-Palacin ⁷ , Raquel De Paz Arias ⁸ , Margarita Codinach ⁹ , Francesc Rudilla ⁹ , Cristina Diaz de Heredia ¹ , Antonio Perez-Martinez ² , Sergio Querol ⁹

¹ Hospital Universitario Materno Infantil Vall d´Hebron, Pediatric Hematology and Oncology Department, Barcelona, Spain; ² Hospital Infantil Universitario La Paz, Servicio de Hemato-Oncología Pediátrica, Madrid, Spain; ³ Santa Creu i Sant Pau Hospital-Universitat Autònoma, Pediatric Hematology, Oncology and HSCT Unit, Barcelona, Spain; ⁴ Hospital Universitario La Paz, Clinical Immunology Department, Madrid, Spain; ⁵ La Paz Biomedical Research Institute-IdiPAZ, Madrid, Spain; ⁶ Hospital Universitario Vall d'Hebron, Hematology, Barcelona, Spain; ⁷ Hospital Universitario Materno Infantil Vall d´Hebron, Unitat de Patologia Infecciosa i Immunodeficiències de Pediatria, Barcelona, Spain; ⁸ Hospital Universitario La Paz, Servicio de Hematología, Madrid, Spain; ⁹ Banc de Sang i Teixits, Barcelona, Spain

Background: Viral infections are a leading fatal complication for immunosuppressed patients, such as patients after hematopoietic stem cell transplantation (HSCT) and patients with primary immunodeficiencies (PIDs). Use of virus-specific T lymphocytes (VSTs) has been successful for the treatment and prevention of viral infections in these settings. Here we describe our experience with the clinical use of VSTs.

Methods: Retrospective review of the lymphoapheresis procedure, IFNγ selection yield and clinical results of the VSTs produced in the Banc de Sang i Teixits and Hospital La Paz between 2014 and 2017.

Results: We assessed the outcome of 15 patients who experienced 17 different viral infections and were treated with virus specific T cells. A total of 27 VST infusions were analysed. Median age was 4.9 years (0.8-36).

Eleven patients received the infusion in the context of post-haematopoietic stem cell transplant viral infections. The indication in the rest of the patients was immunodeficiency- associated CMV infection in 3 patients and EBV chronic active disease in 1.

Donors: Donors were deemed fit to donate and were chosen depending on IFN-γ secretion capacity and HLA match. For 10 of the 11 transplant patients, third-party donors were used. When multiple infusions were required for an episode of infection because of relapse or lack of response, the same donor was used in 5 patients and a different donor in 1.

Toxicity: There were no complications associated to the apheresis procedure. Only one of the donors required the placement of a central venous catheter.

Two patients experienced infusion reaction and one developed aplasia. One of the patients experienced grade 2 acute GVHD secondary to the treatment and died of sepsis.

Response:

1. HSCT patients

EBV VST: One of the patients with EBV PTLD and CSF cleared infection in the CSF but his symptoms did not improve significantly and subsequently relapsed. Two patients did not show any response and received treatment with brentuximab.

CMV VST: Four of the 5 patients with CMV infection cleared the virus, the other patient died of progression.

Adenovirus VST: one patient with viremia responded and cleared the viremia within 3 weeks, the patient with adenovirus disease progressed and died of disease.

EBV + CMV VST: both patients showed response

2. Non-HSCT patients

None of the non-transplant patients with CMV infection responded, 2 of them died of disease. The patient with EBV chronic active disease progressed and died of disease.

Donor cells were detected by chimerism in 4 patients.

Overall outcome: 6 patients died of progression and 1 patient died of sepsis.

Conclusions: The production of VSTs derived from either stem cell donors or third-party donors is feasible and this treatment is likely safe and effective especially for the treatment of viral infections after HSCT.

Clinical Trial Registry: No

Conflict of interest: None

Viral specificity	Number of patients	Number of infusions	Mean CD3 IFNγ/Kg	Mean CD3 nonIFNγ/Kg
CMV	9	16	2,23E+05 (2,29E+03-6,20E+05)	9,89E+03 (5,66E+03-1,40E+04)
EBV	4	6	1,51E+04 (2,57E+03-3,25E+04)	9,38E+03 (2,01E+03-2,50E+04)
Adenovirus	2	3	6,63E+02 (4,02E+02-9,86E+02)	3,28E+03 (2,10E+03-4,46E+03)
CMV + EBV	2	3	1,90E+04 (7,79E+03-3,02E+04)	9,22E+03 (8,51E+03-9,92E+03)

[ [P089 Table] Product characteristics]

P090

Autologous umbilical cord blood in children with autism or in children with cerebral palsy - first applications in Poland

Anna Mucha ¹ , Magda Chroscinska-Krawczyk ² , Monika Kotarska ² , Krystyna Mitosek - Szewczyk ² , Kamila Poteralska ¹ , Joanna Płaczkowska ¹ , Julia Piskuła ¹ , Agnieszka Orczykowska ¹ , Julia Walędzik ¹ , Karolina Zagórska ¹ , Mariusz Grudniak ¹ , Dominika Gładysz ¹ , Iwona Marszałek ¹ , Maciej Boruczkowski ³ , Tomasz Ołdak ¹ , Dariusz Boruczkowski ¹

¹ PBKM, Warsaw, Poland; ² Lublin Medical University, Department of Pediatric Neurology, Lublin, Poland; ³ University of Medical Sciences, Poznan, Poland

Background: Autologous umbilical cord blood (UCB) represents a potentially revolutionary therapy for neurological disorders such as autism spectrum disorders (ASD) or cerebral palsy (CP). Patients with ASD reveal hyperactivity of the immune system, irregular neuronal growth and increased size and number of microglia. Cerebral palsy is the most common cause of physical disability in children. It may occur due to perinatal hypoxic insults, developmental brain abnormalities, genetic diseases, traumatic or infectious causes. The Polish Stem Cell Bank (PBKM) has provided autologous UCB for neuroprotective and neuroregenerative treatment in children with ASD and CP.

Methods: Eleven patients aged from 4 years and 9 months to 10 years and 3 months (medium age: 5 years and 8 months) were enrolled in this medical treatment. Among all patients, there were 4 Romanian children diagnosed with ASD, who received their own UCB collected in years 2010-2012. The UCB units were transferred from Biogenis to PBKM in 2017. The remaining 7 patients were Polish (3 patients with ASD, 4 with CP) and were given own UCB collected in years 2007-2012.

All Romanian patients and 2 Polish patients had 1 UCB injection, the remaining 5 children received 2 UCB injections. Characteristics of autologous cord blood units (median and range) are in the table below.

All patients´ parents provided signed consent before the start of UCB administration.

Results: All Polish patients have been examined by the same psychologist with PEDI-CAT test after 1 and 3 months of UCB administration. Some of the patients are still waiting for the final test (after 6 months), therefore the comprehensive results will be available in 2018. No adverse events were recorded.

Conclusions: The administration, collection and reinfusion of umbilical cord blood (USB) have proven to be safe and well-tolerated treatment possibility for ASD and CP patients.

Conflict of interest:

A. Mucha, K. Poteralska, J. Płaczkowska, J. Piskuła, A. Orczykowska, J. Walędzik, K. Zagórska, M. Grudniak, D. Gladysz, I. Marszalek, T. Oldak, D. Boruczkowski: PBKM employees

WBC [10^3/µl]	CD 34+ [% WBC]	RBC [10^6/µl]	NRBC [10^3/µl]	HCT [%]
26,7 (15,45 - 42,39)	0,23 (0,07 - 0,48)	0,49 (0,11 - 0,94)	0,36 (0,02 - 1,24)	7,11 (1,8 - 14,1)

[ [P090 Table] Table 1]

P091

Automated manufacturing of gene-engineered T cells under serum free conditions

Nadine Mockel-Tenbrinck, Carola Barth, Daniela Mauer, Janina Brauner, Waél Al Rawashdeh, Silke Schult, Olaf Hardt, Georg Rauser, Marion Jurk, Mario Assenmacher, Andrew Kaiser

Miltenyi Biotec GmbH, Bergisch Gladbach, Germany

Background: Manufacturing gene-modified T cells requires robust and reproducible processes that depend on materials and reagents fulfilling specific safety requirements. Generally these processes require the use of human AB serum, a reagent that is commercially limited. For most culture system, it is mandatory to ensure potent T cell expansion yielding high numbers of viable cells. Removing the need for human AB serum, especially in fully automated manufacturing systems during gene-engineered T cell preparation will reduce costs, dependence on reagents of limited availability and reduce potential risks of viral contaminations while increase commercial scalability to wide demand.

Methods: The CliniMACS Prodigy T Cell Transduction process enables the robust generation of gene-modified T cells without the need for serum supplementation when using MACS GMP T Cell TransAct in combination with TexMACS GMP Medium (from Miltenyi Biotec). Furthermore, implementation of this potent soluble polyclonal T cell stimulation reagent, allows for process simplification whereby a manipulation step such as a “bead removal” is rendered unnecessary.

Results: Further developing on a process requiring 3% human AB serum, we now demonstrate the robustness of a serum free process capable of yielding comparably high expansion of viable and functional chimeric-antigen-receptor (CAR) modified T cells. Furthermore, T cell phenotype analysis indicated, that differentiation of selected and activated T cells does not differ for the serum free or the process using serum supplementation. Interestingly, significant higher transduction efficiencies and final CAR T cell numbers were observed using serum free condition. Additionally, no differences in the CAR T cell specificity and functionality, analyzed in vitro and in vivo, were observed.

Conclusions: The fully-automated CliniMACS Prodigy T Cell Transduction process minimize user interactions and enables the preparation of high numbers of functional CAR T cells in a serum free process. Thereby classical process risks, such as viral contamination due to human AB serum, use of different devices or unnecessary manipulations in a non-automated system are reduced to a minimal. Accordingly, these improvements are the next step towards a commercial fully-automated manufacturing of CAR T cells for the treatment of a large number of patients.

Conflict of interest: This project has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 667980 (CARAT).

P092

Bad Enemies, Best Weapons! Cell Therapy in Resistant Cytomegalovirus (CMV) Infection in a Child after Stem Cell Transplant (SCT)

Barbara Torres Guerola ¹ , Juan Montoro Gómez ² , Eduardo Lopez Briz ³ , Sandra Carreño Sánchez ¹ , José María Fernández Navarro ⁴

¹ La Fe, Pediatric SCT Unit, Valencia, Spain; ² La Fe, Hematology, Valencia, Spain; ³ La Fe, Pharmacy, Valencia, Spain; ⁴ La Fe, SCT Unit, Valencia, Spain

Background: Infections remain a major source of morbidity and mortality in the allogeneic stem cell transplant arena. Cord blood transplant recipients account for higher rates of viral infections. Pharmacological treatment is not always successful.

Methods: We present the case of our first patient successfully treated with third party cellular therapy against refractory CMV infection.

4-year-old male with high risk B-ALL, candidate for allogeneic SCT in his second complete remission (CR). No MSD neither MUD available. A 5/8 cord blood unit was chosen. Both donor and recipient were CMV negative. Donor /Receptor = 0+/0+, female/male. Cell dosage= TNC: 14 x 10e7/Kg, CD 34: 6,5 x10e5/Kg. He received prophylactic acyclovir since admission. PCR for CMV detection was performed once a week since conditioning. A positive test in our center is defined by a threshold of 400 copies/mL.

Myeloablative conditioning regimen with thiotepa+ busulfan+fludarabine and ATG (total dose 6 mg/Kg). GVHD prophylaxis: tacrolimus and short-course steroids. No important complications in his early post-transplant period. Engraftment day + 16. Discharged day +18. As per local policy, all blood products were irradiated and filtered before transfusion.

Primary CMV infection on day + 30. Treated with ganciclovir but CMV viremia increased after 2 weeks of adequate therapy. He was then changed to foscarnet. At the same time he presented diarrhea and adenovirus was found in faeces plus mild haematuria with JC/BK viruria, so weekly cidofovir was started in combination with foscarnet. Simultaneously tacrolimus was tapered (by day +60) and discontinued. No evidence of CMV end-organ disease. Blood counts and biochemistry were normal. His leukemia remained in complete remission and the patient had stable full donor chimerism.

The study of CMV resistances showed: CMV Mutations (UL54) A505V, K513R, A688V. CMV antiviral resistance (UL54) FOSCARNET Sensible. GANCICLOVIR Resistent. CIDOFOVIR Resistent.

Facing CMV infection resistant to first and second line treatment, we decided to test the benefit of third party cell therapy. As per local regulations, informed consent was obtained. We contacted with Memorial Sloane Kettering Cancer Center, NY, NY, USA. A suitable cell line of CMV specific Cytotoxic T Lymphocytes (CTL) was identified

Results: Five months post-transplant, our patient received cell therapy. The dose of anti CMV specific CTLs administered was 1x 10e6/kg per day, three doses on days 1, 8 and 15. No immediate side effects registered. A quick decrement on his CMV viral load was observed after second CTL infusion. Further controls have remained negative.

No GVHD has been detected so far nine months post-transplant, four months after specific anti CMV CTL infusion. Regarding immune reconstitution CD4, CD8 and immunoglobulin levels are normal. He remains in complete remission and full donor chimerism. He´s back to school and his performance status is excellent (Lansky 100%).

Conclusions: Specific third party CTLs have played a key role in our patient´s treatment. In the future early inclusion in the management of difficult CMV infections should be considered.

Clinical Trial Registry: NA

Conflict of interest: NA

P093

CAR-T cells targeting α_vβ₃ integrin confer complete remission of epithelial cancers in pre-clinical models in vivo

Lars Wallstabe ¹ , Christoph Rader ² , Hermann Einsele ¹ , Michael Hudecek ¹

¹ Universitätsklinikum Würzburg, Medizinische Klinik und Poliklinik II, Würzburg, Germany; ² The Scripps Research Institute, Department of Immunology and Microbiology, Jupiter, FL, United States

Background: Integrins are heterodimeric transmembrane receptors that act in cell-cell and cell-matrix interactions. Expression of the α_vβ₃ integrin (vitronectin receptor; α_v: CD51 = ITGAV & β₃: CD61 = ITGB3) has been documented in several entities including melanoma, breast and colorectal cancer and glioblastoma where it enhances tumor cell survival and metastasis. Here we evaluated the antitumor function of α_vβ₃-specific CAR-T cells in vitro and a murine xenograft model in vivo.

Methods: Expression of α_vβ₃ integrin was analyzed by flow cytometry. CARs with a targeting domain derived from humanized mAB LM609 (higher vs. lower affinity clone v7 & v11), equipped with either a long or short IgG-Fc spacer (Hinge-CH2-CH3 vs. Hinge-only), each with a signaling module of CD28/CD3zeta (4 constructs) were expressed in lentiviral vectors for gene-transfer into CD8⁺ and CD4⁺ T cells (n=3 donors). CAR⁺ T cells were enriched to >90% purity prior to functional testing.

Results: We confirmed high-level expression of α_vβ₃ integrin on A375 melanoma and MDA-MB-231 breast cancer, and to a lower extent on Jurkat T-cell lymphoma cell lines. We observed specific and very potent lysis mediated by α_vβ₃ CAR-T cells, e.g. A375 there was >60% specific lysis within a 4-hour co-culture with CD8⁺ CAR-T cells (E:T ratio = 20:1). Further, we observed high-level production of IFN-γ by CD8⁺ and CD4⁺ CAR-T cells, and production of IL-2 by CD4⁺ CAR-T cells after stimulation with tumor cells, which translated into sustained viability and productive proliferation of CAR-T cells in vitro. Overall, we observed maximum antitumor reactivity with the α_vβ₃ CAR with short spacer and high-affinity targeting domain (v7).

We engrafted firefly luciferase positive A375 into NSG mice, which led to metastatic tumor growth in all animals. Mice were then treated with a single dose of α_vβ₃ CAR-T cells (5x10⁶, CD8:CD4 ratio = 1:1), which led to complete resolution of all tumor lesions, and complete survival of the CAR-T mouse cohort until the end of the observation period. Mice treated with control T cells showed rapid tumor progression and inferior survival (n=6 mice per group; p< 0.05). CAR-T cell engraftment and in vivo persistence for >30 days after adoptive transfer was documented by flow cytometry, and cytokine production was documented by multiplex cytokine analysis from mouse serum. The strongest antitumor effect in vivo was mediated by the α_vβ₃ CAR with short spacer and low-affinity targeting domain (v11).

Conclusions: Our data demonstrate that CAR-T cells targeting α_vβ₃ integrin confer potent antitumor functions in vitro and in vivo. In line with our previous work, we found that extracellular spacer design and receptor affinity affected tumor recognition and were optimized in order to derive a CAR construct with maximum reactivity. Intriguingly, α_vβ₃ is also expressed on tumor vasculature and cancer-associated fibroblasts, suggesting α_vβ₃ CAR-T cells may be capable of efficiently migrating through endothelial and connective tissue barriers in the tumor microenvironment to exert their antitumor effect.

Conflict of interest: None

P094

CAR-T therapy shapes T cell repertoire in B-cell acute lymphocytic leukemia

Xiujian Wang ¹ , Tao Sun ² , Xiao Liu ³ , Pengfei Xu ⁴ , Chao Jin ⁴ , Lijuan Ding ¹ , Fang Ni ¹ , Yongxian Hu ¹ , Jian Yu ¹ , Yunyun Deng ⁴ , Hao Zhang ¹ , Zuyu Liang ¹ , He Huang ¹

¹ Zhejiang University School of Medicine, Bone Marrow Transplantation Center, The First Affiliated Hospital, Hangzhou, China; ² University of Chicago, Department of Surgery, Chicago, IL, United States; ³ University of Chicago, Institute for Molecular Engineering, Chicago, IL, United States; ⁴ Immuquad Biotechnologies, Hangzhou, China

Background: Chimeric antigen receptor (CAR) T cells targeting the B-cell antigen CD19 have displayed potent anti-leukemia activities in refractory/relapsed acute lymphocytic leukemia (ALL). However, the influence on autologous immune system caused by CAR-T cells has not been well examined and elucidated. Here, we applied high-throughput TCR sequencing technology to assess the systemic and dynamic change of T cell repertoire induced by CAR-T cell therapy in Chinses B-ALL patients.

Methods: 3 female patients in our clinical trial were under observation before and after CAR-T administration, and achieved complete remission (CR) as clinical termination. During CAR-T treatment, they all experienced grade 2 to 3 cytokine release syndrome (CRS). The average age of the patient is 36 years old. The peripheral blood sample, bone marrow sample and the CAR-T sample before infusion were obtained to be tested. The TCR full length mRNAs were deeply sequenced using the ImmunHub^TM TCR profiling system (ImmunQuad Biotech). Briefly, a 5' RACE unbiased amplification protocol was used. An algorithm was applied to raw sequencing data for PCR and sequencing errors correction and V, D, J, C gene segments mapping with IMGT^®.

Results: We observed the decreased diversity and increased clonality of T cell repertoire in peripheral blood and bone marrow compared to the pre-treatment controls measured by invsimpson. The TCR sequences of the enriched T cell clones in blood and bone marrow after CAR-T therapy were not found in CAR-T cell pool. Meanwhile, we found these treatment-associated top TCR sequences were at lower frequency in pre-treatment samples. What's more, The TCR sequences of the enriched T cell clones could maintain after CAR-T therapy. Additionally, CAR-T therapy increased the similarity, which was calculated using BUB index, of T cell repertoires between bone marrow and blood and substantially influenced the diversity and clonality of T cell repertoire.

Conclusions: Our investigation showed the big changes of T-cell diversity and clonality in patient's immune system and the post-treatment homogenization of T cell repertoires with clonal expansion in bone marrow and blood companying CAR-T cell administration. It implied that the lysis of tumor by CAR-T cell therapy might facilitate the priming and clonal expansion of patient's tumor antigen-specific T cells systematically. These enriched autologous T cells may enlarge the therapeutic effect with synergy.

Conflict of interest: None of the authors has anything to disclose.

[P094 Figure]

$$$$

P095

changes of T lymphocyte subsets after CAR-T treatment and the possible mechanisms

Lijuan Ding ¹ , Zuyu Liang ¹ , Yongxian Hu ² , Jian Yu ² , He Huang ²

¹ Zhejiang University, Hangzhou, China; ² Bone Marrow Transplantation Center, The First Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, China

Background: Chimeric antigen receptor-modified T cell (CAR-T) therapy, as one of the most promising targeted immunotherapies, has drawn extensive attention and resulted in multiple applications. According to published studies, CD19-directed CAR-T cells (CD19 CAR-T) can reach a complete remission (CR) rate of 93% and 63% in refractory/relapsed ALL and lymphoma, respectively. In tumor cell immunity, T cells play a central regulatory role. T cells are divided into several major subclasses according to different cell surface molecules, maintaining the body´s immune balance.

Methods: We collected peripheral blood samples from patients with refractory relapsed acute lymphoblastic leukemia and lymphoma at different stages before and after CAR-T treatment. Flow cytometry was used to detect T lymphocyte subsets in peripheral blood, in order to investigate the impact of CAR-T treatment on tumor cell immunity and its possible mechanisms.

Results: The figures below show the changes of T lymphocyte subsets after CD19 CAR-T

therapy. Figure A: CD4+ and CD8+ T lymphocytes did not increase significantly, while activated CD4+ and CD8+ T lymphocytes increased, with more obvious elevation in CD8+ subset. Figure B: Th1 was the only helper T lymphocyte subset during CRS, which explained why IFN-γ became the dominant cytokine.

Conclusions: Activated CD4+ and CD8+ T lymphocyte were significantly increased during CRS, as well as Th1 subset. The three cell subsets bear the task of killing tumor cells and releasing cytokines.

Conflict of interest: None of the authors has anything to disclose.

[P095 Figure]

$$$$

P096

clinical-grade Aspergillus-specific T cells: Discontinued production of the lysate enforces the use of peptide pools and the Switch from magnetic selection to short term expansion

René Geyeregger ^1,2 , Sabine Tischer ^3,4 , Nelli Frank ¹ , Christine Hoffmann-Freimüller ¹ , Julia Stemberger ¹ , Britta Maecker-Kolhoff ^4,5 , Rainer Blasczyk ^3,4 , Gerhard Fritsch ^1,2 , Britta Eiz-Vesper ^3,4

¹ Children´s Cancer Research Institute, Clinical Cell Biology and FACS Core Unit, Vienna, Austria; ² Medical University of Vienna, Vienna, Austria; ³ Hannover Medical School, Institute for Transfusion Medicine, Hannover, Germany; ⁴ Hannover Medical School, Integrated Research and Treatment Center (IFB-Tx), Hannover, Germany; ⁵ Hannover Medical School, Department of Pediatric Hematology and Oncology, Hannover, Germany

Background: Aspergillus-fumigatus (Asp) infections constitute a major cause of morbidity and mortality in patients after haematopoietic stem cell transplantation. We obtained the manufactering liscence to generate clinical-grade Asp-specific T cells via the Interferon-gamma (IFN-g) Cytokine Capture System(CCS) using a GMP-compliant Asp-specific lysate. Due to an unexpected discontinued production of the lysate with the end of 2017, we started to find alternatives and tested the use of Asp-specific peptide-pools for cytokine secretion assay (CSA) as well as with the already established GMP-coform short term expansion (STE) method.

Methods: Healthy volunteers of the alloCELL registry were screened for the presence of Asp-T-cells via IFN-g Elispot assay by stimulating PBMCs with either lysate (n=320) or 7 different Asp-specific overlapping peptide-pools (n=30) (Cat, Crf1, f22, Gel1, pmp20, SHMT and SOD). PBMCs from leukapheresis (n = 8) were prepared in Hannover for the IFN-g-CSA, remaining cells were sent to Vienna to perform STE. For the IFN-g-CSA, 1x10⁷ cells were stimulated for 16h with GMP-conform lysate or for 4h peptide-pools followed by magnetic selection of IFN-g-producing Tcells. Cells were characterized for phenotype and function by flow cytometry. For the STE, 2x10⁷ cells were stimulated for 12 days with either the lysate or peptide-pools and IL-15. The final cell products were characterized via flow cytometry, IFN-g-EliSpot and IFN-g/granzyme-B FluroSpot assays.

Results: Asp-T-cells were detectable (detection limit: >2 Spots/2.5x10E5 PBMCs) in 63% of all volunteers after lysate stimulation but undetectable in >85% of healthy donors if peptide-pools were used. The purity of lysate-stimulated IFN-g+ Asp-T-cells after magnetic enrichment using the IFN-g CSA was 49% (range 17.8%-84.5%),while stimulation with the overlapping peptide pools resulted in a non sastisfactory enrichment. In contrast, the STE revealed highly specific Asp-T-cell spot forming colonies(SFC)/10⁵ cells) for both the lysate (863±153/10⁵ cells) and combined peptide-pools (833±138 SFC/10⁵ cells), respectively. Independent of the stimuli used, predominantly CD4+ T-cells were expanded (80%±3,3 vs 78%±6,3) compared to CD8+ T-cells (17,1%±3,1 vs 18,4%±5,6) among CD3. Whereas CD4+ T-cells were mainly of central-memory type, CD8+ T-cells comprise mainly effector-memory T-cells. Target cells were highly functional and cytotoxic as determined by secretion of granzyme-B and IFN-g.

Conclusions: Despite the discontinuation of lysate production and consequently enrichment of Asp-T-cells via the IFN-g CCS, clinical-scale production of peptide pool-expanded Asp-T-cells using the STE method might enable an efficient adoptive T-cell immunotherapy.

Conflict of interest: No conflict of interest

P097

Clinical-grade depletion of T-cell receptor alpha/beta and CD19 positive cells from stem cell products with the CliniMACS Plus and CliniMACS Prodigy devices

Marienn Reti ¹ , Attila Szederjesi ¹ , Szabolcs Tasnady ¹ , Zita Farkas ¹ , Laszlo Gopcsa ¹ , Arpad Batai ¹ , Aniko Barta ¹ , Csaba Kassa ² , Krisztian Kallay ² , Gergely Krivan ² , Tamas Masszi ³ , Peter Remenyi ¹

¹ St Istvan & St.Laszlo Hospital, Hematology and Stem Cell Transplantation, Budapest, Hungary; ² St. Istvan & St.Laszlo Hospital, Pediatric Hematology & Stem Cell Transplantation, Budapest, Hungary; ³ Semmelweis University, 3rd Clinic of Internal Medicine, Budapest, Hungary

Background: The hematopoietic stem cell transplatation is an increasingly used technique for treatment of several hematologic diseases. Cells harvested from peripherial blood (PBSC) or bone marrow (BM) contain cells, which can cause life-threatening graft versus host disease (GvHD). Donor immune cells responsible for GVHD can be selectively removed from the graft using immunomagnetic cell separation. For clinical grade cell depletion, the CliniMACS® Plus (Miltenyi Biotec, Bergisch-Gladbach, Germany) has been the gold standard method so far. However, as a disadvantage, time consuming, extensive manual manipulation is necessary before starting the CliniMACS® Plus separation. Recently, a fully automated, functionally closed system technique - CliniMACS Prodigy^® - was developed, avoiding any manual manipulation requirement. The aim of our study is to analize and compare the results of TCRαβ⁺/CD19⁺ cell depletion obtained with the CliniMACS Plus (CM) and the CliniMACS Prodigy Systems (PD).

Methods: During the last 4 years (Jul 30, 2014 - Dec 13, 2017) 31 TCRαβ⁺/CD19⁺ cell depletions (CM:6, PD:25; BM:1, PBSC:30) were performed for 28 patients (pts) (pediatric :12, adult:16).

Results: The results (median values and ranges are shown) were comparable for nucleated cell viability: CM:98% (93-99), PD:97% (84-99), p=0,56; TCRγδ⁺ cell recovery: CM:65% (31-124), PD:69% (21-97), p=0,81; NK cell recovery: CM:69% (46-94), PD: 70% (33-92), p=0,95; TCRαβ⁺ cell depletion: CM:4,1^log (2,6-4,6), PD: 3,9^log (3,4-4,6), p=0,99; and CD20⁺ cell depletion: CM:3,0^log (2,0-3,1), PD: 2,9^log (1,9-3,6), p=0,80; respectively. However we obtained significantly higher CD34⁺ cell recovery using PD system: CM:70% (67-85), PD:81% (57-95), p=0,05.

The cell compositions of the final products were comparable without statistical difference. The median graft cell counts were the following (x10⁶/recipient body weight kg): CD34⁺ cell content: CM:4,7 (2,03-95,82), PD:7,6 (1,63-24,9) p=0,51; TCRαβ⁺ cell content: CM:0,016 (0,000-0,0879), PD:0,042 (0,016-0,042) p=0,47; TCRγδ⁺ cell content: CM:4,5 (1,0-10,8), PD:4,8 (1,0-57,0) p=0,33; CD20⁺ cell content: CM:0,046 (0,031-0,308), PD:0,088 (0,011-0,442) P=0,57.

30/31 grafts were transfused, 1 pt died just before finishing graft manipulation. 5 pts (CM:3, PD:2) needed a 2^nd graft (CD34⁺ selected, only) due to suboptimal stem cell content.

There were 5 non-takes and 6 deaths. 22/28 pts are alive with a median follow up of 89 days (range:3-1205).

Conclusions: The TCRαβ/CD19 depletion is a increasingly used separation technique in our clinical practice. It could be shown that both, CM and the recently developed fully automated PD system are highly efficient techniques for graft manipulation. PD proved to be at least as good or slightly better than the old CM system. Due to its simplicity and the comparable cost, PD system seems to be the first choice technique in the future.

Conflict of interest: Congress support for M. Reti, A. Szederjesi and S. Tasnady by the Hungarian distributor of Miltenyi Biotec.

P098

Comprehensive standards for extended quality controls of hematopoietic stem and progenitor cells in the context of ATMP production

Lisa Olfe, Marlene Luther, Mohamed El-Missiry, Katy Haussmann, Carola Tietze-Bürger, Lena Oevermann, Kamran Movassaghi, Lutz Uharek

Charité Campus Virchow Berlin, Berlin, Germany

Background: Hematopoietic stem and progenitor cells (HSPCs) are increasingly used as raw material for advanced therapy medical products (ATMPs), i.e. genetic engineered HSPCs for treatment of hemoglobinopathies. However, there are yet no comprehensive and clearly defined standards for extended quality controls. Moreover, successful ATMP development and in process controls depend on optimal subset definition.

Methods: A standardized multicolor flow cytometry panel and gating strategy was developed based on the work of Fritsch (Transfusion, 2017). We validated this method by characterizing the subset distribution of CD34+ HSPCs in different compartments: bone marrow (BM, n=5, median age=8yr), autologous stem cell grafts mobilized with G-CSF alone (autoG, n=18, median age=52yr) or G-CSF plus Plerixafor (autoGP, n=10, median age=56yr) and G-CSF mobilized healthy allogeneic stem cell donors (alloG, n=5, median age=34yr).

Results: Most importantly, our data confirmed the feasibility, robustness and reproducibility of the approach. In particular, we found significant differences in the distribution of CD34+ subsets: the percentage of CD10+ B-lymphoid precursors was high in BM while negligible in mobilized peripheral blood stem cells (61% in BM vs. 0,3 % in autoG, 3,4% in autoGP and 0,6% in alloG, p< 0,05). Frequency of primitive multi-potent progenitors (MPP) was significantly higher in autoG and alloG compared to BM (52% and 64% vs. 7,5%, p< 0,05). Interestingly, frequencies of MPPs also differed between autoG and alloG (p< 0,05). Frequencies of CD38-CD45RA-CD90+ long-term repopulating cells were comparable in autoG (9,9%), autoGP (15,5%) and alloG (12,2%), but appeared to be lower in BM (5,1%). CD184 expression was significantly higher on BM CD34+ cells (63,7%) compared to peripheral blood stem cells (autoG 4%, autoGP 12,9%, alloG 5,3%, p< 0,05) and significantly higher in autoGP compared to autoG (p< 0,05).

Conclusions: Our approach allows to access differences in subset distribution of HSPCs (i.e CD34+CD38-CD45RA-CD90+ HSCs) for quality control and product characterization in the context of ATMP production. So far, our data indicate that there are important differences in subset composition with regard to source and mobilization agent. Furthermore, our data revealed differences between healthy donors and pretreated patients which might be of critical impact for further processing/genetic engineering and in vivo functionality.

Conflict of interest: None of the authors has anything to disclose.

The research being reported in this abstract was supported by Beckman Coulter and may support development of a commercial product. Beckman Coulter neither has influenced nor takes responsibility for the content of this abstract. The terms of this collaboration have been reviewed and approved by the Charite Berlin in accordance with its policy on objectivity in research.

P099

Cord blood donor lymphocyte infusion to boost immune reconstitution and the outcome of an infant ALL with MLL rearrangements after transplantation

Laura Alonso ¹ , Cristina Diaz de Heredia ¹ , Laura Soria Guerrero ² , Isabel Badell Serra ³ , Guillermo Ortí ⁴ , Rodrigo Martino Bufarull ⁵ , Christelle Ferrà Coll ⁶ , Isabel Sanchez-Ortega Sanchez ⁷ , Laura Medina Marrero ⁸ , Nerea Castillo ⁸ , Ramón Gimeno Martínez ² , Sergio Querol Giner ⁸

¹ Hospital Materno-Infantil Vall d'Hebron, Barcelona, Spain; ² Hospital del Mar & IMIM (Hospital del Mar Medical Research Institute), Immunologia, Barcelona, Spain; ³ Hospital de la Santa Creu i Sant Pau, Pediatrics Service, Barcelona, Spain; ⁴ Hospital Vall d’Hebron, Hematology Service, Barcelona, Spain; ⁵ Hospital de la Santa Creu i Sant Pau, Hematology Service, Barcelona, Spain; ⁶ Institut Català d’Oncologia Trias i Pujol, Badalona, Spain; ⁷ Institut Català d’Oncologia Duran i Reynals, L'Hospitalet de Llobregat, Spain; ⁸ Banc de Sang i Teixits, Cell Therapy Services, Barcelona, Spain

Background: Donor lymphocyte infusion (DLI) is commonly used to treat incipient relapse after transplantation but not possible using cord blood as a source of cells. We have developed a phase I-II clinical trial (NCT02328885) to assess the ability of cord blood donor lymphocyte infusion (CB-DLI) to boost immune vigilance using the 20% fraction when a 80/20 cryobag is available. This clinical trial is under evaluation. Here, we present a case report to show a potential effect of the CB lymphocytes in this setting.

Methods: Clinical case. Three month-old girl diagnosed with Infant ALL with MLL rearrangement was made to proceed to allogeneic hematopoietic stem cell transplant in first CR, when she was 8 months old, after poor response to chemotherapy. The source was a 5/6 matched unrelated cord blood unit. Patient fulfilled inclusion criteria to this clinical trial Conditioning regimen consisted of ATG, thiotepa, busulfan and fludarabine. Cells infused using the 80% fraction of a CB unit were 22.6E7 TNC and 6.8E5 CD34/kg resulting in an early engraftment. Post-transplant course was complicated with grade 2 acute skin and gut GVHD which responded to methylprednisolone. Disease assessment at 3 months post-transplant demonstrated full donor chimerism; flow cytometry was negative and MLL PCR was 0.2%. Immunosuppression was successfully weaned off and 4 months after HSCT, she was free of GVHD and off immunosuppression, she had persistent full donor chimerism and flow cytometry on bone marrow aspirates was negative for leukemic cells. However, MLL PCR was still detectable at a level of 0.067% and her CD4 counts were very low (0.08 x10E9/l). Given the poor immune reconstitution and the risk of relapse, decision was made to infuse the 20% fraction of the cord blood unit (day +140). CD3+ cells/kg infused were adjusted to 1E6 CD3/kg.

Results: A rapid expansion on the T lymphocyte compartment was observed as early as 15 days after infusion. Three-weeks after infusion patient developed gut GVHD that required treatment with corticosteroids and cyclosporine until 6 months after the infusion. This T lymphocyte expansion was transient and numbers went back to previous levels. A second wave of expansion was noted two months later and was equally of short length. In both cases a KI67+ population was observed. The expansion was dominated by the CD4 subset of T lymphocytes, mostly by those bearing memory markers. Interestingly, her MLL PCR levels have fluctuated and became negative after 6-months of cord blood DLI. She is now leukaemia-free almost 3 years post-HSCT. An ELISPOT assay was routinely performed to assess reactivity in front of a panel of defined antigens. Interestingly, reactivity against a tumor peptide collection was detected along the follow-up.

Conclusions: Cord blood DLI seems to be feasible when a CBU is stored in a 80/20 bag and there is sufficient cell dose in the 80% fraction. In spite there is a risk for GVHD that may need an adjustment of CD3 dose, this strategy could be an easy way to boost immunity to control minimal residual disease after engraftment.

Clinical Trial Registry: ClinicalTrials.gov Identifier: NCT02328885

https://www.clinicaltrials.gov/ct2/show/NCT02328885?term=NCT02328885&rank=1

Conflict of interest: Authors declare no conflict of interest

P100

Co-transplantation of third party donors-derived mesenchymal stem cells in hematopoietic stem cell transplantation in patients with severe aplastic anemia: A single institution experience

Arzu Akcay ¹ , Didem Atay ¹ , Fatih Erbey ¹ , Ercument Ovali ² , Gulyuz Ozturk ¹

¹ Acibadem University, Istanbul, Turkey; ² Acibadem Altunizade HospitalLabcell, Istanbul, Turkey

Background: Hematopoietic stem cell transplantation (HSCT) for severe aplastic anemia (SAA) is mainly limited by the high incidence of graft failure and GvHD. Mesenchymal stem cells (MSCs) have been shown to support hematopoiesis in vivo and to display potent immunosuppressive effects to prevent or treat GvHD after HSCT. In this study, we aimed to determine the potential benefits of MSC co-transplantation in patients with SAA undergoing HSCT, regard to the outcomes engraftment and GVHD.

Methods: A total of 10 patients (median age 117 months, range 21-213) were carried out: 9 allogenic (match sibling/family donor=4, match unrelated donor=5, haploidentical donor=1). The conditioning regimen included fludarabin, cyclophosphamide in 7 patients. Thymoglobulin (ATG) were used in all patients. The recipients received cyclosporin A (CsA) and short-term methotrexate for GvHD prophylaxis in 7 patients, the other drugs in 3 patients. MSCs (1×10⁶cells/kg) were administered by venous infusion 24 h before the hematopoietic graft. Bone marrow (in 6 pts) and peripheral blood (in 4 pts) CD34⁺ cells were infused intravenously. Median CD34⁺ counts were 4.18 × 10⁶/kg.

Results: No patients experienced infusional toxicity during the infusion of MSCs. In all patients achieved hematopoietic reconstitution and sustained full donor chimerism in 9 pts, mix chimerism in 1 pts. The median time for myeloid engraftment was 13 days (range 9-25 days) and for platelet engraftment was 23 days (range 10-108 days). The incidence was 10% for grade I- II, 20% for III-IV acute GvHD and 10% for chronic limited GvHD. Nine patients were alive, but a patient died from infection. Patients were followed-up for a median of 15,5-month (range 2-60).

Conclusions: Despite limited cases enrolled in the study, our findings support previous studies showing that HSCT combined with MSCs infusion could be an effective approach to reduce the risk of graft failure and GvHD in HSCT for SAA. MSC occupy a hematopoiesis-supportive capacity.

Conflict of interest: None of the authors has anything to disclose.

P101

Dissecting EBV-specific T-cell responses after successful allogeneic EBV-specific T-cell transfer for CNS PTLD

Britta Maecker-Kolhoff ^1,2 , Rebecca E. Schultze-Florey ^1,2 , Sabine Tischer ^2,3 , Leonie Kuhlmann ⁴ , Patrick Hundsdoerfer ^5,6 , Arend Koch ^6,7 , Sarina Ravens ⁴ , Lilia Goudeva ³ , Christian Schultze-Florey ⁸ , Christian Koenecke ⁸ , Rainer Blasczyk ³ , Ulrike Koehl ^2,9 , Hans-Gert Heuft ³ , Immo Prinz ⁴ , Britta Eiz-Vesper ^2,3

¹ Hannover Medical School, Pediatric Hematology and Oncology, Hannover, Germany; ² Integrated Research and Treatment Center Transplantation (IFB-Tx), Hannover, Germany; ³ Hannover Medical School, Institute for Transfusion Medicine, Hannover, Germany; ⁴ Hannover Medical School, Institute of Immunology, Hannover, Germany; ⁵ Charité Universitaetsmedizin Berlin, Pediatric Hematology and Oncology, Berlin, Germany; ⁶ Berlin Institute of Health, Berlin, Germany; ⁷ Charité Universitaetsmedizin Berlin, Neuropathology, Berlin, Germany; ⁸ Hannover Medical School, Hematology, Oncology, Hemostaseology and Stem Cell Transplantation, Hannover, Germany; ⁹ Hannover Medical School, Cell Therapy Center, Hannover, Germany

Background: EBV-associated post-transplant lymphoproliferative disease (PTLD) with CNS involvement is a severe complication after solid organ transplantation. Standard treatment with reduction of immunosuppression, anti-CD20 antibody application and/or chemotherapy often fails leading to poor outcome. Transfer of EBV-directed T-cells has shown promise, however, contributions of transferred and endogenous T-cells are largely unknown.

Methods: Manufacturing of clinical-grade EBV-specific CD4+ and CD8+ T-cells from a 5/10 HLA-matched third party donor was conducted using ppEBNA-1 and ppSelect and the IFN-γ Cytokine Capture System (Miltenyi Biotech) on a CliniMACS device.

Monitoring of EBV load was performed by routine qPCR. EBV-specific T-cell frequencies pre/post transfer were quantified by IFN-γ ELISpot using the peptide pools ppEBNA1, ppSelect, ppLMP2a, ppBZLF1 (Miltenyi Biotec). To enhance sensitivity analysis was repeated after 7-day-expansion in the presence of ppEBNA1 or ppConsenus and 50U/ml IL-2 (Peprotec).

For TCR β chain repertoire analysis expanded PBMCs were stained with antibodies against DAPI, hCD45, hCD3, hTCR αβ and hCD8. Cells were sorted into TCRαβ+/CD8+ T-cells on a FACS Aria Fusion flow cytometer. RNA extraction (RNeasy Plus Micro Kit, QIAGEN) and reverse transcription (SMARTer RACE 5'-3' PCR Kit, Clontech) were carried out according to the manufacturer's recommendation. For combined amplification of the TCR β CDR3-region and Illumina adaptor sequences the Advantage-2 PCR Kit (Clontech) was used. Indexing of the samples was performed with Nextera Primer Kit (Illumina) and the product was purified with the Agencourt AMPure XP Kit. Denaturation and dilution of the pool was done as described (Illumina MiSeq Dilution and Denaturation Guide). Next generation sequencing was performed on the Illumina MiSeq System. For analysis the FastQ files were annotated at IMGT/HighV-Quest database and processed with tcR-package and VDJtools.

Results: An 11-year-old boy developed multilocular EBV-positive CNS PTLD ten years after liver transplantation. Complete remission was achieved by repeated intravenous and intrathecal anti-CD20 antibody rituximab administration combined with intrathecal chemotherapy (methotrexate, cytarabine, prednisone) over a time period of three months. Due to the poor prognosis of CNS PTLD and lack of EBV-specific T-cells in patient's blood, we decided to perform EBV-directed T-cell immunotherapy as a consolidating treatment. The patient received five infusions of allogeneic EBV-specific T-cells (25.000 / kg bw each) manufactured by Cytokine Capture System from a 5/10 HLA-matched unrelated third party donor. No relevant acute toxicity was observed. EBV-specific T-cells became detectable after first injection and increased during the treatment course. NGS TCR-profiling verified the persistence and expansion of donor-derived EBV-specific clones. After two transfers epitope spreading to unrelated EBV antigens occurred suggesting onset of endogenous T-cell production, which was supported by detection of recipient-derived clones in NGS TCR-profiling. Continuous complete remission was confirmed 27 months after initial diagnosis.

Conclusions: Repetitive transfer of partially HLA-matched EBV-specific T-cells from a third party donor evoked robust EBV-specific T-cell responses in a high risk PTLD patient. While T-cell specificity broadened over the treatment course, individual TCR β sequences could be attributed to both donor and recipient progeny. Molecular tracking of individual T-cells will help understand the contributions of transferred vs. endogenous virus-specific T-cells in patients receiving adoptive T-cell transfer.

Conflict of interest: The authors declare no conflict of interest. This work was supported by the German children's cancer fund (DKS 2011.05) and the German Federal Ministry of Education and Research (01EO1302). PH is participant in the BIH‐Charité Clinical Fellows Program funded by the Charité - Universitätsmedizin Berlin and the Berlin Institute of Health.

P102

Abstract previously published

P103

Flow cytometric analysis of HLA disparities as a highly efficient tool to detect chimerism in rare cell populations after non-identical transplantation and adoptive cell transfers

Rebecca Schultze-Florey ¹ , Sabine Tischer ² , Wolfgang Kühnau ³ , Britta Eiz-Vesper ² , Britta Maecker-Kolhoff ¹

¹ Hannover Medical School, Ped. Hematology Oncology, Hannover, Germany; ² Hannover Medical School, Transfusion Medecine, Hannover, Germany; ³ Hannover Medical School, Institut for Human Genetics, Hannover, Germany

Background: Chimerism analysis after HLA-mismatched stem cell transplantation is an important diagnostic tool for clinical decisions regarding modification of immunosuppression. In particular, flow-cytometry based detection of HLA disparities offers an attractive option to detect chimerism in lymphocyte subsets following HLA-mismatched cell transfers. Here we present a broadening of the technique to detect rare virus-specific T cells in peripheral blood after adoptive cell transfer by flow cytometry.

Methods: PBMCs from residual blood samples from platelet (PLT) apheresis HLA-A01⁺/A02^- CMV⁺ (donor 1) and HLA-A01^-/A02⁺ CMV⁺ (donor 2) individuals from the alloCELL donor registry were isolated by Ficoll density centrifugation. 1x10^7 PBMC from donor 1 and 2 were stimulated separately with pp65 and incubated overnight in a 24-well plate. After stimulation total cells of the donor 1 and 2 were mixed in ratios of 1:1 and 1:10 (donor 2: donor 1). The IFN-γ positive T cells were isolated via cytokine secretion assay (CSA, Miltenyi biotech) and analyzed via flow cytometry (mAb against CD45, CD3, CD4, CD8, IFN-γ, HLA-A01, HLA-A02). Simultaneously a sample of the enriched IFN- γ + CMV specific T cells was analyzed for molecular chimerism by STR PCR.

Results: Before magnetic enrichment the 1:1 mixed IFN-γ positive T cells (CD45+, CD3+, CD56-) displayed a 17 % HLA-A01+ (donor 1) phenotype and 59% HLA-A02+(donor 2) phenotype. After magnetic enrichment the distribution of HLA-A01+/HLA-A02+ cells was similar (17% and 57%, respectively; 1x10^6 cells analyzed). The 1:10 (donor 2 : donor 1) mixed IFN-γ positive cell suspension showed before enrichment a distribution of 12% and 75% (after enrichment 11% and 71%). In comparison, STR analysis of enriched cells (1x10^4 cells analyzed) revealed for the 1:1 mixture 30% of the donor 1 cells and 70% of the donor 2 cells. For the 1:10 mixture cells from donor 1 were 70% and from donor 2 30 %. Similar results were obtained in a second HLA-A2 mismatched donor pair. While results of FACS chimerism were available within 36 hours, molecular chimerism after magnetic enrichment took 4 days.

Conclusions: The increasing usage of HLA-mismatched transplantations requires fast and reliable determinations of cell chimerism. FACS analysis of mismatched HLA alleles provides a rapid technique to detect chimerism in subpopulations avoiding the need for cell purification. Sensitivity is comparable to STR chimerism analysis. FACS chimerism can be combined with techniques of functional single cell identification such as cytokine secretion, which provides a valuable tool for monitoring of antigen-specific cells after partially HLA-matched transplantation and adoptive cell transfer. Prospective studies are needed to determine the fate of partially HLA-matched antigen-specific T cells in adoptive transfer situations.

Conflict of interest: All authors declare that there is no conflict of interest.

P104

Flow cytometry or Luminescence assay: Which the best quality control of extracorporeal photopheresis ?

Judith Desoutter, Christèle Ossart, Aline Regnier, Lavinia Merlusca, Delphine Lebon, Jean Pierre Marolleau, Marie Noëlle Lacassagne

CHU Amiens Picardie, Amiens Cedex, France

Background: In agreement with good practices for therapeutic use of human cells, quality control has to be performed to valid the process of extracorporeal photopheresis (ECP). Currently, the reference method to highlight inhibition of cell division after ECP, remains used of tritiated thymidine incorporation that requires a suitable environment for radioactivity assays. In last few years, alternative methods such as CFSE staining by flow cytometry or luminescent cell viability assay had been developed. Here, we report results, advantages and drawbacks about this both assays designed in our laboratory.

Methods: To perform CFSE staining by flow cytometry, we opted for CellTrace™ CFSE kit (Invitrogen) that, by binding covalently to intracellular amines, used to assess cell proliferation. For the luminescent cell viability test, we used CellTiter-Glo® Assay (Promega) which allows determining number of viable cells in culture, based on quantitation by a luminometer of the ATP present for reporting the presence of metabolically active cells. Both methods were performed using patients cells collected by cytapheresis at the beginning (PRE-ECP) and after 8-MOP (Metoxsalene, SALF ; Macopharma) irradiation cells (POST-ECP) reinjection to patient. Cells obtained were typed initially by flow cytometry to appreciate their viability (7-AAD) and their distribution of mononuclear cells (MNCs) subpopulations. Then, they were plated and incubated at 37°C with 5% CO₂ in divided two groups : non stimulated cells and stimulated cells with a solution of 10 µg/mL PHA (Phytohemagglutinin-P, Sigma-Aldrich) and 50 ng/mL of IL-2 (Human IL-2, Miltenyi Biotec). To demonstrate ECP efficacy, to our specifications, cell proliferation difference between stimulated and non-stimulated PRE-PCE cells (PROLIF) must be greater than 3% and in the same time, inhibition of cell division between PRE-PCE and POST-PCE cells (INH) will be greater than 97%.

Results: CFSE assays were performed, specifically to CD3⁺ PRE-ECP and POST-ECP viable cells (CD3⁺/7-AAD^-) and were read at day 0, to verify CFSE labelling, and at day 2-3 after ECP. In our 9 patients already analyzed, median of PROLIF was at 54.6 + 8.8% and median of INH was 98.8 + 1.0%. In contrast, luminescent tests analyzed all type of viable MNCs (7-AAD^-) collected and were read at day 5-6. In our 9 patients studying, median of cell PROLIF was at 3.9 + 0.6% and median of INH was 100 + 0.17%. To note, in CFSE test, one patient was out of INH specification (INH = 90.9%), with a hematocrit at 2.1% (range to 0.5% to 2.1% across assays).

Conclusions: We show a low PROLIF in luminescent assay, staying in our specification, caused by entire MNCs analyzed without targeting, whereas our stimulation by PHA/IL-2 focuses on CD3⁺ lymphocytes, whose rate is highly heterogeneous (9.7% to 71%) in our cohort. As for the patient out of INH specification, this observation is probably due to hematocrit greater than 2.0%, known to reduce irradiation efficacy. However, it turns out that, both methods answer our laboratory specifications criteria. The choice of assay will depended on laboratory objectives, ECP pathology treated and further progress in understanding POST-ECP cells mechanisms of action in vivo.

Conflict of interest: No conflict of interest to disclose

[P104 Figure]

$$$$

P105

Fluorescence Based Reporter Cells Rapidly Identify and Distinguish Functional Chimeric Antigen Receptors (CARs) based on NF-κB and NFAT signaling

Julian Rydzek ¹ , Sabrina Jutz ² , Judith Leitner ² , Lars Wallstabe ¹ , Thomas Nerreter ¹ , Johannes Huppa ³ , Peter Steinberger ² , Hermann Einsele ¹ , Michael Hudecek ¹

¹ Medizinische Klinik II, Universitätsklinikum, Würzburg, Germany; ² Institut für Immunologie, Medizinische Universität, Wien, Austria; ³ Institut für Hygiene und Angewandte Immunologie, Medizinische Universität, Wien, Austria

Background: Immunotherapy with chimeric antigen receptor (CAR)-modified T-cells is under intense pre-clinical and clinical investigation involving a rapidly increasing spectrum of new CAR designs and target antigens. Here, we present a novel dual parameter reporter (DPR) cell line that enables rapid high-throughput testing and functional characterization of CARs.

Methods: DPRs were derived from the Jurkat T-cell lymphoma line which was modified with NF-κB- and NFAT-inducible CFP and GFP reporter genes. CARs were expressed by lentiviral gene-transfer and CAR positive DPRs enriched to >90% purity using an EGFRt selection marker. CAR DPRs were co-cultured with antigen positive stimulator cells and reporter gene-activation was analyzed by flow cytometry. Stimulation using T-cell stimulator (TCS) cells with membrane-bound anti-CD3 scFv that triggers CD3 on DPRs was used as reference.

Results: We transduced DPRs with a prototypic ROR1-specific CAR with 4-1BB co-stimulation and detected highest NF-κB and NFAT reporter gene-induction after 24 to 48 hours of co-culture with ROR1 positive stimulator cells. Activation of NF-κB and NFAT by CAR engagement was also confirmed in primary T-cells using Western Blot. Next, we expressed a panel of ROR1-CARs with variable spacer length (derived from IgG4-Fc) in DPRs based on our previous work in primary T-cells showing the importance of a particular spacer design for optimal anti-tumor function. DPRs rapidly identified the optimal construct from this panel, shortening the time required for analysis from ~3 weeks to ~6 days. We were also interested in determining the NF-κB and NFAT activation signature in DPRs modified with a 1st (no co-stimulation), 2nd (CD28 or 4-1BB) and 3rd generation (CD28 plus 4-1BB) ROR1-CAR. We detected similar levels of NFAT induction by each of these receptors (Figure 1). However, the ROR1-CAR construct comprising a 4-1BB rather than CD28 co-stimulatory moiety showed significantly stronger induction of NF-κB (n=3, p< 0.05). Notably, the 2nd generation ROR1-CAR with CD28 did not induce higher levels of NF-κB than the 1st generation construct and the 3rd generation construct did not induce higher levels of NF-κB than the 2nd generation ROR1-CAR with 4-1BB. This indicated that the CD28 co-stimulatory domain in the CAR format did not contribute to NF-κB signaling. To corroborate these data, we confirmed that stimulation of DPRs through TCR and endogenous CD28 (independent from the CAR) readily induced a NF-κB signal.

Conclusions: Our data demonstrate the potential to use DPRs to rapidly evaluate panels of CAR constructs based on NFAT and NF-κB signaling. Our data suggest that the conventional position of the CD28 co-stimulatory moiety in the CAR construct, located between transmembrane domain and CD3 zeta, is suboptimal for fully exploiting its co-stimulatory capacity. This finding has triggered the design of novel CAR formats that are currently being evaluated in our laboratory. Further, we are currently integrating DPRs into the analysis of CAR libraries with different specificity, affinity and receptor designs and are confident this novel DPR platform will accelerate the pre-clinical development and selection of CARs with optimal anti-tumor function for clinical translation.

Conflict of interest: None

[P105 Figure]

$$$$

P106

Immune reconstitution and graft correlations of innate effector cells after hematopoietic stem cell transplantation

Lia Minculescu ¹ , Hanne Vibeke Marquart ¹ , Lars Peter Ryder ¹ , Anne Fischer-Nielsen ¹ , Eva Kannik Haastrup ¹ , Ida Schjødt ² , Søren Lykke Petersen ² , Lone Smidstrup Friis ² , Brian Thomas Kornblit ² , Niels Smedegaard Andersen ² , Henrik Sengeløv ²

¹ Copenhagen University Hospital, Rigshospitalet, Department of Clinical Immunology, Copenhagen, Denmark; ² Copenhagen University Hospital, Rigshospitalet, Department of Hematology, Copenhagen, Denmark

Background: Immune reconstitution of γδ T cells and NK cells are recognized as important for clinical outcomes after hematopoietic stem cell transplantation (HSCT) and could be dependent on contents and subset distribution of these cells in the stem cell graft.

Methods: From 2015 to 2017 108 consecutive patients with hematological malignancies were transplanted at the Stem Cell Transplant Unit, Rigshospitalet, Denmark. The conditioning was myeloablative in 51 patients and nonmyeloablative in 57 patients. Donors were HLA-identical sibling (24), haploidentical (2) and 9/10 or 10/10 matched unrelated (82). T cell replete grafts were derived from bone marrow (17) and peripheral blood (91). Fresh graft samples and patient blood samples on day 28, 56, 91 and 180 after HSCT were analyzed for absolute concentrations of T- and NK cells (BD TruCount^TM beads) and subset phenotypes by multicolor flow cytometry panels performed on BD FACSCanto™ II. Subset markers included TCRαβ, TCR γδ, CD3, CD4, CD8, Vd1, Vd2, CD16, CD56, HLA-DR, CD45RO, CD45RA, CD197, CD314 and CD337.

Results: Median concentrations of CD3 T cells reached normal range by day 180 after transplantation while NK cells remained within normal range at all time points.

The fraction of γδ T cells out of CD3 T cells was stable, while the distribution of Vδ1 and Vδ2 showed increasing fractions of Vδ1 from day 28 towards day 180. Within the NK cell compartment, the CD16bright fraction remained stable, while the CD56bright fraction decreased along with a parallel increase in the CD16+CD56+ fraction towards day 180. Immune reconstitution of γδ T cells, Vδ1 and V2δ fractions correlated highly significantly with the graft contents of the corresponding cell subsets, p< 0.001 at all analyzed time points. The correlation to graft contents within the NK cell compartment showed greater variations; CD16bright cells only correlated on day 28 (p=0.05), CD56bright cells only on day 28 (p< 0.01) and day 56 (p=0.01), and CD16+CD56+ cells only significantly on day 91 (p=0.03)

Conclusions: Immune reconstitution showed compelling correlation to graft contents; the correlation was stronger for γδ T cells than for NK cells. The results are relevant for the increasing focus on innate effector cells and graft manipulation in HSCT.

Conflict of interest: All authors: nothing to disclose.

Cell concentrations, 10x7/L, median(range)	n	Day 28	n	Day 56	n	Day 91	n	Day 180
CD3 (normal range 69-270)	108	41.5(0-200)	107	43.0(0-340)	103	44.0(0-980)	88	80.5(5.6-410)
NK (normal range 8.0-56)	108	28.0(2.0-92.0)	107	19.0(0-160)	103	16.0(0-90.0)	88	23.0(2.0-78.0)
Relative concentrations percent, median (range)
TCRgd/CD3	106	4.2(0-50.4)	106	4.3(0-43.6)	101	4.0(0.1-47.8)	88	3.7(0.4-45.3)
Vd1/TCRgd	105	10.0(0-73.8)	105	11.6(0-82.4)	101	12.8(0-89.3)	87	26.3(0-95.8)
Vd2/TCRgd	105	87.6(0-100)	105	85.7(5.1-100)	101	82.8(3.7-100)	87	67.5(2.4-98.9)
CD16++/NK cells	108	10.1(0.5-72.2)	106	14.2(2.1-51.8)	102	14.7(0.7-59.1)	87	12.3(1.3-84.8)
CD16+56+/NK cells	108	58.3(6.7-88.7)	106	67.8(28.9-90.4)	102	67.8(32.4-90.7)	87	75.8(13.6-96.7)
CD56++/NK cells	108	23.4(3.8-71.7)	106	14.8(3.4-56.4)	102	12.5(1.2-59.8)	87	8.7(1.5-51.8)

[ [P106 Table] Table 1.]

[P106 Figure]

$$$$

P107

Immuno-T, a motion comic educating patients and caregivers on how immunotherapy works

Chiel Gharakhani Nia ¹ , Stijn Calis ² , Yves Bertrand ² , Lieve Achten ³ , Jasmien Soetens ⁴ , Barbara Casteels ⁴ , Daisy Flamez ⁵ , Bruno Bollaert ⁶ , Tessa Kerre ¹

¹ Ghent University Hospital, Hematology, Gent, Belgium; ² Holofarm, Gent, Belgium; ³ Telma, Gent, Belgium; ⁴ Ghent University Hospital, Cancer Centre, Gent, Belgium; ⁵ Ghent University, Biomarked IOF Consortium and CRIG, Gent, Belgium; ⁶ WahWah vzw, Gent, Belgium

Background: The introduction of immunotherapy has been a major breakthrough in the treatment of cancer. In the general public and the average patient populations, the overall knowledge of the immune system and its role in cancer is very limited. Moreover, explaining the immune system and how immunotherapy works to patients and their caregivers, is challenging because of the extreme complexity.

Serious games and virtual aids have only been scarcely used for cancer patients or patients in general. However, the limited amount of literature evinces the benefit patients and caregivers experience when these services are employed.

We now have developed Immuno-T: a motion comic, explaining the genesis of cancer and the working mechanisms of three types of immunotherapy. A motion comic combines a digital comic with animation, sound effects, voice-over and/or a musical score.

Methods: The scenario, story board and basic character design were developed by a hematologist, executed by game developers, and further refined by small focus groups consisting of hematologists-oncologists, psychologists, study coordinators, medical students, nurses and communication managers. The result was a first usable version. One hundred testers were asked to give their opinion in a anonimous questionnaire. The first version was eventually evaluated by 40 people of different backgrounds (hematologists-oncologists (both pediatric and adult), psychologists, study coordinators, medical students, nurses, communication managers, game developers, laymen and a small number of patients). Taking all their remarks and observations into account, Immuno-T was further adapted and improved into the final version.

Results: Immuno-T explains three types of immunotherapy. In an engaging, interactive and easy-to-understand way, the complex theory of checkpoint inhibitors, CAR T-cells and BiTEs is illustrated. The majority of the evaluators (94.9%) considered the motion comic a good tool to explain immunotherapy to patients, in addition to the physician consultation. In the first version, the level of understanding differed depending on the type of immunotherapy that was explained: BiTes were the least clear to the evaluators, prompting us to adapt this part of the motion comic. Almost 40% of the evaluators felt hopeful after watching the motion comic, 30% experienced other positive emotions (happy, well-informed, strengthened).

Conclusions: Although motion comics carry important advantages (a clear and graphic presentation of complex information and a more engaging way of delivering information), they have only rarely been used in healthcare settings. Immuno-T is ready to be rolled out in the first real hospital setting. We are currently preparing a clinical trial to evaluate the effectiveness of information transmission using Immuno-T: does our motion book lead to patients´ better understanding of the given treatment? Increasing patients' knowledge of disease and treatment has a proven impact on hope, strength and empowerment, and on adherence to the treatment.

Meanwhile, Immuno-T also keeps evolving: the motion book will be enriched with an augmented reality function to improve the understanding and deepen the user experience. Furthermore multiple translations are forthcoming, to facilitate the international career of our motion comic. Lastly, Immuno-T will also be evaluated in the context of education at different levels.

Clinical Trial Registry: Non applicable

Conflict of interest: The development of the motion comic was supported partly by a number of pharmaceutical companies having some forms of immunotherapy in their protfolio (Celgene, Celyad, Amgen, BMS, Roche, Novartis, MSD), but the major part was supported by the Innovation Fund of the Ghent University Hospital. The supporting companies had no influence on the development of the motion comic.

The individual authors have nothing to disclose.

1771

Improved Efficacy of Granulocyte Collections Obviating the Need for Addition of Hydroxyethyl Starch by Use of the Terumo BCT Spectra Optia Device

Dimitra Oikonomopoulou ¹ , Ioannis Tsonis ¹ , Vassilis Anyfantis ² , Chara Giatra ¹ , Maria Vardaka ¹ , Zoi Poulopoulou ¹ , Eirini Bika ¹ , Zois Mellios ¹ , Ioannis Konstantellos ¹ , Tatiana Tzenou ¹ , Stavros Gigantes ¹ , Ioannis Baltadakis ¹ , Nicholas Harhalakis ¹ , Dimitrios Karakasis ¹

¹ Evaggelismos General Hospital, Department of Haematology and Bone Marrow Transplantation Unit, Athens, Greece; ² Rontis Hellas S.A., Athens, Greece

Background: Granulocyte transfusions (GTs) represent a potentially life-saving intervention in patients with severe neutropenic infections and anticipated bone marrow recovery as it applies in the setting of hematopoietic stem cell transplantation (HSCT). However, their use has been limited by lack of urgent donor availability as well as by aspects pertaining to the collection and administration of adequate granulocyte doses in a safe manner. We explored the feasibility and safety of GTs in intensive chemotherapy and/or allogeneic HSCT recipients by the use of the Spectra Optia device. Our aim was to optimize collection efficacy, while avoiding the addition of Hydroxyethyl starch (HES) for sedimentation of red blood cells to prevent anaphylactoid reactions in recipients.

Methods: A total of 28 granulocyte collections (GCs) were obtained for 5 patients with grade IV neutropenia and life-threatening infections in the context of induction chemotherapy for acute leukemia (n=1) or allogeneic HSCT (n=4). Granulocyte donors (4 females/24 males) were either volunteer donors or friends or relatives of the patient, selected on the basis of ABO compatibility. For granulocyte mobilization, granulocyte-colony stimulating factor (G-CSF, 6-8 µg/kg, subcutaneously) plus dexamethasone (8 mg, intramuscularly) were administered 10-12 hours prior to collection. All collections were performed with the Spectra Optia device, v11.2 PMN collection protocol, without using HES. Peripheral veins were used as vascular access in all procedures. Calcium supplementation was given orally to all donors. The target dose was 3.5-5.0×10^10 granulocytes per transfusion. Granulocyte collection products were irradiated prior to infusion to patients.

Results: The median granulocyte yield per collection procedure was 3.78 (range, 1.55-8.42) ×10^10, with a median volume of 455 (range, 165-650) ml and a median hematocrit of 19.55% (range, 7.5-33.5%). The median blood volume processed was 6000 (range, 2528-8418) ml or 1.17 (range 0.39-1.62) total blood volumes (TBV). There was a median platelet count reduction in donors of 29% (range 11-51%) from baseline values, which was mainly dependent on blood volume processed. No remarkable side effects were recorded in donors in relation to granulocyte mobilization and collection. GTs were well tolerated by recipients without febrile or anaphylactoid reactions. A clinical benefit, defined as control of neutropenic infection until hematopoietic recovery or engraftment, was achieved by granulocyte transfusions in 3 out of 5 patients.

Conclusions: Granulocyte collection with the fully automated Spectra Optia device is a safe procedure for donors, and results in transfusion of adequately high doses of granulocytes. Comparable yields of granulocytes can be achieved with or without the addition of HES suggesting enhanced safety for recipients. Based on the above data, a broader application of GTs may be warranted especially in the context of HSCT.

Conflict of interest: None of the authors has anything to disclose.

P109

In vivo patterns of dendritic cell and monocyte populations in paediatric patients with acute GVHD treated with ECP

Aisling M Flinn, Tom J Altmann, Venetia Bigley, Xiao Nong Wang, Andrew R Gennery

Newcastle University, Institute of Cellular Medicine, Newcastle upon Tyne, United Kingdom

Background: Extracorporeal photopheresis (ECP) is a second line immune-modulating therapy used in the treatment of acute GVHD. Previous studies suggest a role of dendritic cells (DC) in the underlying tolerance-inducing mechanisms of action, and monocyte differentiation into DCs during the ECP process, although little is known about the true in vivo effects on DC and monocyte populations.

Methods: Six paediatric patients received ECP for the treatment of corticosteroid-refractory/dependent acute GVHD. Four patients have completed therapy, one was withdrawn after 16 cycles of ECP and one remains on ECP (week 72). Whole blood samples were analysed before ECP cycles at regular intervals and absolute numbers of myeloid DC (mDC, CD3^-CD19^-CD20^-HLA-DR⁺CD14^-CD16^-CD11c⁺ cells), plasmacytoid DC (pDC, CD3^-CD19^-CD20^-HLA-DR⁺CD14^-CD16^-CD11c^-CD123⁺ cells), and monocytes populations (including classical monocytes, CD3^-CD19^-CD20^-HLA-DR⁺CD14⁺CD16^- cells), were enumerated in TruCOUNT® (BD Biosciences) tubes using flow cytometry. Results were compared to two control groups at 4, 8 and 12 months post HSCT; paediatric patients post HSCT who did not develop acute GVHD (group 1, n=6) and paediatric patients with acute GVHD who did not receive ECP (group 2, n=6).

Results: Among the patients who have completed ECP therapy, an overall increase in pDC numbers was observed, with a reduction in the mDC/pDC ratio. In contrast, the patient who remains on ECP exhibits consistently low pDCs, with an increasing mDC/pDC ratio. There was an overall downward trend of the classical monocyte population seen in all ECP patients. The median mDC/pDC ratio was higher in the ECP group at 4 months but higher in the acute GVHD control group 2 at 8 months (Figure A). A steady rise was seen in mDC and pDC numbers in both the no acute GVHD control group 1 and ECP groups, while this was more static in the acute GVHD control group 2 (Figure A).

Conclusions: Increasing the pDC population, and decreasing the mDC/pDC ratio, overall promoting a favourable pDC environment may be an underlying mechanism of action behind ECP. Plasmacytoid DCs have been reported to have tolerance-inducing effects, including stimulation of regulatory T-lymphocytes and causing T-lymphocyte anergy. Decreasing classical monocyte populations may be due to DC differentiation and/or deviation from an inflammatory environment. Further research examining the phenotype of the DC populations and the concurrent impact on regulatory T-lymphocytes is needed.

Conflict of interest: None

[P109 Figure]

$$$$

P110

Individual differences of multipotent mesenchymal stromal cells detected by interaction with lymphocytes

Nikolay Kapranov, Yulia Davydova, Irina Galtzeva, Nataliya Petinati, Natalia Sats, Nina Drize, Larisa Kuzmina, Elena Parovichnikova, Valeriy Savchenko

National Research Center for Hematology, Moscow, Russian Federation

Background: Multipotent mesenchymal stromal cells (MSCs) are used for the treatment and prevention of acute graft versus host disease (GVHD). Not all MSCs samples are effective. The study of MSCs interaction with lymphocytes in vitro revealed 2 types of changes in the subpopulations of lymphocytes. The expression of HLA-DR on lymphocytes was significantly increased in the co-cultivation with a part of the MSCs samples (group A), whereas co-cultivation of the same lymphocytes with other MSCs samples did not (group B). The aim of the study was to investigate the differences in the two identified groups of MSCs.

Methods: MSCs were derived from bone marrow (BM) of 26 donors (15 male and 11 female aged 11 to 62 years, median 27 years) and on 2^nd-3^rd passages were co-cultured with non-activated and activated with 5 mg/ml of PHA (PHA-lymphocytes) allogeneic lymphocytes in a ratio of about 1:10 for 4 days. Subpopulations of lymphocytes and MSCs surface markers were analyzed over time by flow cytometry. Determination of gene expression levels (REL) was performed by real time PCR.

Results: There were no differences between BM donors for groups A and B by sex, age and concentration of CFUs in their BM. Significant differences in the characteristics of MSCs were found in the total cell production for 4 passages, which was in group A 2.5 times lower than in group B (8.1±1.5 vs. 20.3±3.2). After a day of co-cultivation with lymphocytes, the REL of IDO1 in group A was 6.5 times higher than group B (10058±2695 vs. 1554±716), and REL of PPARg was 3.4 times lower (0.35±0.06 vs. 1.18±0.32). After 4 days of co-cultivation in group A, the REL of ICAM1 was 3.2 times higher (11.3±2.8 vs. 3.5±0.9). Thus, the immunomodulatory ability of MSCs from group A should be significantly higher than from group B. This is also indicated by differences in the subpopulations of lymphocytes co-cultivated with MSCs from these groups.

The following significant differences in the subpopulations of lymphocytes were revealed. After 4 days of co-cultivation with MSCs of group A, the fraction of CD4+HLA-DR+ was higher 2.8 times (20.4±0.4 vs. 7.4±1.0); CD8+HLA-DR+ 3.5 times (35.5±1.3 vs. 10.0±2.9); CD4+EM (effector memory cells) 1.2 times (20.2±0.9 vs. 16.6±1.6); CD4+TM ("transitional" memory cells) 3.5 times (1.2±0.01 vs. 0.3±0.2) for non-activated lymphocytes, for activated lymphocytes proportion of CD4+HLA-DR+ was higher 2.4 times (72.1±3.6 vs. 29.7±9.3); CD8+HLA-DR+ 2.2 times (76.2±4.6 vs. 35.0±13.1), CD4+CM (central memory cells) 1.6 times (53.4±2.7 vs. 32.7±5.5); CD8+CM 2.4 times (35.8±2.7 vs. 15.1±6.0); CD4+TM 6.5 times (1.8±0.2 vs. 0.2 ± 0.1); CD8+EM 1.6 times (16.3±2.3 vs. 10.1±1.4); CD4+PD1+ 2.5 times (44.2±5.6 vs. 17.6±3.2); CD8+PD1+ 2.2 times (23.4±3.5 vs. 10.6±2.6) and CD8+TE (terminal effector cells) was 2.5 times lower (6.2±0.7 vs. 14.1±2.9) compared with lymphocytes co-cultured with MSCs from group B.

Conclusions: The revealed individual differences in MSCs can explain why not all MSC samples are effective for the treatment of autoimmune diseases, acute GVHD and others.

Conflict of interest: The study was financially supported through a grant from the Russian Science Foundation, Project № 16-15-00102. The individual authors have nothing to disclose.

P111

Induction of donor-specific tolerance in living donor kidney transplant recipients through clinical MIC cell infusion - a phase I study (TOL-1)

Christian Morath ¹ , Anita Schmitt ² , Christian Kleist ³ , Volker Daniel ⁴ , Gerhard Opelz ⁴ , Caner Süsal ⁴ , Florian Kälble ¹ , Claudia Sommerer ¹ , Lei Wang ² , Angela Hückelhoven ² , Arianeb Mehrabi ⁵ , Uta Merle ⁶ , Luiza Pego da Silva ¹ , Carsten Müller-Tidow ² , Martin Zeier ¹ , Matthias Schaier ¹ , Michael Schmitt ² , Peter Terness ⁴

¹ University Clinic Heidelberg, Division of Nephrology, Heidelberg, Germany; ² University Clinic Heidelberg, Department of Internal Medicine V, Heidelberg, Germany; ³ University Clinic Heidelberg, Department of Nuclear Medicine, Heidelberg, Germany; ⁴ University of Heidelberg, Institute of Immunology, Heidelberg, Germany; ⁵ University Clinic Heidelberg, Allgemein-, Viszeral- und Transplantationschirurgie, Heidelberg, Germany; ⁶ University Clinic Heidelberg, Department of Internal Medicine IV, Heidelberg, Germany

Background: Major limitations of long-term allograft survival are life-threatening side-effects by immunosuppressive medication and chronic allograft injury. Therefore there is a fervent need for the induction of a donor-specific tolerance without broad unspecific immunosuppression. To this aim a clinical phase I study with so called MIC cells was performed, i.e. donor-derived monocytes that gain immunosuppressive properties after incubation with the proliferation inhibitor mitomycin C.

Methods: For 10 kidney transplant recipients, MIC cells were manufactured under Good Manufacturing Practice (GMP) conditions out of leukapheresis products from the living donors. The amount of MIC cells administered to the patient was escalated from 1.5x10^6 MIC cells per kg body weight at day -2 (N=3, group A), to 1.5x10^8 MIC cells per kg body weight at day -2 (N=3, group B) or on day -7 (N=4, group C) before living donor kidney transplantation. After kidney transplantation, patients received a standard triple drug immunosuppressive therapy. The frequency of adverse events (AE) was measured as primary outcome at day 30.

Results: Clinically, all kidney transplant recipients had a median serum creatinine of 1.4 mg/dL at day 30 and remained stable with a median creatinine of 1.48 mg/dL at day 180 without significant proteinuria (median 10 g/mol creatinine at day 180) and without rejection episode. The patients experienced in total 70 AEs including four severe AEs not related to the MIC cell infusion. Besides two infectious complications no other AEs like positive cross match results, de novo donor-specific antibodies or rejection episodes were recorded.

In patients of group C, a reduction of immunosuppressive therapy was effective in the observational phase with low-dose cyclosporine A and low-dose enteric-coated mycophenolate sodium.

Major immunological changes were observed in patients of group C with an increase of CD19+ B cells up to a median of 300/µL until day 30 and a decrease thereafter to a median of 35/µL at day 180. Furthermore CD19+CD24highCD38high transitional Bregs increased from a median of 2% at day 30 to a median of 20% of the total CD19+ B cell pool on day 180. And the plasma IL-10/TNFalpha ratio increased from a median of 0.05 before cell therapy to a median of 0.11 at day 180. In an in vitro mixed lymphocyte culture assay lymphocytes of these patients showed no or only minimal reactivity against irradiated donor lymphocytes while reactivity against 3rd party lymphocytes was preserved.

Furthermore the analysis of MIC cell products in in vitro assays demonstrated the capability of MIC cells to induce tolerogenic dendritic cells by down-regulating costimulatory molecules CD80 and CD86 and the maturation molecule CD83 while up-regulating the immunosuppressive molecule CD103.

Conclusions: All kidney transplant recipients receiving the MIC cell product showed a stable renal function without any allograft injury nor rejection episodes even under reduction of conventional therapy with immunosuppressive drugs. Therefore MIC cells represent an innovative and extremely promising option for donor-specific immunosuppression in living donor kidney transplantation and beyond.

Clinical Trial Registry: EudraCT number 2014-002086-30

Conflict of interest: C. Morath, A. Schmitt, C. Kleist, G. Opelz, M. Zeier, M. Schaier, M. Schmitt, and P. Terness are founders of TolerogenixX GmbH. L. Wang is an employee of the company.

P112

Infusion of Donor Memory T-Cells as a safe strategy to improve immune reconstitution after Haploidentical Stem Cell Transplant

Andreina Figueira ¹ , Alba Perez ¹ , Blanca Rosich ¹ , Yasmina Mozo ¹ , David Bueno ¹ , Sonsoles San Roman ¹ , Berta Gonzalez ¹ , Pedro Rubio ¹ , Diego Plaza ¹ , Ana Sastre ¹ , Raquel De Paz ² , Antonio Perez Martinez ¹

¹ Hospital Universitario La Paz, Pediatric Hemato-Oncology and Stem Cell Transplantation Unit, Madrid, Spain; ² Hospital Universitario La Paz, Hematology Department, Madrid, Spain

Background: The use of T-cell depleted grafts in haploidentical stem cell transplantation (HSCT) has been proven efficient in preventing graft versus host disease (GvHD), although it is associated with a delay in early T-cell recovery which increases the risk of viral infections, leukemia relapse or graft rejection. Conventional donor lymphocyte infusion (DLI) after HSCT transplantation has proven to be an effective strategy; nevertheless, it is conditioned because of a high prevalence of GvHD even with low dose of T-cells. The infusion of selected lymphocyte subpopulations with low aloreactivity is emerging as an effective strategy to rectify this issue. The depletion of CD45RA+ naïve lymphocytes, while preserving CD45RO+ memory T-cells, could provide a safe source of functional lymphocytes with anti-infection, anti-leukemic and anti-rejection properties, and a lower rate of adverse effects.

Methods: We present data of CD45RO+ memory T-cells DLI (mDLI) performed after HSCT in cases of mixed chimerism, persistent lymphopenia, presence of viral/opportunistic infections or as a strategy to accelerate immune reconstitution.

Results: Fifteen patients with diagnosis of Acute Lymphoblastic Leukemia (n=6), Acute Myeloblastic Leukemia (n=3), Myelodisplastic Syndrome (n=1), Severe Aplastic Anemia (n=3), Sideroblastic Anemia (n=1) and Chronic Granulomatous Disease (n=1), received one or more mDLI after HSCT. Of the fifteen patients, twelve received a HSCT with a graft consisting of CD34+ and CD45RA- cells, and three with a TCRab+ cells depleted graft.

A total of forty-three mDLI were infused. The median dose of CD45RO+ memory T-cells infused was 5.00x10⁷/kg (range:4.8x10⁴-4.25x10⁸/Kg), with a median dose of CD45RA+ naïve T-cells of 3.90x10²/kg (range:0-1.3x10⁴/Kg).

The mDLI were infused at a median of seventy-seven days after HSCT (range:14-407 days), with a median interval between mDLI of thirty-four days (range:4-159 days). 60% of patients received 2-3 mDLI (n=9), while four patients (26.67%) were infused in more than 3 occasions and two patients (13.33%) were infused only once.

Twenty-one mDLI (49%) were administered because of lymphopenia, fourteen of them (33%) in patients with concomitant viral/opportunistic infections (CMV, EBV, BKV, Adenovirus, VZV, toxoplasmosis). Mixed chimerism/graft failure was the motive of 37% of the mDLI (n=16) and six (14%) were administered as an attempt to accelerate immune reconstitution.

All infusions were well tolerated without any side effects during the infusion or appearance/worsening of GvHD.

A progressive increase in T-cell counts was observed following six mDLI (28.57%), although it was a transitory response (3-8 weeks) in five cases. Viral/opportunistic infections were controlled in five cases (35.71%), requiring a median of 2 mDLI to achieve this response. As for the mDLI administered in cases of mixed chimerism/graft failure, none were effective in reverting this situation.

Conclusions: Our preliminary data suggests that mDLI, is a safe adoptive immunotherapy strategy even with high dose of T-cells without infusion side effects or GvHD complications. Some efficacy has been observed in patients with lymphopenia and opportunistic infections, while there have been no positive results in patients with mixed chimerism/graft failure, up to date. However, to determine the real efficacy of this strategy, prospective studies are required.

Conflict of interest: None of the authors has anything to disclose.

P113

Long lasting responses to virus-targeted T-cell therapy in oligometastatic EBV-related nasopharyngeal carcinoma

Simona Secondino ¹ , Antonella Gurrado ² , Sabrina Basso ² , Gianpiero Rizzo ¹ , Ilaria Imarisio ¹ , Ilaria Pisani ² , Jessica Bagnarino ² , Marco Zecca ³ , Paolo Pedrazzoli ¹ , Patrizia Comoli ²

¹ Fondazione IRCCS Policlinico San Matteo, Oncology Unit, Pavia, Italy; ² Fondazione IRCCS Policlinico San Matteo, Pediatric Hematology/Oncology & Cell Factory, Pavia, Italy; ³ Fondazione IRCCS Policlinico San Matteo, Pediatric Hematology/Oncology, Pavia, Italy

Background: Nasopharyngeal carcinoma (NPC) is a Epstein-Barr virus (EBV)-related, highly chemo-radiosensitive malignancy. However, one-third of patients are considered to be incurable because of metastatic or recurrence disease. Published experiences reported 7-14% metastatic NPC patients surviving ≥2 years. Expression of antigenic viral proteins by malignant cells constitutes a good target for immunotherapeutic strategies, and we have previously showed disease control by autologous EBV-targeted cell therapy in refractory/relapsed patients failing conventional treatment.

Methods: We describe the results of a T-cell therapy program for patients with oligometastatic relapsed NPC patients, treated at completion of second-line chemotherapy. The patients received 2 administrations of EBV-specific cytotoxic T lymphocytes (CTL) at a total cell dose/infusion of 1.5-3 x 10⁸, repeatable in case of objective response.

Results: Starting from 12/2011, we have treated 12 patients with refractory/metastatic disease. Six patients had more than 2 organ involved and >3 metastatic lesions, while 6 patients had oligometastatic disease, with less than 3 lesions. The best response after second-line treatment was progressive disease (PD) in 2 patients, stable disease (SD) in 3 patients, and partial (PR) or complete (CR) response in 4 and 3 patients, respectively. After CTL therapy, no severe adverse events were observed. The 5 patients treated with progressive or stable disease and more than 9 metastatic lesions showed NPC progression and died of their disease. The patient with >3 matastases who reached PR after second-line chemotherapy, stabilized disease for 2 years after T-cell therapy. Upon disease progression, she was treated with third-line capecitabine-based chemotherapy and reached long-lasting CR. She is alive without evidence of disease at 85 months from diagnosis. Of the 3 patients who reached CR after second-line chemotherapy and received adjuvant CTL treatment, 2 persist in CR at 5- and 3-year follow-up, while one patient had disease relapse, and reached long-lasting CR after treatment with 3^rd-line gemcitabine-based chemotherapy. Of the 3 patients with oligometastatic disease who were treated with EBV-CTL while in PR after second-line chemotherapy, one reached long-lasting CR, while the other 2 had disease progression after initial stabilization of PR. They received conventional treatments (one patient radiotherapy on site of disease and the other 3^rd line chemotherapy) followed by T cell therapy, achieving a CR. At a median follow-up of 58 months, 7/12 patients are alive with no evidence of disease.

Conclusions: EBV-specific CTL therapy is safe and associated with clinical benefit in patients with refractory or metastatic NPC. Combination of second-line chemotherapy and CTL therapy seems to yield the best results, which compare favourably to the outcomes described in the literature.

Clinical Trial Registry: N/A

Conflict of interest: none to declare

P114

Mixed chimerism reversal after cord blood donor lymphocyte infusion using a 20% fraction of the original CB unit

Albert Esquirol Sanfeliu ¹ , Rodrigo Martino Bufarull ¹ , Laura Soria Guerrero ² , Guillermo Ortí ³ , Christelle Ferrà Coll ⁴ , Isabel Sanchez-Ortega Sanchez ⁵ , Cristina Diaz de Heredia ⁶ , Isabel Badell Serra ⁷ , Laura Medina Marrero ⁸ , Nerea Castillo ⁸ , Ramón Gimeno Martínez ² , Sergio Querol Giner ⁸

¹ Hospital de la Santa Creu i Sant Pau, Hematology Service, Barcelona, Spain; ² Hospital del Mar & IMIM (Hospital del Mar Medical Research Institute), Immunologia, Barcelona, Spain; ³ Hospital Vall d'Hebron, Hematology Service, Barcelona, Spain; ⁴ Institut Català d'Oncologia Trias i Pujol, Badalona, Spain; ⁵ Institut Català d'Oncologia Duran i Reynals, L'Hospitalet de Llobregat, Spain; ⁶ Hospital Materno-Infantil Vall d'Hebron, Barcelona, Spain; ⁷ Hospital de la Santa Creu i Sant Pau, Pediatrics Service, Barcelona, Spain; ⁸ Banc de Sang i Teixits, Cell Therapy Services, Barcelona, Spain

Background: Donor lymphocyte infusion (DLI) has not been tested in CB transplantation. We have developed a phase I-II clinical trial (NCT02328885). to assess the ability of DLI from a 20% spared fraction after engraftment of the 80% original product. This clinical trial is under evaluation. Here, we present a case report to show a potential effect of the CB lymphocytes in reverting mixed chimera.

Methods: Clinical case: A 46 years old male was diagnosed with ALL and chemotherapy treatment was started using a high risk ALL scheme (PETHEMA group). The most critical complication during the treatment was a central vein thrombosis as a consequence of L-asparaginase coagulopathies. The evaluation at the end of chemotherapy was a first CR with a negative minimal residual disease (MRD). He received umbilical CB transplant due to lack of a related and unrelated suitable donor. Patient fulfilled criteria to enter this clinical trial.

Conditioning regiment consisted of thiotepa (10 mg/kg), fludarabine (150 mg/m2) and busulphan (12 mg/kg, oral administration). GvHD prophylaxis was late administration of ATG (6 mg/kg), cyclosporine and short pulses of prednisone thereafter. The first part (80%) of umbilical CB containing 2.3E7 TNC and 1.1E5 CD34/kg was infused. Neutrophils (0.5x10⁶/L) and platelets (>20x10⁶/L) recovery were +21 and +47 days respectively. The main complication was a cytomegalovirus reactivation at day +55 that required a valganciclovir/ganciclovir treatment associated with corticoids due to a differential diagnosis with GvHD. Corticotherapy was fastly decreased and stopped before a second part infusion of CB.

Results: Bone marrow aspirates showed CR with negative MRD. However, donor lymphocytes population in the chimerism decreased (see figure) and the DLI was scheduled for day +106. The second part (20%) of umbilical cord blood was infused containing 7E5 CD3/kg. Already one month after the infusion a 4-fold increase in the number of T cells was apparent. These numbers amplified further in the next 3 months (from 70 T cells/microliter at the moment of infusion to 255 T cells/microliter 30 days later, or 933 T cells/microliter 3 months post-DLI). Concomitantly, an intense proliferative activity was noted. It was detected on CD4+ T cells and with a certain delay also on CD8+ cells. In this period a reversal in the CD4/CD8 ratio was observed, and even although most circulating T cells were CD8+ before the infusion (ratio equal to 0.1), a normal ratio was achieved in only 2 months At all time points analyzed, T cell memory phenotypes predominated.

Conclusions: In summary, we observed a vigorous increase in the presence of circulating T lymphocytes after DLI. The consequence of second CB infusion was a chimerism improvement until achieving total donor lymphocytes chimerism that was maintained. Also, and a new cutaneous and digestive grade II GvHD that responded to medium doses of corticosteroids was shown. In the last follow up, the patient was alive, in CR with full donor chimerism. Therefore, a potential role for this cell therapy intervention is proposed.

Clinical Trial Registry: NCT02328885

https://www.clinicaltrials.gov/ct2/results?term=NCT02328885&Search=Search

Conflict of interest: Authors declare no conflict of interest

[P114 Figure]

$$$$

P115

NK cell specific checkpoint inhibition improves anti-tumor efficacy of patient derived NK cells against autologous Multiple Myeloma

Sara Tognarelli¹, Ivana Von Metzler², Sebastian Wirsching¹, Benedikt Jakobs³, Andreas Mackensen³, Hubert Serve², Peter Bader¹, Evelyn Ullrich¹

¹Goethe University Frankfurt am Main, Division for Stem Cell Transplantation and Immunology, Department for Children and Adolescents, Frankfurt, Germany; ²Goethe University Frankfurt am Main, Department of Hematology and Oncology, Frankfurt, Germany; ³University Hospital Erlangen, Department of Medicine 5, Erlangen, Germany

Background: Natural Killer (NK) cells are innate lymphocytes with a strong anti-tumor ability. In tumor patients, such as multiple myeloma (MM) patients, an elevated number of NK cells after stem cell transplantation (SCT) has been reported to have benefits for the patients. We recently showed that, at an early time point after auto-SCT, CD56^bright CD16^-/+ NK cells represented the main NK cell subset. Remarkably, this NK cell population, considered to be immature, expressed maturation markers and exerted important cytotoxic functions, when challenged with K562 tumor cells, indicating a potential impact on antitumor immunity.

Methods: In this preclinical immune monitoring and scientific research study, NK cells have been isolated kits from peripheral blood and bone marrow of healthy donors and MM patients at defined time points using NK cell negative isolation. Also patient-derived tumor cells have been sorted from MM patient bone marrow samples at first diagnosis. Phenotypic and functional characterization has been performed by in vitro multi-color flow cytometry assays.

Results: With the aim of the NK cell use for adoptive cell therapy, we also addressed the cytotoxicity of patient-derived, cytokine-stimulated NK cells against MM cells at diagnosis, before and after auto-SCT. We detected changes in NK cell phenotype as well as cytotoxic function before and after expansion. Remarkably, after cytokine stimulation patient NK cells did not significantly differ from healthy donor NK cells, showed a highly activated phenotype and were able to significantly enhance the lysis of MM cells. In a smaller cohort of MM patients we were able to isolate autologous tumor cells, as well as bone marrow NK cells and we could show that our cytokine-based ex vivo activation protocol was able to significantly improve even the lysis of autologous tumor cells, suggesting a potential use of NK cells as adoptive therapy for MM patients. Moreover, in an effort to further enhance the lysis of MM cells, we also used an “NK check point inhibitor” antibody, targeting the inhibitory NKG2A receptor. Blocking the NKG2A-HLA-E pathway induced a significant increase in the NK cell mediated cytotoxicity against Multiple Myeloma.

Conclusions: In addition to autologous stem cell transplantation, adoptive NK cell therapeutic protocols that imply a cytokine-based ex vivo NK cell expansion procedure might be of high interest for the treatment of Multiple Myeloma patients. Furthermore, our functional assays revealed NKG2A blocking antibodies as promising “NK check point inhibitors”.

Conflict of interest: The authors have nothing to disclose.

P116

Abstract previously published

P117

Abstract previously published.

P118

Potent anti-leukemia activities of humanized chimeric antigen receptor modified T(CAR-T) cell therapy in Chinese patients with relapsed/refractory acute lymphoblast leukemia

Lei Xiao¹, Qing Zhang², Xiaofan Zhu³, Zhao Wu¹

¹Innovative Cellular Therapeutics Co.,Ltd., Shanghai, China; ²Guangdong No.2 Province People 's Hospital,, Guangzhou, China; ³Institute of Hematology& Blood Diseases Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Tianjin, China

Background: Although highlighted breakthrough of CD19 targeted chimeric antigen receptor (CAR) T cell on blood malignancies, high relapse rate is a big challenge for CART cells with single chain variable fragment (scFv) domains of murine origin. Our previously murine CART cells achieved high complete remissions (CRs) on patients with relapsed/refractory hematological malignances, as 83% complete remission rate in acute lymphoblastic leukemia and over 57% complete remission rate in non-hodgkin's lymphoma (NCT 02813837). However, unsatisfactory relapse rate was encountered as with most CAR with single chain variable fragment (scFv) domains of murine origin. Human anti-mouse antibody (HAMA) may be the important culprit. Therefore, we constructed CD19-directed CAR T cells with a humanized scFv domain to overcome the potential for anti-murine immune-mediated rejection.

Methods: We have constructed humanized anti-CD19 CAR with lentiviral vector encoding a CAR composed of murine or humanized anti-CD19 scFv, CD3z, and 4-1BB domains. To estimate the anti-tumor activity and safety in vitro, they were co-cultured with tumor cell lines with or without CD19 expression. To evaluate the safety and efficiency in vivo, human tumor cell transplanted immune deficient mice were divided into four groups including humanized CAR-T cells, murine derived CART cells, normal T cells and buffer. What's more, a preliminary clinical trial on four patients with refractory or relapsed acute lymphoblast leukemia were conducted afterwards. Four subjects with r/r B-ALL were treated with humanized CAR-T cells from December 16, 2016 to December 27, 2016. Both the murine and humanized CAR-T cells were infused with dose range between 0.45×10E6 CAR-T cells/kg and 10.51×10E6 CAR-T cells/kg. All subjects were monitored closely during the trial.

Results: CD19 positive tumor cells co-cultured with CAR-T cells showed that most of the tumor cells were killed by CAR-T cells while CD19 negtive tumor cells remained alive. For tumor cell transplantation experiment, the group of humanized CD19 CAR achieved a longer survival. Pre-clinical experiments showed the anti-malignancy ability and safety of CD19-directed CAR T cells with a humanized scFv domain. In the preliminary clinical trial, all four subjects with B-ALL treated by humanized CAR-T cells achieved CR as well as MRD negative, between 14 to 41 days after CART cell infusion. And none of them suffered from irreversible neurotoxicity.

Conclusions: This is the first multicentre report to our knowledge of successful treatment of r/r B-ALL with humanized CD19 CAR T cells in China. Even r/r B-ALL with high-burden leukemia patients also was effective and associated with a high remission rate after autologous CD19 CAR-T infusion. Though the follow-up is short, CD19 humanized CAR-T cell demonstrated its anti-malignances activity in subjects with ALL.

Clinical Trial Registry: NCT 02813837

Conflict of interest:

Q. Zhang and X. Zhu: nothing to disclose;

Z. Wu and L. Xiao: stockholders of Innovative Cellular Therapeutics Co., Ltd.

P119

Processing of Peripheral Blood Hematopoietic Stem Cell Using a GMP- Automated Closed System in a Clean Room Facility

Naziha Menasria ¹ , Mohammed Bakr ² , Rola Sami El Chahab ¹ , Wedad S Hamdi ¹ , Shondalyn Tan Daquila ¹ , Ajaeb Dakhilalla M H Al-Nabet ³

¹ Hamad Medical Corporation, GMP Cellular Therapy Lab, Department Laboratory Medecine & Pathology, Doha, Qatar; ² National Center for Cancer Care and Research (NCCCR), Doha, Qatar; ³ Hamad Medical Corporation, Department Laboratory Medecine and Pathology, Doha, Qatar

Background: Autologous hematopoietic progenitor cells collected through apheresis (HPC-A) must be cryopreserved and stored before transplantation. Cell processing before HPC-A cryopreservation includes volume reduction that reduces the amount of dimethyl sulfoxide used as cryoprotectant, and decrease the needs for storage space in liquid nitrogen tanks.

Methods: We retrospectively compared the Pericell protocol performed with an automated GMP closed system: the Sepax 2 device (Biosafe SA / GE Healthcare) and the common manual technique that uses refrigerated centrifugation at 850 × g prior to plasma removal and volume adjustment. We comparatively evaluated Total Nucleated Cells (TNC) and CD34+ cell recovery & viability, as well as technologist time needed to complete the procedure. A total of 19 procedures were carried out, to process apheresis products obtained from adult patients: eleven HPC-A units were processed with the historical technique and eight HPC-A units were processed with the Pericell protocol in an attempt to validate this technique.

Results: Automated processing required 15 minutes per procedure whereas manual processing required 20 to 25 minutes. The percentage volume reduction using the Sepax 2 Technology was 58.3% (range, 36.7-76.0 %; SD, SD, ± 13.8%), starting from 301 mL (range, 190- 430 mL; SD, ± 76.9 mL) to reach a final value of 126.5 mL (range, 60- 200 ml; SD, ± 53.2 mL). This is to be compared with the performances of the manual method :mean, 45.2% (range, 20.3-63.1 %; SD, ± 16.7%), downsizing the volume of the cell suspension from 260.5 mL (range, 141- 340 mL; SD, ± 72.3 mL) to 140.3 mL (range, 60- 240 ml; SD, ± 54.1 mL)),(P value = 0.088). There was no significant differences in the recovery of nucleated cells (TNC) for cell products processed with the Sepax 2Technology (mean 97.4%; range from 89.4%-110.9%; SD, ± 6.8%), CD34⁺ cells (mean 105.2%; range from 98.6%- 119%; SD, ± 7.0%) or cell viability (mean, 99.8%; range from 98.5-101; SD ± 0.83%) compared with the centrifuge method:100.3% (range from 92.3%- 112.5%; SD, ± 5.7%), mean; 107.2% (range, 91.3%-132.1%; SD, ± 13.4%) and mean :100.1% (range from 99-101.6; SD ± 0.86%),(P value > 0.3). Moreover, the automated processing demonstrated time effective results compared with a semi-automated centrifuge processing method. Although it is not significant, there was an increase in the volume reduction percentage using Sepax 2 Technology.

Conclusions: Our data demonstrates the effectiveness and suitability of an automated procedure for volume reduction of HPC-A collected by apheresis, prior to addition of the cryoprotectant and controlled-rate freezing. The Pericell protocol performed with Sepax 2 Technology shortens processing time, includes features that facilitate compliance with FACT-JACIE standards, can be operated in a GMP environment, and provides similar cell recovery and viability than the reference technique.

Conflict of interest: This study has been done without conflict of interest. The authors acknowledge the support of members of the Cell Processing Facility (Centre de Thérapie Cellulaire) at Institut Paoli-Calmettes and Inserm CBT-1409, Marseille, France for personnel training and support.

P120

Quantitative determination of virus-specific (CMV, EBV, Adeno) lymphocytes in donor peripheral blood and in CD45RA-depleted T-cell products before and after cryopreservation

Viktoriia Kiseleva, Svetlana Glushkova, Sergei Blagov, Elena Kurnikova, Yakov Muzalevskii, Aleksei Kazachenok, Yana Baizyanova, Rimma Khismatullina, Julia Panchenko, Natalia Khripkova, Elena Osipova, Michael Maschan

Dmitry Rogachev National Research Center of Pediatric Hematology, Oncology and Immunology, Moscow, Russian Federation

Background: Viral infections remain one of the leading cause of morbidity and mortality after hematopoietic stem cell transplantation (HSCT). Although pharmacotherapy may help to prevent or treat viral disease, these drugs are expensive, toxic, and often ineffective due to primary or secondary resistance. These deficiencies in conventional therapeutics have increased interest in adoptive cell therapy. Adoptive transfer of virus-specific cytotoxic T lymphocytes (VSTs) can rapidly reconstitute antiviral immunity without causing graft-versus-host disease. In this regard, evaluation of the numbers of VSTs in memory T-cell infusions is important to control viral infections before recovery of broad repertoire of T-cells after T-depleted haploidentical and unrelated transplantation.

Methods: T-cells were derived from G-CSF stimulated (n=74) or unstimulated (n=5) apheresis of healthy donors. Apheresis product was processed with single-step CD45RA depletion procedure on CliniMACS Plus or Prodigy instrument. CD45RA-depleted fraction was aliquoted and cryopreserved for further use. The frequency of CMV, EBV, ADV -specific T cells in donor peripheral blood (n=142), CD45RA-depleted T-cells products before (n=20) and after (n=79) cryopreservation were monitored by gamma-interferon enzyme-linked immunospot (ELISPOT) assays with CMVpp65, AdV5 Hexon, EBV consensus respectively. Excess monocytes from the cell suspension were depleted by adhesion on plastic dishes for 2 hours. The CD3 proportion and monocyte /CD3 (Mon/CD3) ratio in the cell suspension were determined by flow cytometry.

Results: In CD45RA-depleted T-cell products median Mon/CD3 was 0,5 and the efficiency of VSTs detection was dramatically lower in comparison with VSTs detection in peripheral blood (Median Mon/CD3=0,14). The median frequency per 300 000 MNC of CMV, EBV, ADV-specific T cells in the peripheral blood was 148; 54; 30 and in CD45RA-depleted T-cells products 4,0; 3,0; 0 respectively. Monocytes contained in CD45RA-depleted T-cells products can mediate suppression of interferon-gamma production by T cells after G-CSF stimulation of donors. After depletion of monocytes median Mon/CD3 ratio in cell suspension was 0,05 (P< 0,0001) and the detection of VSTs significantly increased to 119; 88 and 16, respectively (p< 0,0001). Among 79 DLI products after cryopreservation the median frequency of CMV, EBV, ADV-specific T cells was 43; 16; 6 per 300 000 CD3 lymphocytes, respectively, median Mon/CD3 was 0,13. After adhesion on plastic the number of monocytes decreased by 4,8 times (Mon/CD3=0,027) and the frequency of VSTs detection increased approximately 5 times (p< 0,0001). The median frequency of CMV, EBV, ADV-specific T cells after monocyte depletion was 255; 153; 11 per 300 000 CD3 lymphocytes, respectively (p< 0,0001).

Conclusions: The excess of monocytes in cell suspension leads to inefficient detection of VSTs by ELISPOT assay and prior adhesion of monocytes can be recommended for correct VSTs detection in CD45RA-depleted T-cells products. The analysis of DLI aliquots after cryopreservation allows accurate quantification of specific antiviral T-lymphocytes administered to a patient and optimize antiviral therapy.

Conflict of interest: nothing to disclose

P121

Recovery of thymopoiesis in a paediatric cohort with acute GVHD - Can ECP help?

Aisling M Flinn, Catherine F Roberts, Xiao Nong Wang, Andrew R Gennery

Newcastle University, Institute of Cellular Medicine, Newcastle upon Tyne, United Kingdom

Background: Thymic damage inflicted by acute GVHD, corticosteroids and other non-selective immunosuppressive agents negatively impacts T-lymphocyte reconstitution following HSCT, essential for a successful clinical outcome. Extracorporeal photopheresis (ECP) is an alternative immune-modulating therapy that could, by promoting immune tolerance, reducing acute GVHD and allowing weaning of concurrent immunosuppression, facilitate thymic recovery and restoration of normal thymopoiesis.

Methods: Whole blood samples were taken from 8 paediatric patients with corticosteroid-refractory/dependent acute GVHD before ECP cycles at regular intervals. Thymopoiesis was prospectively assessed quantitatively by sequential monitoring of CD3⁺CD4⁺CD45RA⁺CD31⁺ naïve T-lymphocytes using flow cytometry and T-cell receptor excision circles (TRECs) using PCR, and qualitatively using T-cell receptor (TCR) DNA spectratyping analysis. Serum IL-7 was measured using ELISA as an indirect reflection of thymic output. Thymic recovery was compared to two control groups at 4, 8 and 12 months post HSCT; paediatric patients post HSCT who did not develop acute GVHD (group 1, n=6) and paediatric patients with acute GVHD who did not receive ECP (group 2, n=6). Statistical analysis was performed using the Kruskal-Wallis test, with p< 0.05 being significant.

Results: All ECP patients (median age at HSCT = 5.8 years), with grade II-IV acute GVHD, demonstrated low thymic output at the beginning of ECP treatment, regardless of timing post HSCT (median absolute count of naïve T-lymphocytes = 5 cells/µL, median number of TRECs = 191/ml of blood) and a highly abnormal TCR repertoire. Four patients have completed ECP therapy (median number of ECP cycles =21), treatment is ongoing for 3, and one patient has been withdrawn from treatment. Qualitative and quantitative improvement in thymic output is seen in those who have completed therapy, with an inverse relationship with IL-7 evident (patient example in Figure A). Faster thymic recovery is evident in those who started ECP earlier. Among those who continue on ECP, one patient shows evidence of improving thymic output to date. One patient has demonstrated ongoing negligible thymic output 20 months post HSCT despite 72 weeks of ECP therapy, suggesting irreversible thymic damage. Median age at HSCT of control group 1 was 7.3 years and group 2 was 8.1 years. Comparing thymic recovery between the three groups showed no significant difference in naïve T-lymphocyte absolute counts (p=0.61, 0.31, 0.84 at 4, 8 and 12 months respectively).

Conclusions: This demonstrates, for the first time, that ECP can facilitate thymopoiesis in paediatric patients with aGVHD, potentially to the same rate as patients without acute GVHD, although further prospective data are needed. Promoting an immune tolerant environment post HSCT may be more conducive to faster T-lymphocyte reconstitution and further approaches to achieve this are needed, rather than our current armamentarium of immunosuppressive agents. Improvement is not seen in all patients, suggesting a point of thymic 'no return'; when excessive damage sustained by the thymus results in loss of intrinsic regenerative capacity and persistently inadequate T-lymphocyte reconstitution. A biomarker to diagnose and monitor for thymic damage would be valuable.

Conflict of interest: None

[P121 Figure]

$$$$

P122

Retrospective comparison of DLI and g-DLI for lymphoid malignancies, myeloma and myeloid malignancies

Sylvain Lamure ^1,2 , Franciane Paul ^1,2 , Anne-Laure Gagez ¹ , Jérémy Delage ^1,3 , Nathalie Fegueux ¹ , Guillaume Cartron ^1,2 , Lu Zaho-Yang ⁴ , Jean-Luc Veyrune ⁴ , Patrice Ceballos ¹ , John De Vos ^2,4,5

¹ Montpellier University Hospital, Clinical Hematology, Montpellier, France; ² University of Montpellier, Montpellier, France; ³ Clinique du Parc, Montpellier, France; ⁴ Montpellier University Hospital, Unit for Cell Therapy, Montpellier, France; ⁵ Montpellier University Hospital, Institute for Regenerative Medicine and Biotherapy, Montpellier, France

Background: Donor lymphocyte infusion (DLI) can be use to prevent or cure hematological malignancies relapse after allogeneic stem cell transplantation, using antitumoral effect of donor T and NK cells. DLI's efficacy is correlated with low tumour burden, hence prophylactic and pre-emptive use for mixed chimerism and detectable minimal residual disease had been developped. This procedure is effective alone or associated with antitumoral chemotherapy or targeted therapy.

Recombinant human granulocyte colony stimulated factor (G-CSF) primed DLI (gDLI) derives from frozen aliquots of the initial peripheral hematological stem cell (PHSC) bag collection. PHSC often exceeds the required amount for transplantation (5 to 10x10⁶/kg), and cryoconservation is not pernicious for T lymphocyte. Based on anti-tumour outcome of T lymphocyte infused with PHSC, some team described efficacy and safety of gDLI in different situation. Biological effects of G-CSF on peripheral blood stem cells suggests a decrease in anti-tumoral activity of such primed DLI, such as: Th2 polarization, promotion of regulatory T cell and tolerogenic dendritic cell differentiation. Two retrospective studies compared the clinical equivalence for activity and toxicity of gDLI and classical DLI. They included 63 and 67 patients; mainly suffering of AML/MDS, and no difference had been shown on clinical outcome.

In this monocentric retrospective study, we aim to compare the response rate, the overall survival (OS), the GVHD rate and toxicity for patient who received gDLI and classical DLI. We intend to compare clinical outcome for different malignancies (ALL, myeloid malignancies, lymphoid malignancies and myeloma) and different situation (clinical relapse, molecular relapse and mixed chimerism).

Methods: We performed systematic review of medical charts of patient who underwent hematological stem cell transplantation and donor lymphocyte infusion for hematological malignancies in Montpellier University Hospital between 1998 and 2016.

Results: We retrospectively evaluated data of 140 patients. Initial diseases were myeloid malignancies (N=56), lymphoid malignancies (N=27), acute lymphoid leukemia (N=21) and myeloma (N=36). Indications for DLI were relapse (N=95), pre-emptive (detectable minimal residual disease, N=18, mixed chimerism, N=23) and prophylaxis (N=4). Sixty for patients of them had classical DLI, 66 had gDLI, and 10 had successively gDLI and classical DLI. Response rate was 37.4%, median overall survival (OS) was 21.8 month, and 5-years OS was 29.7%. Graft versus host disease occurred for 47,9% of them, treatment-requiring GVHD for 24.3%, and 9,4% died from toxicity. There was no differences between classical and gDLI in term of response (p=0.31), OS (p=0.8), GVHD (p=0,44), treated GVHD (p=0.78) and cause for mortality (p=0.14). Factor associated with poor response were 3-month mixed chimerism and relapse indication for DLI, myeloid malignancies and relapse indication were associated with shorter OS. Sub-group analysis shows no difference between classical and gDLI in outcome for lymphoid or myeloid malignancies, myeloma or ALL, neither in curative or pre-emptive indication.

Conclusions: Taken together our results and those previously published strongly suggests similar efficiency and toxicity of gDLI and classical DLI, for myeloid malignancies as well as multiple myeloma, ALL and lymphoma, to treat or prevent clinical or molecular relapse. G-DLI may substitute classical DLI.

Conflict of interest: No conflict of interest to declare.

P123

Risk factors and novel biomarkers for cytokine release syndrome after CD19-targeted chimeric antigen receptor T cell therapy

Yongxian Hu ¹ , Lijuan Ding ¹ , Lei Xiao ² , AlexH Zhang ³ , He Huang ¹

¹ Zhejiang University, Hangzhou, China; ² Shanghai Innovative Cellular Therapy, CO LMD, Shanghai, China; ³ Tongji University, School of Medicine, Shanghai, China

Background: CD19-targeted chimeric antigen receptor T cells (CART19s) are a highly effective novel immunotherapy for relapsed/refractory B-cell malignancies. Cytokine release syndrome (CRS) is the most significant and life-threatening complication. Our understanding of CRS continues to evolve, and identification of risk factors and novel laboratory biomarkers are needed to evaluate strategies to mitigate toxicity.

Methods: Patient clinical characteristics were analyzed in 42 patients who received CART19 therapy. A multiplex cytokine array was performed to detect 900 cytokine levels in sera from 3 stages (pre-CART19 infusion, peak CRS and CRS restoration) of 3 patients with grade 3 CRS. Based on these data, 16 selected cytokines in 42 patients undergoing CART19 therapy were evaluated by ELISA.

Results: CRS developed in 81% of patients (grade 1, 14.3%; grade 2, 28.6%; grade 3, 38.1%). Multivariable analysis of patient clinical characteristics identified high tumor burden, thrombocytopenia before lymphodepletion and number of previous relapse as independent risk factors of CRS. 16 cytokines (TECK, TGFα, IP-10, MCP-2, FCAR, CXCL13, XCL1, IL-1 R4, AGRP, BAFF-R, BDNF, BRCK, CCL2, CCL3, GPC2, MMP-13) were identified as possible novel biomarkers by cytokine array (Figure 1A). IL-1 R4, MCP-1 and XCL1 were confirmed as novel biomarkers of CRS by ELISA. At stages of pre-infusion, peak CRS and CRS restoration, mean serum levels of IL-1 R4 were 56.6±49.2pg/ml, 2473.8±449.2pg/ml and 45±39.5pg/ml respectively; mean serum levels of MCP-1 were 16.6±15.2pg/ml, 476.9±341.2pg/ml and 21.5±27.6pg/ml respectively; mean serum levels of XCL1 were 9.6±7.2pg/ml, 77.9±45.8pg/ml and 5.9±3.5pg/ml respectively. All the above mean cytokine levels at stage of peak CRS were significantly higher than stages of pre-infusion and CRS restoration respectively (P< 0.01, Figure 1B). IFN-γ is one of the classical cytokine signature for CRS. Thus paired peak serum levels of IFN-γ, IL-1 R4, MCP-1 and XCL1 showed strong correlations (IFN-γ VS IL-1 R4, Spearman r=0.342, p=0.023; IFN-γ VS MCP-1, Spearman r=0.234, p=0.035; IFN-γ VS XCL1, Spearman r=0.532, p=0.014) implying that MCP-1, IL-1 R4 and XCL1 are novel biomarkers for CRS similar with IFN-γ.

Conclusions: Our data provide novel clues for prevention and treatment of CRS after CART19 therapy. Also our study lay the foundation for clarification of CRS pathophysiology.

Conflict of interest: None of the authors has anything to disclose.

[P123 Figure]

$$$$

P124

Safety and feasibility of extracorporeal photopheresis (ECP) as second-line therapy for treatment of pediatric patients with acute GvHD

Marlene Luther ¹ , Lisa Olfe ² , Kamran Movassaghi ² , Ina Lange ² , Annette Künkele ² , Johannes Schulte ² , Angelika Eggert ² , Agnieszka Blum ² , Lutz Uharek ² , Carola Tietze-Bürger ²

¹ Charité Campus Virchow-Klinikum, Stem Cell Facility, Berlin, Germany; ² Charité Campus Virchow-Klinikum, Berlin, Germany

Background: Only 30 - 60 % of patients with severe acute GvHD respond to the first-line therapy with systemic steroid treatment. In adult patients, ECP has been established as a promising second-line therapy with an immunomodulatory but no generalized immunosuppressive effect. As there is so far only limited experience in pediatric patients, we evaluated safety and outcome in a retrospective study.

Methods: Between 2014 and 2017 we treated 10 children with at least 3 cycles of ECP (median 9 cycles) for aGvHD. The median weight was 25 kg (range 11 - 91 kg) with a median age of 11yr (range 1 - 15 yr). For children weighing less than 25 kg the leukapheresis circuit was primed with RBC. A total of 89 cycles were analyzed.

Results: We saw no adverse events CTC grade 3/4 during ECP treatment, in particular regarding hypotension and anaphylactic reactions. A severe sepsis which was unrelated to ECP, resulted in a fatal outcome in one patient. During 89 cycles of ECP only 3 RBC transfusions and 4 PLT transfusions had to be given because of anemia or thrombocytopenia. At the end of ECP treatment, 9 of 10 children improved with 4 complete remissions of GvHD and 5 partial responses with at least 50% improvement. Also, a ≥ 50% reduction in steroid dose at the end of ECP treatment was possible in 9 out of 10 patients. The median time from onset of ECP treatment to end of immunosuppressive medication was 5 months for steroids and 9 months for all immunosuppressive medication, respectively.

Conclusions: ECP is a feasible and sufficiently safe treatment for pediatric patients with acute refractory GvHD. However, treatment of younger children requires special operating standards, a well trained staff, and is technically and logistically more demanding as compared to ECP in adults. Our preliminary clinical data confirm previous studies and suggest that the effort is justified because most children clearly improved after ECP.

Conflict of interest: None of the authors has anything to disclose.

P125

Small scale production of clinical grade MSCs for a single center setting

Adomas Bukauskas ¹ , Mindaugas Stoskus ¹ , Vilma Valceckiene ¹ , Laimonas Griskevicius ^1,2

¹ Vilnius University Hospital Santaros Klinikos, Hematology, Oncology and Transfusion Medicine Center, Vilnius, Lithuania; ² Vilnius University, Department of Internal, Family Medicine and Oncology, Faculty of Medicine, Vilnius, Lithuania

Background: Classic cell culture flask method is an effective way to expand mesenchymal stromal cells (MSCs) but the demand for dedicated facilities, personnel and time-to-production is a significant hurdle for the small transplant center. Commercially available self-contained bioreactors increase cell yield and decrease manual work yet at the increased cost of disposables and culture medium. Herein we report a successful integration of both methods for clinical scale MSCs expansion following bone marrow aspirate processing through a filter-based device.

Methods: Human bone marrow was collected from healthy donors after informed consent was obtained. To retrieve MSCs fraction, filter-based Bone Marrow MSC Separation Device (Kaneka Corporation, Osaka, Japan) was used. All cells were grown in high glucose DMEM medium supplemented with 5% human platelet lysate and 5 U/ml heparin. Cell medium was exchanged every 3 days. After initial culture in flasks (Passage 0), cells were harvested and loaded on Quantum Cell Expansion System (Terumo BCT, Inc, Lakewood, CO, USA) (Passage 1). Following expansion in bioreactor MSCs were harvested and cryopreserved for therapeutic use. Samples were taken for quality control (sterility, mycoplasma DNA, endotoxin, genomic integrity and immunophenotype).

Results: 18 donors were 25.5 (range 22-35) years of age. The median of aspirated bone marrow fluid volume was only 20 ml (range 15-35). Bone marrow mononuclear cells (BM MNCS) culture in flasks (Passage 0) took a median of 14 days (range 13-24). Of note, 3 of 18 (17%) BM MNSCS cultures did not expand. The median yield of MSCs expanded in T-flasks was 38.4x106 (range 10-85 x10⁶). MSCs expansion in bioreactor took a median of 9 days (range 6-17). Doubling times ranged from 36.83 to 80.24 h with the median of 57.12 h (or 2.38 days). The final yield ranged from 275 to 1800 x10⁶ cells with the median of 550 x10⁶ cells per culture. The median time from bone marrow aspiration to final yield was 23 days (range 20-41). At harvesting, cell viability was ≥ 95.9%. All cells were compliant with flow cytometric ISCT criteria and were negative for both mycoplasma and endotoxin. One batch was contaminated with Staphylococcus Warneri and was not released for clinical use. 14 out of 18 (78%) batches were released for clinical use.

Conclusions: Filter based device allows initial reduction of flask surface area (up to 600 cm²) required for culture of MSCs and simplifies dramatically Passage 0 process. Withdrawal of small amount of bone marrow fluid may decrease the pain associated with aspiration procedure and the bone marrow dilution with peripheral blood. This approach could be successfully adapted in small-scale hospital laboratories for clinical expansion of MSCs.

Conflict of interest: None

P126

SUCCESSFUL MANAGEMENT OF SEVERE ADVERSE EFFECTS DURING AUTOLOGOUS PERIPHERAL BLOOD STEM CELL INFUSION AND POLICY IMPLEMENTATION REGARDING WASHING PRODUCTS

Laura Medina Marrero ¹ , José García-Arroba Peinado ¹ , Carme Talarn Forcadell ² , Laia Ramiro Infante ¹ , Virginia Callao Molina ¹ , Carmen Azqueta Molluna ¹ , Nuria Martínez Llonch ¹ , Elena Valdivia García ¹ , Margarita Codinach Creus ¹ , Margarita Blanco García ¹ , Susana García Gómez ¹ , Luciano Rodríguez Gómez ¹ , Lourdes Escoda Teigell ² , Sergi Querol Giner ¹

¹ Banc de Sang i Teixits, Barcelona, Spain; ² Hospital Joan XXIII, Tarragona, Spain

Background: Thawing and immediate infusion is the most common method used for autologous stem cell transplantation. Adverse reactions associated to these infusions have been well described and mainly associated to DMSO, despite other origins have also been reported, such as granulocyte contain or red blood cells osmotic lysis. Developing an infusion policy including post-thaw cell washing is important to allow administration of cell graft at special risk.

Methods: We present two cases of severe adverse reactions occurring during infusion of autologous stem cell grafts in two patients with non-Hodgkin's lymphoma.

Case 1 was a 58-year-old, 76 kg b/w patient, that experienced itchy and sore throat after infusion of half of the first bag. The reaction was partially solved by reducing the rate of infusion. Thereafter, facial flashing and erythema suddenly occurred, which were accompanied by low blood pressure (60/30 mmHg), tachycardia (110/minute), dizziness, nausea and vomits, and the infusion was interrupted. Immediately, a Trendelemburg position was adopted while saline solution, hydrocortisone and oxygen were administrated. One hour was required to recover normal vital signs, so one third of the bag could not be infused.

Case 2 was a 65-year-old, 58kg b/w patient that after ten minutes of infusion of the first bag developed low blood pressure (70/40 mmHg), profuse sweating and dizziness. Saline solution and hydrocortisone were administered with successful resolution of the clinical picture.

Results: In both cases, ward physicians immediately contacted with the cell processing facility to report the severe adverse effects, so that an emergency strategy was considered. Teams agreed returning the remaining cryopreserved bags to the core cell therapy lab for graft washing. Procedure was performed using the Sepax® device using an automated protocol (Smartwash) that resulted in a satisfactory cell yield and viability. Afterwards, patient 1 did not develop any further complication; meanwhile, patient 2 exhibited a mild cutaneous reaction that disappeared after hydrocortisone, but allowing the infusion of the whole volume in the bag. Both patients had a successful engraftment.

In both cases, the quantity of DMSO was compliant with our infusion policy (less than 1 gr per kg patient b/w) and other quality variables including viability and colony growth were correct. However, the amount of granulocytes per kg in the collected products was high, 6.07E8 and 6.59E8/kg, respectively. After revision, this was considered as the most probable cause of the SARs observed.

Conclusions: Post-thaw washing allowed that almost all cryopreserved cells were finally infused. Consequently, after reviewing scientific literature and our own experience, we established a policy to prevent these severe reactions. Cell products with more than 4E8 granulocytes/kg are now recommended for post-thaw washing or multi-day infusion according CD34 cell numbers.

Conflict of interest: Nothing to disclose

	CN/kg (E8)	CMN/kg (E8)	CD34/kg (E6)	GM/kg (E5)	CFU/CD34 (% eclone)
CASE 1	16.80	10.73	8.45	11.5	20
CASE 2	15.58	8.99	2.19	2.62	49

[ [P126 Table] PRODUCTS CHARACTERISTICS]

P127

Validation of a highly scalable CD34 selection procedure to support cell therapy strategies in haematopoietic progenitor cell transplantation

Carmen Azqueta Molluna ¹ , Nuria Martinez Llonch ¹ , Elena Valdivia Garcia ¹ , Margarita Codinach Creus ¹ , Margarita Blanco Garcia ¹ , Pere Barba ² , Maria Isabel Benitez Carabante ³ , Luisa Sisinni ⁴ , Júlia Marsal ⁵ , Cristina Diaz de Heredia ³ , Isabel Badell Serra ⁴ , Rodrigo Martino Bufarull ⁶ , Rocio Parody Porras ⁷ , Christelle Ferrà Coll ⁸ , Laura Medina Marrero ¹ , Susana G Gómez ¹ , Sergio Querol ¹

¹ Banc de Sang i Teixits, Cell Therapy Services, Barcelona, Spain; ² Hospital Vall d'Hebron, Hematology Service, Barcelona, Spain; ³ Hospital Materno-Infantil Vall d'Hebron, Barcelona, Spain; ⁴ Hospital de la Santa Creu i Sant Pau, Pediatrics Service, Barcelona, Spain; ⁵ Hospital Sant Joan de Deu, Barcelona, Spain; ⁶ Hospital de la Santa Creu i Sant Pau, Hematology Service, Barcelona, Spain; ⁷ Institut Català d'Oncologia Duran i Reynals, L'Hospitalet de Llobregat, Spain; ⁸ Institut Català d'Oncologia Trias i Pujol, Badalona, Spain

Background: CD34+ selection remains a very attractive method for different interventions in hematopitic stem cell transplantation. We currently use this methodology as a method of GVHD prevention or to boost haematopoiesis in the treatment of poor graft function (PGF) after transplantation. In order to standardize a high-throughput methodology for a core cell therapy lab, we developed and validated an automated cell processing procedure combining the Sepax® and the CliniMACS® systems.

Methods: We included 44 consecutive procedures. This new, automatic procedure is applied to peripheral blood apheresis. After product reception, first step consist on platelet depletion, and volume adjustment using the Smartredux program for Sepax device. Then, the cell product was incubated with the anti-CD34 monoclonal antibody and washed out using again the Sepax washing system, to further adjust the volume to this required by CliniMACS. Finally, CliniMACS tubing set was installed and the cells were introduced for CD34 selection procedure using the CD34 Selection program. Indications for CD34 selection were: CD34 enrichment, first step of CD45RA depletion both in the context of GVHD prophylaxis, and CD34+ cell boost related to PGF.

Results: Median time from starting to end was 4 hours. Table 1 shows the post-Sepax cells recovered wit purities and yields. Before selection, platelet depletion and monoclonal antibody washing using Sepax device achieved extensive platelet depletion (up to a median of 68%) with almost complete recovery of TNC and CD34 cells. Then, cells were adjusted to the required volume for immunomagnetic selection that achieved a consistent final product with high CD34 purity and a very low CD3 contain (5.2 CD3 log depletion).

Importantly, CD34 potency was maintained. CFUGM recovered after selection were 1.4E7 (0.1-5). Median clonogenic efficiency of seeded CD34 was 18% (5-38) that resulted in an optimal engraftment.

Conclusions: Here, we present a standardized procedure that resulted in a robust and fast, large-scale CD34 selection using the combination of two automated devices. This approach allowed an improvement of our previous results with a reproducible output of purity and yield, maintaining optimal levels of cell viability and potency that resulted in an efficient myeloid engraftment. This new automatic procedure allows a high throughput in a central core cell therapy lab.

Conflict of interest: The authors declare no conflict of interest

Variable	Post-Sepax	Post-CliniMACS
Total CD34+ cells	4.8E8 (2.0-7.6)	3.3E8 (0.5-6.3)
CD34+ cell purity	0.92% (0.74-0.94)	94% (65-99)
CD34+ cell viability	99% (94-100)	95% (83-100)
CD34+ cell yield	93% (76-107)	81% (36-108)
Platelet yield	32% (4-80)	N.A.
Total CD3+ cells	N.A.	8E4 (0.5-16)

[ [P127 Table] Cell purity and yields during positive selection]

P128

Abstract previously published

P129

Wharton’s jelly-derived mesenchymal stem cells treatment in adult patients in Poland

Anna Mucha ¹ , Mariusz Grudniak ¹ , Magda Murzyn ¹ , Jolanta Koryga ¹ , Artur Olkowicz ¹ , Karolina Popławska ¹ , Dominika Gładysz ¹ , Iwona Marszałek ¹ , Maciej Boruczkowski ² , Tomasz Ołdak ¹ , Dariusz Boruczkowski ¹

¹ PBKM, Warszawa, Poland; ² University of Medical Sciences, Poznań, Poland

Background: Wharton´s jelly-derived mesenchymal stem cells (WJ-MSCs) are a type of cell population with significant self-renewal and multi-lineage differentiation properties. Their high proliferation rate, immune privileged status, nontumorigenic properties make them ideal for both autologous and allogeneic use in regenerative medicine applications. WJ-MSCs therapy, based on cells prepared by Polish Bank of Stem Cells (PBKM), has shown promising effects in the treatment of many diseases and injuries.

Methods: From 2007 to 2017 PBKM has delivered WJ-MSCs products for 371 adult patients aged from 19 years and 2 months to 86 years and 5 months (median age: 47 years and 2 months). The patients after Bioethical Committee approval, received intravenous or intrathecal injections of WJ-MSCs, obtained from third party donor (TPD). The cells were collected from healthy newborns, then processed, screened for bacterial contamination and endotoxin content, and finally frozen in liquid nitrogen vapours. WJ-MSCs immunophenotype was confirmed using flow cytometry assay. The patients received from 1 to 5 injections in one medical course. The average cell dose per infusion was 1 x 10^6/kg of body weight. All patients were monitored for the procedure related adverse events.

Results: WJ-MSCs have been used in adult patients for more than 50 diseases and were most often used in neurology (66% of all patients), ophthalmology (19%) as well as orthopedics (13%). The stem cells were also given to patients in hematology (4 patients) and gynecology (4 patients). The most common indications were amyotrophic lateral sclerosis (153 patients), retinitis pigmentosa (29 patients) and spinal cord injury (27 patients). All WJ-MSCs administrations took place in Poland. Immediately after stem cell infusion 16 adverse events (AE) occured and another 40 AE at least 24 hours after WJ-MSC administration were observed. All AE were mild or moderate, eg. the most common was headache or nausea, which were easily managed with medications.

Conclusions: WJ-MSCs have a wide implementation in the medical experimental procedures and show promising preliminary results with high therapeutic value in adult patients. However further research is needed and long-term results have to be collected.

Conflict of interest: All the authors are employed at PBKM.

P130

Wharton’s jelly-derived mesenchymal stem cells treatment in children with neurological diseases

Anna Mucha ¹ , Mariusz Grudniak ¹ , Magda Chrościńska-Krawczyk ² , Monika Kotarska ² , Krystyna Mitosek - Szewczyk ² , Magda Murzyn ¹ , Jolanta Koryga ¹ , Artur Olkowicz ¹ , Karolina Popławska ¹ , Dominika Gładysz ¹ , Iwona Marszałek ¹ , Maciej Boruczkowski ³ , Tomasz Ołdak ¹ , Dariusz Boruczkowski ⁴

¹ PBKM, Warsaw, Poland; ² Lublin Medical University, Department of Pediatric Neurology, Lublin, Poland; ³ University of Medical Sciences, Poznan, Poland; ⁴ PBKM, Warszawa, Poland

Background: Many neurological diseases cannot be efficiently treated, because of their unknown pathogenesis and the lack of appropriate medications. The concept of repairing the nervous system by stem cells has recently evolved from hypothesis that tissue regeneration could have been achieved through stem cell differentiation into neural cells, acting mainly by the means of paracrine mechanisms. Transplantations of stem cells capable of producing neurotrophic factors may be an efficient method for improving brain function. Wharton´s jelly mesenchymal stem cells (WJ-MSCs), thanks to Polish Bank of Stem Cells (PBKM), have been in use in children´s neurology in Poland since 2014.

Methods: Two hundred and forty three neurological patients aged from 5 months to 18 years and 8 months (median age: 7 years and 2 months), after Bioethical Committee approval, received intravenous or intrathecal injections of WJ-MSCs, obtained from third-party donor (TPD). The cells were previously collected from healthy newborns, then processed, screened for bacterial contamination as well as endotoxin content, and frozen in liquid nitrogen vapour. WJ-MSCs immunophenotype was confirmed using flow cytometry assay. The patients received from 1 to 10 injections (5 in one medical course). The average cell dose per intravenous infusion was 1 x 10^6/kg of body weight. Each patient was examined by the same neurologist at the day of infusion.

Results: Neurological patients constisted of 87% of all pediatric patients listed in PBKM in years 2011- 2017. The most common indications were cerebral palsy (74 patients), autism (41 patients) and spina bifida (22 patients). The remaining 86 patients have been diagnosed with other neurological diseases (total number of diagnosis: 46). All WJ-MSCs administrations took place in Poland. Adverse events (AE) recorded during injections were easily managed with medications and occured 8 times (headache - 4, nausea - 2, lumbar puncture headache -1, blood pressure increase -1).

Conclusions: The administrations of third-party donor WJ-MSCs seem to be safe and efficient procedure with promising preliminary results in neurological patients. However more insight in the study will be beneficial, therefore long-term results still need to be collected.

Conflict of interest: All the authors are employed at PBKM.

Chronic leukaemia and other myeloproliferative disorders

P131

Allogeneic stem cell transplantation in Philadelphia-negative chronic myeloproliferative neoplasms. A centre experience

Alejandro Avendaño Pita ¹ , Javier Carrillo Checa ¹ , Daniel Rivera Delgado ¹ , Veiga Vaz Alvaro ² , Arratibel Zalacain Nerea ² , Baile González Mónica ² , Félix López Cadenas ² , Perez López Estefania ² , Lucia López Corral ¹ , Cabrero Calvo Monica ¹ , Jose Ramón González Porras ¹ , Maria Dolores Caballero Barrigon ¹ , Ana Africa Martin Lopez ¹

¹ Hospital Universitario de Salamanca/IBSAL, Salamanca, Spain; ² Hospital Universitario de Salamanca, Salamanca, Spain

Background: Nowadays, allogeneic stem cell transplantation (ASCT) is the only curative strategy in patients with high risk Philadelphia-negative chronic myeloproliferative neoplasms (CMN).

Methods: From 867 patients who underwent to an ASCT in our centre, 21 (2.42%) were CMN different from chronic myeloid leukaemia. We retrospectively analysed this subgroup of patients: 8 (38.1%) primary myelofibrosis (MF), 11 (52.4%) secondary MF (6 post PV, 4 post ET and 1 post other CMN) and 2 CMN different from MF (1 CMN JAK2+ and 1 CMN/MDS JAK2+ with fibrosis).

Results: The clinical and biological features and ASCT characteristics are summarized in table 1.

All patients, except two that were not evaluable, achieved neutrophil and platelet engraftment (only two patients did not reach 50.000 platelets). Concerning the complications related to ASCT: 71.4% of patients suffered from mucositis (86.7% grade I and II), 23.8% developed CMV reactivation, 9.5% haemorrhagic cystitis, 4.8% obstruction sinusoidal syndrome and none developed transplant associated microangiopathy. A total of 16 patients (76.2%) developed acute graft versus host disease (GVHD) (18.8 % grades III-IV) with a median of presentation of 26 days (15-141) and 11 patients (52.4%) developed chronic GVHD (63.6% extensive) with a median of 203 days (120-566). Transplant related mortality (TRM) at day +100 was 4.8% and the global TRM.

At day +100 post ASCT, 17 patients (80.9%) achieved complete remission (CR) (two of them with minimal residual disease) and 3 (14.3%) patients were in relapse situation (one case not evaluable because of early dead). Six (28.6%) of these 20 patients relapsed after a median of 150 days (53-2295) from ASCT. In four of them, the approach of relapse was modulating immunosuppression +/- donor lymphocyte infusion and hypometilating agents in the cases of MRD, reaching CR and complete chimerism.

In the last follow-up, 13 patients (61.9%) were alive (11 in CR and 2 with active disease) and 8 (38.1%) were dead from different causes: relapse in 1 case (4.8%), infection in 5 (23.8%), bleeding in 1 (4.8%) and the last one from another causes (hepatic failure).

With a median follow-up of 35 months (16-121) the overall survival (OS) and event free survival (EFS) at 3 years was 61% and 57%, respectively.

Conclusions: The results of our series confirms the potential curative of ASCT. Taking into account that TRM is not minimal, we observe that more than a half of patients are long survivors. Modulating immunosuppression plays an important role in the setting of post ASCT relapse.

Conflict of interest: Nothing to disclose

Clinical and biological characteristics (N=21)	N (%)	Median (range)	ASCT characteristics	N (%)	Median (range)
Age (years)		60 (29-68)	Age		61 (30-69)
Sex: Male Female	11 (52.4) 10 (47.6)		Status at ASCT: CR PR Clinical improvement Active disease Progression	1 (4.8) 5 (23.8) 3 (14.3) 8 (38.1) 4 (19.0)
AML transformation pre-ASCT	5 (23.8)		Peripheral blood as source of stem cells	21 (100)
Time to AML transformation (months)		8.8 (2.7-117.2)	Number of CD34+ (106/kg)	6.59
Driver mutations (N=18): JAK2 CALR MPL Triple negative	13 (72.2) 2 (11.1) 2 (11.1) 1 (5.6)		Type of donor: Sibling donor Not sibling donor Haploidentical	15 (71.4) 4 (19) 2 (9.5)
IPSS (N=18): Low/int-1 Int-2/high	4 (22.2) 14 (77.8)		Donor median age		51 (30-65)
DIPSS-plus (N=18) Low/int-1 Int-2/high	3 (16.7) 15 (83.3)		Conditioning: MAC RIC	5 (23.8) 16 (76.2)
Lines of treatment pre-ASCT		1 (0-4)	Status CMV: +/+ -/+ -/-	16 (76.2) 4 (19) 1 (4.8)
Type of treatment: Hidroxyurea Splenectomy JAK inhibitor Chemotherapy Previous trasplant	9 (42.9) 4 (19) 11 (52.4) 6 (28.6) 2 (9.5)		ABO: ABO identical Major incomp. Minor incomp. Major & Minor	15 (71.4) 2 (9.5) 1 (4.8) 3 (14.3)

[ [P131 Table] Table 1]

P132

Chronic Neutrophilic Leukemia, An Extremely Rare Cause Of Neutrophilia In Childhood: Cure With Hematopoetic Stem Cell Transplantation

Akif Yeşilipek ¹ , Vedat Uygun ¹ , Seda Öztürkmen ¹ , Hayriye Daloğlu ¹ , Neşe Yaralı ² , Gülsün Karasu ¹

¹ MedicalPark Antalya, Pediatric Bone Marrow Transplantation Unit, Antalya, Turkey; ² Ankara Child Health and Diseases Haematology Oncology Training and Research Hospital, Department of Paediatric Haematology and Oncology, Ankara, Turkey

Background: Chronic neutrophilic leukemia (CNL) is a rare myeloproliferative disorder that occurs frequently after the mutations in gene for receptor of colony-stimulating factor 3 (CSF3R). Here we report a case of an Fanconi Anemia (FA) patient who developed CNL during follow-up and was cured with Hematopoetic Stem Cell Transplantation (HSCT).

Methods: The patient was referred to pediatric hematology department at the age of 8 years because of recurrent pulmonary infections and accompanying leukopenia. A workup revealed an elevated DEB-induced chromosome breakage leading to diagnosis of FA. At 12 years of age, WBC count increased to 15×10⁹/ L; however, at the next visit 2 months later, the WBC count was increased to 63.8×109/L, which required further examination. Infection and reactive causes were investigated because of the presence of neutrophils representing 70% of the leukocytes in the peripheral smear, but no causative agent was detected. T618I mutation was detected in the CSF3R gene, which is frequently reported in patients with CNL. HSCT was planned because the patient had FA, which could accelerate the progression to AML.

Results: The patient was scheduled for transplantation from a 9/10 compatible unrelated donor. In the preparation regimen, 180 mg/m² fludarabine, 40 mg/kg cyclophosphamide, and 30 mg/kg anti-thymocyte globulin (ATG fresenius) were used. Cyclosporine and mycophenolate were used for graft versus host disease (GVHD) prophylaxis.GCSF was not used due to the CSF3R mutation history; the neutrophils and platelets engrafted at +18 and +9, respectively. Grade II acute GVHD developed at the third week and resolved with methylprednisolone. However, chronic GVHD findings appeared on the eye, mouth, and skin at month 7 of the transplant, which was accepted as moderate chronic GVHD. Currently, at month 9 of the transplant, the patient is still experiencing mild chronic GVHD with full chimerism and is receiving tacrolimus and methylprednisolone. He now has no CSF3R mutation and has a leukocyte level of 7.5×10⁹/L.

Conclusions: The association between FA and CSF3R mutation was not reported thus far. To our knowledge, the patient studied is the youngest CNL patient with a possible acquired mutation; moreover, it is the first case of CNL reported in a patient with FA. The CSF3R mutation should not be overlooked in patients with FA who have long-standing neutrophilia, because the mutation is a predisposing factor for developing AML.

Conflict of interest: Nothing to disclose

P133

CML: Stem Cell Transplantation in Children and Adolescents with Reduced Intensity Conditioning - a Multicentric Prospective Trial

Susanne Matthes ^1,2 , Petr Sedlacek ³ , Roland Meisel ⁴ , Rita Beier ⁵ , Maura Faraci ⁶ , Krzysztof Kalwak ⁷ , Marianne Ifversen ⁸ , Ingo Müller ⁹ , Jerry Stein ¹⁰ , Kim Vettenranta ¹¹ , Gabriele Kropshofer ¹² , Sabina Sufliarska ¹³ , Adriana Balduzzi ¹⁴

¹ Medical Center – University of FreiburgUniversity, Vienna, Stem Cell Transplantation, Vienna, Austria; ² St Anna Childrens Hospital, Vienna, Austria; ³ Charles University Motol Prague, Pediatric Hemato-Oncology, Prague, Czech Republic; ⁴ Universitätsklinikum Düsseldorf, Pediatric Hemato-Oncology, Düsseldorf, Germany; ⁵ University Children HospitalChildrens Hospital, Essen, Germany; ⁶ Istituto di Ematologia, Università Cattolica del S. CuoreG. Gaslini, Genova, Italy; ⁷ Wroclaw Medical University, Pediatric BMT, Wroclaw, Poland; ⁸ Children and Adolescent Medicine Rigshosptilet, Copenhagen, Denmark; ⁹ University Medical Center Hamburg-Eppendorf, Pediatric Stem Cell Transplantation, Hamburg, Germany; ¹⁰ Schneider Children's Medical Center, BMT Unit, Petach Tikva, Israel; ¹¹ University of Adelaide, Hospital for Children and Adolescents, Helsinki, Finland; ¹² University Children Hospital, Hematology Oncology, Innsbruck, Austria; ¹³ University Childrens Hospital, Pediatric Hematology/Oncology, Bratislava, Slovakia; ¹⁴ Clinica di Oncoematologia Pediatrica Ospedale San Gerardo, Monza, Italy

Background: In the era of TKIs life expectancy of adult/elderly patients with CML, responding to Imatinib approaches that of the general population. Only selected patients are being treated with SCT, although PFS and OS >80% have been reported for patients with an EBMT-score of 0-2.

In children and adolescents with CML treated with Imatinib, 10-year PFS is 60% and TKI-free survival is 6% (Giona et al 2015). Response rates in children treated with second line TKIs are comparable to those observed with Imatinib. Besides the fact that life-long TKI for children with CML implicates treatment for 50-70 years, non-target effects of Imatinib such as growth deficiency and chronic immunosuppression, treatment failures due to non-compliance in adolescents remain major concerns. In a prospective multicentric trial (ClinicalTrials.gov NCT02707393) the SCT commettee of the IBFM study group evaluated the safety and efficacy of SCT following a reduced intensity conditioning regimen (RIC) in children and adolescents with CML.

Methods: In order to reduce TRM, late effects and chronic GvHD the protocol was based on an ATG-containing RIC regimen (Fludarabin 4 x 40 mg/m², Thiotepa 2 x 5 mg/kg, Melphalan 140 mg/m²). The protocol included frequent BCR/ABL Monitoring post SCT and treatment with Imatinib in the first year in case of increasing BCR/ABL. Patients with persisting BCR/ABL > 0,02% received DLI in the second year after SCT.

Results: A total of 30 patients have been included. Donors were matched related (10) and matched unrelated (20). Median age at SCT was 14 years. All patients were in CP1 and had received Imatinib or second line TKI. BCR/ABL at the time of SCT ranged from 0,001-33%. Indication for SCT was insufficient molecular response/secondary response loss in 19 patients and patient/physician choice for the remaining patients. All patients engrafted, 7/30 patients displayed mixed chimerism (8-80% donor chimerism). TRM was 4% (16 year old boy with BCR/ABL > 30% at time of SCT who died on day + 180 of invasive adenovirus infection). Moderate/severe chronic GvHD was observed in 6/29 patients which resolved in all cases, with no patients receiving immunosuppression beyond 2 years after SCT. There is no correlation between mixed chimerism and the need of secondary intervention. In six patients Imatinib treatment was restarted after a median of 4 months. Three patients received additional DLI in the second year after SCT, 2/3 being off TKI and in MR 5.5 at the time of analysis. No DLI-associated GvHD ocurred.

After a median follow-up of 52 months, current survival with BCR/ABL < 0,01% is 97% and current TKI-free survival is 90%.

Conclusions: SCT following RIC is feasible in children and adolescents with CML and associated with a disease-free and TKI-free survival of > 90%. Patients with insufficient molecular Response or secondary response loss can be treated successfully with RIC-SCT. RIC-SCT may be a valid alternative to life-long treatment with TKIs in children and adolescents with CML.

Clinical Trial Registry: ClinicalTrials.gov NCT02707393

Conflict of interest: None of the authors has anything to disclose.

P134

GETH-CLL4: Long-term follow-up of a pilot trial evaluating the addition of ofatumumab to the conditioning regimen of patients with chronic lymphocytic leukemia undergoing allogeneic HCT

Pau Montesinos ¹ , Lucia Lopez-Corral ² , David Valcarcel ³ , Montserrat Rovira ⁴ , Jose Antonio Garcia-Marco ⁵ , Javier Loscertales ⁶ , Carol Moreno ⁷ , Rafael Duarte ⁸ , Maria Jose Terol ⁹ , Pau Abrisqueta ³ , Monica Cabrero ² , Jaime Sanz ¹ , Julio Delgado ⁴

¹ Hospital La Fe, Valencia, Spain; ² Hospital Clinico, Salamanca, Spain; ³ Hospital Vall d'Hebron, Barcelona, Spain; ⁴ Hospital Clinic, Barcelona, Spain; ⁵ Hospital Puerta de Hierro, Madrid, Spain; ⁶ Hospital La Princesa, Madrid, Spain; ⁷ Hospital Santa Creu i Sant Pau, Barcelona, Spain; ⁸ ICO - Hospital Duran i Reynals, Barcelona, Spain; ⁹ Hospital Clinic, Valencia, Spain

Background: Patients with CLL whose tumor cells harbor TP53 aberrations or relapsing early after chemoimmunotherapy are considered to have an adverse prognosis with conventional agents. These patients have been traditionally offered an alloHCT according to EBMT guidelines. The addition of rituximab to the typical fludarabine-based conditioning regimen has apparently improved the outcome of patients with CLL. In this investigator-initiated trial we hypothesized that the addition of ofatumumab, a novel anti-CD20 monoclonal antibody with greater anti-CLL activity, to the conditioning regimen could also improve the outcome of these patients.

Methods: Patients with CD20-positive CLL were eligible in case of TP53 aberrations or disease relapse within 12 months of the last chemoimmunotherapy. The disease had to be in complete or partial response at the time of transplantation. Donors were siblings or unrelated volunteers with 9-10/10 HLA compatibility. The conditioning regimen comprised ofatumumab (300 mg on day −20, 2000 mg on days −13 and −6 and 1000 mg on days +1 and +8); fludarabine (150 mg/m2); and melphalan (70-140 mg/m2). GvHD prophylaxis was with cyclosporine A and mycophenolate mofetil. Patients with one-allele mismatch also received rabbit thymoglobulin (8 mg/kg). The primary end point of the study was progression-free survival (PFS) at 3 years.

Results: Median age of the 18 recruited patients was 57 (range, 35-62) years, and 13 (72%) were male. Donors were HLA-matched in all but two cases. All patients proceeded to alloHCT except one whose donor became pregnant after initiation of ofatumumab therapy and was withdrawn from the study. The remaining 17 patients completed all ofatumumab infusions. All patients who underwent alloHCT engrafted. The cumulative incidence of developing any kind of GvHD was 76% at 1 and 3 years, respectively. Moreover, the cumulative incidence of developing severe GvHD (defined as grade 3-4 acute or extensive chronic GvHD) was 53% and 65% at 1 and 3 years, respectively. With a median follow-up of 53 (range, 36-60) months for survivors, there have been no relapses on study. As no relapses were observed in the study, both PFS and overall survival were identical: 47% (95% CI 28-78%) and 40% (95% CI 22-73%) at 3 and 5 years from transplantation, respectively (Figure 1). Causes of death were septicemia (3), pneumonia (2), GvHD (2), cerebral hemorrhage (1), PML (1) and hemoptysis (1).

Conclusions: The addition of ofatumumab to the conditioning regimen of CLL patients undergoing alloHCT reduced the relapse rate completely, but was not able to reduce the GvHD rate, thus leading to a very high non-relapse mortality. However, the long-term follow-up of this trial shows a clear plateau in the survival curve and therefore alloHCT may be still recommended in a restricted group of patients. There is, however, a clear need to identify those patients with a low risk of non-relapse mortality, as has been the aim of recent EBMT initiatives.

Clinical Trial Registry: This study was registered at clinicaltrials.gov (NCT01455051) and performed in accordance with the Declaration of Helsinki and the International Conference on Harmonized Guidelines for Good Clinical Practice.

Conflict of interest: For the purpose of this trial, ofatumumab was provided free of charge together with an unrestricted grant from GSK/Novartis. None of the authors has anything to disclose.

[P134 Figure]

$$$$

P135

Impact of target therapy on outcome after allogeneic stem cell transplantation in patients with BCR-ABL-negative myeloproliferative neoplasms: a single center experience

Maria Barabanshikova, Elena Morozova, Ivan Moiseev, Julia Vlasova, Vadim Baikov, Ildar Barkhatov, Elena Darskaya, Sergey Bondarenko, Boris Afanasyev

First Pavlov State Medical University of Saint Petersburg, Raisa Gorbacheva Memorial Institute of Children Oncology, Hematology and Transplantation, Saint Petersburg, Russian Federation

Background: Post-polycythemia vera (post-PVMF), post-essential thrombocythemia (post-ETMF), primary myelofibrosis (PMF) and myelodysplastic/myeloproliferative neoplasms (MDS/MPN) are BCR-ABL-negative hematopoietic stem cell disorders with generally poor outcome. The aim of the study was to compare the results of allogeneic hematopoietic stem cell transplantation (alloHSCT) in patients with myelofibrosis (MF) and MDS/MPN and evaluate if target therapy with ruxolitinib improves the outcome in MF.

Methods: We compared 3 groups of patients: MDS/MPN, MF without history of ruxolitinib, and MF with pre- and post-transplant ruxolitininb. 40 patients aged 49 (30-61) years were included in the analysis. 26 patients were diagnosed with MF (3 - post-ETMF, 6 - post-PVMF, 20 - PMF), 11 - MDS/MPN (7 - СMML, 1 - aCML, 3 - MDS/MPN unclassifiable). All patients received conditioning regimen with fludarabine 180 mg/m² plus busulfan 8-10 mg/kg. In 24 MFand MDS/MPN-patients without ruxolitinib graft versus host disease prophylaxis (GVHD) consisted of antithymocytic globulin (ATGAM 60 mg/kg or thymoglobuline 5 mg/kg) and tacrolimus/mophethyl-mycophenolate. 16 MF-patients received ruxolitinib 30-45 mg daily at least 3 months until D-1 and 10-15 mg from D+5 to D+100 and GVHD prophylaxis with cyclophosphamide 50 mg/kg D+3,+4. Among them disease stabilization before alloHSCT occurred in 11 patients, disease progression - 1, clinical improvement - 4 according to ELN criteria. All MDS/MPN-patients had active disease at the moment of alloHSCT. AlloHSCT was performed from unrelated (29), related (7) and haploidentical (4) donor. Stem cell source was granulocyte colony-stimulating factor mobilized peripheral blood progenitor cell (PBSC) (29) and bone marrow (11). Median number of CD34+cells/kg was 6,5 x 10⁶ (3,0 - 11,9) in PBSC group and 2,7 x 10⁶ (1,0 - 8,0) in bone marrow group.

Results: Median follow up was 13,7 months (3,6-126,1) Primary engraftment was documented in 30 patients. The rate of primary graft failure was significantly higher in MDS/MPN- compared to MF-patients, 54% versus 16% subsequently (p=0.014). All of them died due to infection (8) or disease progression (2).

Acute GVHD grade 2-4 and chronic GVHD was documented in 33% and 50% of MF and MDS/MPN-patients without history of ruxolitinib therapy, 26% and 33% of MF-patients who received ruxolitinib (p=ns). The highest incidence of severe sepsis was observed in MDS/MPN-patients - 63% compared to MF-patients with or without history of target therapy - 24% (p=0.029). The rate of toxic hepatitis grade 3-4 and venoocclusive liver disease was similar between groups: 21% and 12% in patients with no history of ruxolitinib, 25% and 13% in patients treated with ruxolitinib.

We found no significant difference in 2-year OS between MDS/MPN- and MF-patients treated with conventional therapy (38% vs 18%, p=0.195), but the 2-year OS of MF-patients treated with ruxolitinib pre- and posttransplant was significantly superior (93% vs 38%, p=0.011) (figure 1).

Conclusions: In our small patient cohort there was no difference between MPN and MF on conventional therapy. However introduction of JAK inhibitor in the pre- and posttransplant period might improve results of alloHSCT in such unfavorable disease as MF. Larger multicenter studies are required to confirm this assumption.

Conflict of interest: None of the authors has anything to disclose.

[P135 Figure]

$$$$

P136

Abstract previously published

P137

Abstract previously published

P138

Outcome of Allogeneic Haematopoietic Stem Cell Transplantation in Myeloproliferative Neoplasms-Unclassified: A Retrospective Study by the Chronic Malignancies Working Party of the EBMT

Donal McLornan ^1,2 , Richard Szydlo ³ , Marie Robin ⁴ , Linda Koster ⁵ , Gernot Stuhler ⁶ , Renate Arnold ⁷ , Dietger Niederwieser ⁸ , Ghulam Mufti ⁹ , Dietrich Beelen ¹⁰ , Arnold Ganser ¹¹ , Aleks Radujkovic ¹² , Donald Bunjes ¹³ , Henrik Sengeloev ¹⁴ , Ibrahim Yakoub-Agha ¹⁵ , Ram Malladi ¹⁶ , Mahmoud Aljurf ¹⁷ , Igor Wolfgang Blau ⁷ , Peter Kalhs ¹⁸ , Stig Lenhoff ¹⁹ , Antonio Campos ²⁰ , Gerald G Wulf ²¹ , Mareike Verbeek ²² , Matthias Stelljes ²³ , Yves Beguin ²⁴ , Yves Chalandon ²⁵ , Nicolaus Kroger ²⁶

¹ Guy’s and St.Thomas NHS Foundation Trust, Haematology, London, United Kingdom; ² King's College London, London, United Kingdom; ³ Imperial College Healthcare NHS Trust, Department of Clinical Sciences, London, United Kingdom; ⁴ Hôpital Saint-Louis, Service d'Hématologie-Greffe, Paris, France; ⁵ EBMT Data Office, Leiden, Netherlands; ⁶ Deutsche Klinik für Diagnostik, Wiesbade, Germany; ⁷ Charité Universitaetsmedizin Berlin, Berlin, Germany; ⁸ University Hospital Leipzig, Division of Haematology & Oncology, Leipzig, Germany; ⁹ King's College Hospital/ King's College London, London, United Kingdom; ¹⁰ Univeristy Hospital Essen, Haematology, Essen, Germany; ¹¹ University of Hannover, Hannover, Germany; ¹² University of Heidelberg, Medizinische Klinik u. Poliklinik V, Heidelberg, Germany; ¹³ Universitätsklinikum Ulm, Ulm, Germany; ¹⁴ Rigshospitalet, Copenhagen, Denmark; ¹⁵ CHU de Lille, LIRIC, INSERM U 995, Universite de Lille, Lille, France; ¹⁶ Queen Elizabeth Hospital, Haematology, Birmingham, United Kingdom; ¹⁷ King Faisal Specialist Hospital & Research Center, Riyadh, Saudi Arabia; ¹⁸ Medizinische Universitaet Wien, Vienna, Austria; ¹⁹ Skanes University Hospital, Lund, Sweden; ²⁰ Inst. Português de Oncologia do Porto, Porto, Portugal; ²¹ Universitaetsklinikum Goettingen, Goettingen, Germany; ²² Klinkum Rechts der Isar, Munich, Germany; ²³ University of Münster, Munster, Germany; ²⁴ University of Liege, Liege, Belgium; ²⁵ Hôpitaux Universitaires de Genève, Geneva, Switzerland; ²⁶ University Hospital Eppendorf, Hamburg, Germany

Background: Myeloproliferative Neoplasm (MPN), unclassified (MPN-U) is a rare and inherently heterogeneous disease entity that presents with an MPN-type clinical/ histological phenotype yet fails to meet the diagnostic criteria for the other MPN variants. The clinical course is highly varied, no well-defined treatment algorithm exists and the outcome of allogeneic haematopoietic stem cell transplantation (allo-HSCT) has never been documented in detail. We hereby report on a multicentre, EBMT-registry based study of adult MPN-U patients.

Methods: All patients who underwent allo-HSCT for MPN-U between 2000-2015 (Reduced Intensity Conditioning (RIC) or Myeloablative Conditioning (MAC)) utilising either bone marrow or peripheral blood stem cells were selected. Statistical analyses were performed with SPSS 22 (SPSS Inc./IBM, Armonk, NY). All patients provided informed consent according to the Declaration of Helsinki.

Results: A total of 229 patients were analysed. MAC regimens, as defined by EBMT criteria, were utilised in 111 patients while 118 received RIC. Median age at allo-HSCT was 51 years (range(r), 19-71) in MAC and 58 (r,24-74) in RIC cohorts. Median time from diagnosis to allo-HSCT in the MAC and RIC cohorts= 12 months (r, 2-282) and 16 months (r,2-244) respectively. Donor types were as follows: MAC; matched sibling donors (MSD) n=40 (36%), unrelated donors (URD) n= 65 (59%) and mismatched related (MMR) n=6 (5%); RIC cohort MSD; n=36 (31%), URD; n=75 (63%) and MMR n= 7(6%). Median times to neutrophil engraftment for MAC and RIC platforms were identical at 18 days (p=0.97). Acute (a) and Chronic (c) Graft Versus Host Disease (GVHD) status was documented in 96% and 64% of the entire cohort. Rates of grade II-IV aGVHD were higher in the MAC cohort (33%) when compared to the RIC cohort (21%; p=0.05) whereas cGVHD rates (limited/ extensive) were similar: 21%/30% in the MAC cohort versus 18%/20% in the RIC cohort (p=0.87). The median follow up period of surviving patients was 30 months (range 1-182). Non-relapse mortality (NRM) probabilities at 1,3 and 5-years were considerable in both groups: 32%, 42%, and 52% (MAC) and 35%, 44% and 44% (RIC); p=0.81. Cumulative incidence of relapse (CIR) at 1,3 and 5-years was 22%, 35% and 39% (MAC) and 24%, 41% and 41% (RIC) (p=0.59). Moreover, the 1,3 and 5 year overall survivals (OS) rates for MAC and RIC were 64%, 46% and 34% and 59%, 40% and 39% respectively (p=0.68). Regarding the best time to transplant in the disease course, there was a near significant trend for improved outcome in the MAC setting when recipients underwent allo-HSCT within the first year of diagnosis compared to those transplanted later (p=0.08).

Conclusions: This is the largest study to date reporting on outcomes of allo-HSCT for MPN-U. Given the heterogeneity of the disease course, OS rates appear acceptable but it is of note that NRM and CIR rates in both settings are considerable. In the MAC setting, transplanting earlier in the disease course appears to be associated with improved survival. Of note, there was no significant difference in relapse rates between RIC and MAC. Multivariate and subgroup analyses will be presented.

Clinical Trial Registry: Not relevant

Conflict of interest: Nothing relevant to disclose and no conflict of interests for the authors

P139

Outcome of Allogeneic Transplantation in CML in the Tyrosinkinase Inhibitors Era, Single Centre Evaluation

Markéta Šťastná-Marková, Jan Vydra, Veronika Válková, Hana Klamová, Ludmila Nováková, Barbora Čemusová, Michal Kolář, Petr Cetkovský, Antonín Vítek

Institute of Haematology and Blood Transfusion, Praha, Czech Republic

Background: The introduction of tyrosinkinase inhibitors (TKIs) to the CML treatment has been a destiny changing event for the majority of patients. Allogeneic transplantation then remains a therapeutic option for patients, in which the use of TKIs is not suitable for any reason.

Methods: 46 patients with the diagnosis of CML transplanted between the years 2005-2017, which means in the period when at least imatinib was available in the front line therapy, were evaluated. 11 patients were transplanted before the full effect of TKI was waited for, in the era of transition from up front transplantation to TKIs treatment only, 8 patients were diagnosed in advanced phase of the disease, 5 patients showed severe, otherwise not resolvable TKIs toxicity and 18 patients progressed on TKIs treatment. Except of 2 patients transplanted in RIC all the others were prepared with full conditioning either with busulfan or TBI 12 Gy, using ATG in unrelated donors. The overall survival and incidence of relapse were evaluated considering the reason for transplantation indication, disease status and GVHD occurrence.

Results: The overall survival in 2 and 10 years was 75% resp.70%. There was naturally a significant difference in overall survival between up-front transplanted patients (who would not be transplanted at present) and the patients with advanced phase or not sufficient TKIs response, in 5 years 100% resp. 64% and in 10 years 92% resp. 62%. The survival of patients diagnosed in advanced phase (AP), who all received induction chemotherapy together with TKIs before entering the transplantation, in the comparison to the patients diagnosed in chronic phase (CP) but failing the TKIs response was in 2 years 88% resp. 65%, in 5 years 85% resp. 58% and in 10 years 63% resp. 58%. Statistically in these small numbers not significant. When comparing patients diagnosed in CP and staying still in CP before transplant (excluding up-front transplants) (CP1), those diagnosed in CP and progressing to AP during TKIs treatment (CP2) and those diagnosed as late as AP(AP), the statistically worst surviving group was the prior transplant accelerating one. (92%- CP1, 27%-CP2, 87% AP) in 2 years.

In the occurrence of aGVHD gr. 2-3 during the treatment the survival in one year was 76% vs 96% without the complication, p =0,07, on the contrary the occurrence of cGVHD improved the survival in 5 years to 83% vs 70% (not significant in this cohort).

Conclusions: The introduction of TKIs to the CML treatment has selected the patients with biologically poor character of the disease, for whom the allogeneic transplantation is the only curative approach. It seems that the patients diagnosed in the advanced phase but transplanted in chemotherapy and TKIs induced remission have better results than those primary resistant or who have lost the treatment response to TKIs. It looks like the graft versus leukaemia effect connected with the occurrence of cGVHD, known to be effective in CML in the pre-TKIs era, has a role also in these patients, although we do not statistically prove it in this small cohort.

Conflict of interest: Nothing to disclose

P140

Outcomes of Allogenic Hematopoietic Cell Transplantation for Childhood Chronic Myeloid Leukemia: Single Center Experience

Hanafy Hafez ^1,2 , Amr Abdalla ^1,2 , Mahmoud Hammad ^1,2 , Nayera Hamdy ^3,4 , Dina Yassin ⁴ , Sherin Ibrahim ⁴ , Lama El Halaby ⁵ , Alaa Elhaddad ^1,2

¹ National Cancer Institute, Cairo University, Pediatric Oncology and BMT, Cairo, Egypt; ² Children Cancer Hospital Egypt CCHE 57357, Pediatric Oncology and BMT, Cairo, Egypt; ³ National Cancer Institute, Cairo University, Clinical Pathology, Cairo, Egypt; ⁴ Children Cancer Hospital Egypt CCHE 57357, Clinical Pathology, Cairo, Egypt; ⁵ Children Cancer Hospital Egypt CCHE 57357, Research Department, Cairo, Egypt

Background: Despite the apparent efficacy and favorable toxicity profile of TKIs, allogeneic SCT remains the only curative treatment for CML especially in younger patients, but transplant related morbidity and mortality should be considered. We evaluated the clinical characteristics and outcomes of pediatric CML patients who had SCT in our center.

Methods: This retrospective study included pediatric patients (below 18 years) with confirmed CML, who received an allogeneic SCT at the stem cell transplantation unit, Children Cancer Hospital Egypt (CCHE 57357) between August 2007 and June 2017 with follow up till December 2017. All patients received Myeloablative conditioning chemotherapy containing Busulfan/Cyclophosphamide followed by stem cell infusion from matched related donors.

Results: From the whole 110 patients diagnosed with CML, 41 had available matched related donors and subjected to HSCT during the study period while 69 patients continued TKI therapy. The median age at diagnosis was 12 years and M/F ratio was 1.4:1. Thirty-seven patients were diagnosed with chronic phase, one case with accelerated disease and 3 patients with blastic crises and average duration before transplant was 12 months. Bone marrow was the stem cell source in 30 patients and peripheral blood was used in 11 cases and the average stem cell dose was 4.5x10⁶/kg. The 3-year overall survival and event free survival were 97% and 86% respectively, and the transplant related mortality at d100 and one year were 0% each. The incidence of chronic graft versus- host disease (cGVHD) was 82% in cases with peripheral blood source compared to 30% in cases who received bone marrow stem cells (P. value 0.005), also cGVHD was higher in F/M donor/recipient transplant (62%) as compared to M/F (46%) and M/M or F/F (35%) but not statistically significant.

Conclusions: Considering the good survival rates and very low TRM associated with HSCT in our pediatric CML patients, our data suggest offering transplant for all patients with available matched related donor, without need for prolonged exposure to TKIs. Bone marrow stem cell source should be preferred in those young patients to avoid a significant risk of cGVHD and its related complications.

Conflict of interest: All authors declare that nothing to disclose

P141

Patients with T315I mutation CML: characteristics and outcomes of the treatment

Julia Vlasova, Elena Morozova, Maria Barabanshchikova, Tatyana Gindina, Ildar Barhatov, Ivan Moiseev, Alexander Alyanskiy, Elena Darskaya, Boris Afanasyev

R.M. Gorbacheva Institute of Pediatric Oncology Hematology and Transplantation, I.P.Pavlov First St. Petersburg State Medical University, Hematology, St. Petersburg, Russian Federation

Background: Resistance to tyrosine kinase inhibitors (TKI) in patients with chronic myeloid leukemia (CML) is frequently caused by point mutations in the BCR-ABL kinase domain, including the gatekeeper mutant T315I, which confers a high degree of resistance to all currently approved tyrosine kinase inhibitors except ponatinib.

Methods: Retrospective analysis of 79 BCR-ABL^T315I -positive CML patients (pts) was done. Allogeneic bone marrow transplantation (allo-HSCT) was made in 20 pts, 59 pts received only pharmacological therapy (33 pts received TKI as monotherapy or in combination with other drugs other 26 pts received hydroxyurea, interferon- α or chemotherapy). At the time of T315I detection 51 (65%) pts were in CP, 22 (28%) pts had AP and 6 (7%) pts were in BC. Median (Me) age at the time of mutation detected was 42 years (13-75) (38 years in HSCT-group). In allo-HSCT group 12 (60%) pts had unrelated donors, 17 (85%) pts received more than 2 lines TKIs before HSCT, 2 pts were in BC at the time of HSCT, 5 pts were in AP, 8 pts were in CP≥2, 5 pts were in CP1. The number of points on EBMT scale: 3-4 points - 16(80%) pts, 5-7 points - 4(20%) pts. Conditioning regimen in 14 (70%) pts had reduced intensity. Me time to HSCT after T315I detection was 7 months (0-72). Mutation analysis was performed by Sanger sequencing. Overall survival (OS) was estimated by Kaplan-Meier method with log-rank test for comparison between groups. Cox regression was used for multivariate survival analysis that included next covariates: age, phase on the time of mutation detection, performance of allo-HSCT, time to T315I detection from TKI start.

Results: Me follow-up time after T315I detection was 21 months (1-100). 5-years OS in whole group was 42% (fig.1A). According to multivariate analysis only CML phase at the time of mutation detection significantly affect to survival in whole group. All pts in BC (n=5, 2 in HSCT group and 3 in non-HSCT group) died within first year after T315I indication wherein Me survival time was 1,3 month (fig.1B). 5-years OS in non-HSCT group (n=59) was 42% with Me survival time 2,8 years. 5-years OS after allo-HSCT (n=20) was 37% with Me survival time 5 months (fig.1C). All living patients after allo-HSCT are in deep molecular response. There was no significant difference in 5-years OS between TKI (n=17) and non-TKI (n=42) pharmacological therapy (non-HSCT) groups (42% and 47% respectively, p=0,53) (fig.1D).

Conclusions: Detection of T315I mutation in TKI-resistant patients is extremely unfavorable factor for survival, especially in the advanced phase CML, and it is a great reason for switching to ponatinib or other new potential investigated drugs if possible. Allo-HSCT can be a potential option for this group of patients in case of good selection taking into consideration transplant risk, especially for patients in CP ≥2.

Conflict of interest: Nothing to disclose

P142

Abstract previously published

P143

Secondary chronic neutrophilic leukemia (CNL) after severe aplastic anaemia (SAA) - a case report

Laurent Schmied ¹ , Pontus Lundberg ¹ , Alexander Tzankov ² , Martina Kleber ¹ , Jakob Passweg ¹ , Beatrice Drexler ¹

¹ University Hospital Basel, Hematology, Basel, Switzerland; ² University Hospital Basel, Pathology, Basel, Switzerland

Background: Chronic neutrophilic leukemia (CNL) and aplastic anemia (AA) are both rare diseases. Here we report the case of a 42 years old male patient with a medical history of 20 years of SAA, who developed a secondary CNL and subsequently received an allogeneic hematopoietic stem cell transplantation (HSCT). To the best of our knowledge a secondary CNL after SAA has not yet been described.

Methods: Case report

Results: After diagnosis of severe aplastic anemia at the age of 22 the subsequent treatment encompassed three cycles of horse ATG and cyclosporine resulting in disease control for nine, three and seven years, respectively. After the third cycle of intensive immunosuppressive therapy, cyclosporine was maintained as long-term therapy. Nineteen years after the initial diagnosis neutrophilia was observed without signs of infection, inflammation or solid tumor. Bone marrow examination revealed hypercellularity with neutrophilic proliferation and discrete dysplasia of the myeloid cell lineage. Molecular genetics detected a SETBP1, EZH2 and CSF3R mutation (T618I). Integrative diagnosis was made according to WHO classification confirming the diagnosis of CNL. Subsequently, the patient was first treated with Ruxolitinib, showing normalization of the neutrophil counts within 3 months, but treatment had to be stopped due to thrombocytopenia after 6 months. Then allogeneic hematopoietic stem cell transplantation from a 12/12 HLA identical unrelated donor after conditioning with Cyclophosphamid (60mg/Kg/day for two days) and Busulphan (4 x 0.8mg/Kg/day for four days) could be performed without relevant toxicity, complications or graft versus host disease (GvHD). The posttransplant follow up (3 months) shows complete remission without any signs of GvHD or other complications.

Conclusions: We describe the first case of CNL that has evolved secondary to SAA nineteen years after the initial diagnosis. Possible pathophysiological mechanisms could include damage of the hematopoetic stem cells by previous cytotoxic therapy or selection of a myeloid clone occurring in aplastic anemia.

Conflict of interest: The authors have no conflicts of interests to declare

[P143 Figure]

$$$$

P144

Validation and comparison of prognostic systems in secondary (post-PV/ET) myelofibrosis after allogeneic stem-cell transplantation. A report of the CMWP-EBMT

Nico Gagelmann ¹ , Dirk-Jan Eikema ² , Linda Koster ² , Christine Wolschke ¹ , Renate Arnold ³ , Lothar Kanz ⁴ , Grant McQuaker ⁵ , Ibrahim Yakoub-Agha ⁶ , Thierry Lamy ⁷ , Gerard Socié ⁸ , Jean Henri Bourhis ⁹ , Mohamad Mohty ¹⁰ , Jan J Cornelissen ¹¹ , Patrice Chevallier ¹² , Paolo Bernasconi ¹³ , Matthias Stelljes ¹⁴ , Pierre-Simon Rohrlich ¹⁵ , Renato Fanin ¹⁶ , Jürgen Finke ¹⁷ , Johan Maertens ¹⁸ , Didier Blaise ¹⁹ , Maija Itälä-Remes ²⁰ , Hélène Labussière-Wallet ²¹ , Marie Robin ⁸ , Yves Chalandon ²² , Nicolaus Kröger ¹

¹ University Medical Center Hamburg Eppendorf, Hamburg, Germany; ² EBMT Data Office, Leiden, Netherlands; ³ Charité Universitaetsmedizin Berlin, Berlin, Germany; ⁴ University Hospital, Tübingen, Germany; ⁵ Gartnaval General Hospital, Glasgow, United Kingdom; ⁶ Hospital CHU de Lille, LIRIC, INSERM U 995, Université de Lille, Lille, France; ⁷ University Hospital, Rennes, France; ⁸ Hopital St. Louis, Paris, France; ⁹ Gustave Roussy, Institut de Cancérologie, Villejuif, France; ¹⁰ Hospital Saint Antoine, Paris, France; ¹¹ Erasmus MC Cancer Institute, Rotterdam, Netherlands; ¹² CHU Nantes, Nantes, France; ¹³ Fondazione IRCCS Policlinico San Matteo, Pavia, Italy; ¹⁴ University of Münster, Münster, Germany; ¹⁵ CHU Nice - Hôpital de l`ARCHET I, Nice, Germany; ¹⁶ Azienda Ospedaliero Universitaria di Udine, Udine, Italy; ¹⁷ University of Freiburg, Freiburg, Germany; ¹⁸ UZ Leuven, Leuven, Belgium; ¹⁹ Institut Paoli Calmettes, Marseille, France; ²⁰ HUCH Comprehensive Cancer Center, Helsinki, Finland; ²¹ Centre Hospitalier Lyon Sud, Lyon, France; ²² Hôpitaux Universitaires de Genève, Geneva, Switzerland

Background: A recent model using clinical-molecular features of patients with secondary myelofibrosis (sMF) was developed by MYSEC to predict outcome from diagnosis. However, it is unclear whether scores, such as the MYSEC or DIPSS are prognostic in the transplant setting in sMF. We aimed to validate and to compare these models in sMF patients who underwent allogeneic hematopoietic stem-cell transplantation.

Methods: We identified 159 sMF patients who received stem-cell transplant from related (n = 59) or unrelated donors (n = 100) between 2007 and 2015 with available data on blood levels at transplant and presence of mutations at diagnosis. The MYSEC model was calculated as follows: one point was assigned to presence of constitutional symptoms and platelets < 150 x10⁹/L. Two points were assigned to hemoglobin < 11 g/dl, circulating blasts ≥3% and a CALR-unmutated genotype, whereas 0.15 points were assigned for each year of age. Risk groups (number of patients) according to MYSEC and DIPSS were: low (n = 27 and n = 16), intermediate-1 (n = 70 and n = 59), intermediate-2 (n = 40 and n = 70), and high (n = 22 and n = 14). Scores were validated using Kaplan-Meier estimates while C-statistics were applied to evaluate their discriminatory power.

Results: Median follow-up was 41 months (range, 29 to 52) while the median time between diagnosis and transplant was 128 months (range, 2 to 526). The median age of sMF patients was 52 years (range, 32 to 75). Overall survival at three years was 55.9% (95% confidence interval [CI], 47.5 to 64.3). By stratifying sMF according to evolution from either polycythemia vera (post-PV) or essential thrombocytopenia (post-ET), no difference was found in survival at three years being 57.7% (95% CI, 45.5 to 70.0) for post-PV and 54.6% (95% CI, 43.0 to 66.2; p = 0.92) for post-ET.

Overall survival rates at three years according to each risk group of the DIPSS were as follows: 79.5% (95% CI, 58.5 to 100) for the low-risk, 56.3% (95% CI, 42.6 to 70.0) for the intermediate-1-risk, 53.9% (95% CI, 41.2 to 66.6) for the intermediate-2-risk, and 40.8% (95% CI, 14.1 to 67.5) for the high-risk group. Overall, DIPSS was not predictive of outcome (p = 0.30).

Regarding MYSEC, probabilities of survival at three years were 69.3% (95% CI, 51.5 to 87.1) for the low-risk, 64.5% (95% CI, 52.5 to 76.5) for the intermediate-1-risk, 46.8 (95% CI, 29.7 to 63.9) for the intermediate-2-risk, and 21.7% (95% CI, 0 to 45.4) for the high-risk group. The MYSEC model was predictive of survival overall (p = 0.03).

When used to assign patients to the four discrete risk categories, the test retained moderate predictivitability (C-index = 0.585). However, prognostic ability was improved in comparison with DIPSS (C-index = 0.546). MYSEC significantly re-classified patients according to risk (p < 0.001).

Conclusions: In comparison to DIPSS, which is validated for primary myelofibrosis but currently also used for risk stratification in sMF patients receiving transplant, the clinical-molecular system by MYSEC provides significant re-classification of patients leading to improved prognostic capability.

Conflict of interest: No conflicts of interest related to the abstract

[P144 Figure]

$$$$

Conditioning regimen

P145

12 month CD4+ T-cell count and DLI treatment for mixed T-cell chimerism is associated with favorable outcomes in Reduced-Intensity Alemtuzumab-based transplant for low grade lymphoma

Claire Horgan, Shereef Elmoamly, Maria Kaparou, David Davies, Bhuvan Kishore, Manos Nikolousis, Richard Lovell, Shankara Paneesha, Katie Randall, Kathy Holder, Julie Suhr, Alex Kanellopoulos

Heart of England NHS Foundation Trust, Haematology, Birmingham, United Kingdom

Background: Upfront reduced intensity conditioned (RIC) allogeneic stem cell transplant (allo-SCT) may confer survival advantage for selected patients with heavily pre-treated low grade lymphoma, although data on Alemtuzumab use for graft versus host disease prophylaxis in this setting is limited. We conducted a retrospective study of patients with heavily pre-treated low-grade lymphoma undergoing RIC-Alemtuzumab allo-SCT at the Heart of England NHS Trust.

Methods: We analyzed 22 consecutive patients, median age 57.5 years, 9 female and 13 male, with a diagnosis of follicular lymphoma, marginal zone lymphoma, or waldenstrom's macroglobulinaemia who underwent Alemtuzumab-based RIC allo-SCT, 8 (36%) sibling-donor and the remaining 14 (64%) volunteer-unrelated donor (VUD), at Heart of England NHS Trust between January 2010 and October 2017. 10 patients were conditioned with FMC whereas 12 patients had BEAM-Campath. Total Alemtuzumab dose was 25mg (d-1) and 50mg (d-2, d-1) for sibling and unrelated donor transplants respectively.10 patients (45%) were in complete remission pre-transplant, with the remaining 12 patients (55%) being in partial remission. Patients had received a median of 3 (1-6) lines of treatment prior to transplant, and 2 of the 22 patients had previously received an autologous transplant.

Results: With a median follow up of 27.4 months, choice of conditioning did not seem to affect progression-free (PFS) (p=0.48), or overall survival (OS) (p= 0.91), albeit statistically significant improvements in OS for DLI (donor lymphocyte infusion)-treated patients were observed (p=0.001, figure 1). Overall, 7 patients (32%) underwent DLI, with the majority (71%) being for mixed chimerism at a median day of +196 post transplant. In terms of immune reconstitution, CD4+ T-cell count >200/µl at 12 months was associated with better OS (mean OS for CD4+ >200/µl patients at 12 months = 61 months vs 18 months for patients with CD4+ < 200/µl at 12 months, (p=0.017)). Besides, median OS for patients who had CD4+ T-cell count >200/µl at 12 months was 61.7 months. Acute GVHD affected 50% (11) patients, with 6 patients (27% of total) having grade 3 or 4, the remaining 5 (23% of total patients) having grade 1 or 2 disease. 4 individuals (18%) exhibited chronic GVHD, with only 1 of these having extensive disease. Interestingly, CD4+ T-cell count < 200/µl at 12 months did not have any association with either acute or chronic GVHD (p=0.56 and p=0.84 respectively). At 3 months, whole blood chimerism was not predictive of either PFS (p= 0.31) or OS (0.9), although there was a trend towards better outcomes for patients with mixed T-cell chimerism at 3 months (p=0.1). Finally, transplant related mortality was 23%.

Conclusions: In our series of Alemtuzumab based RIC allo-SCT for low grade lymphomas, three-month mixed T-cell chimerism necessitating DLI correlates with a favorable outcome. CD4+ T-cell immune reconstitution at 12 months seems to affect overall survival but CD4+ T-cell count < 200/µl at 12 months was not associated with acute or chronic GVHD. In summary, upfront RIC allo-SCT with Alemtuzumab is valid curative treatment for high risk low grade lymphoid malignancies, but prospective studies are needed to identify key determinants of its success.

Clinical Trial Registry: n/a

Conflict of interest: None of the listed authors has anything to disclose

[P145 Figure]

$$$$

P146

A case of a Myelodisplastic disease derived from a previously HLA-haploidentical transplanted marrow followed by a second HLA-haploidentical transplantation completely mismatched respect to the first

Matteo Parma ¹ , Elisabetta Terruzzi ¹ , Elisa Diral ¹ , Andrea Aroldi ¹ , Marilena Fedele ¹ , Pietro Enrico Pioltelli ¹ , Carlo Gambacorti-Passerini ²

¹ Ospedale San Gerardo, Monza, Italy; ² Ospedale San Gerardo, Clinica Ematologica dell`Universita Milano-Bicocca, Monza, Italy

Background: In recent years Haploidentical Bone Marrow Transplantation (Haplo-BMT) has been widely used in patients with high risk hematological malignancies who lack a matched related donor (MRD) and cannot obtain a matched unrelated donor (MUD) in a reasonable time. Here we describe a singular case of a patient who was submitted to Haplo-BMT for an Acute Myeloid Leukemia (AML) and who developed 5 years after a mielodysplastic syndrome. He was submitted to a second Haplo-BMT from a different donor, whose HLA was completely different respect to the first donor.

Methods: A 46-year-old male patient with undifferentiated AML was submitted to Haplo-BMT from his brother in first complete remission in 2011. The outcome was favorable, no major complications occurred, only a mild self-limiting chronic GvHD it was observed.

After 5 years of wellness, the patient presented worsening anaemia and a platelet decrease. A bone marrow aspirate was diagnostic for a three-lineage myelodisplastic syndrome, with 10% blast infiltration and a monosomal karyotype (7-). The chimerism status evaluated on bone marrow and peripheral mononuclear cells remained “full donor” on multiple controls. We concluded for a second haematologic neoplasm derived from the transplanted marrow. It is worth noting that the donor is actually fine.

Based on the good conditions the patient was submitted to a second BMT from a different donor (his son) who was HLA-Haploidentical respect to him but completely mismatched respect to the first donor. Considering the chimaerism status at the transplant, we were in the following situation: Haplo in GvHD direction, completely mismatched in reject direction but also in Graft versus Leukemia direction. In order to reduce the risk of rejection we performed an ATG (7,5 mg/Kg) priming 18 days before BMT. Anti HLA-antibody research revealed only a low title of anti-HLA of class-II. Conditioning regimen was a reduced-intensity Tiotepa-Busulfan-Fludarabine regimen, GvHD prophylaxis was performed with Cyclosporin-A, Micofenolate-Mofetil and post transplant Cyclophosphamide.

Results: No significant complication was observed during the procedure; platelet and neutrophil engraft occurred at day +28 and +32 respectively from BMT. Actually after 1 year the patient is in good conditions, no infections, viral reactivations, GvHD nor other complications appeared during the follow-up. He shows a complete and stable cytogenetic remission and chimaerism analysis reveals a 100% donor engraft either on marrow or on peripheral mono-nucleated cells since day +30.

Conclusions: Here we report our singular experience of a double Haplo-BMT with two completely HLA different donors performed for a secondary neoplasm occurred on the transplanted cells. In our knowledge there is not similar case reported on literature.

Conflict of interest: C. Giambacorti-Passerini: consultancy, honoraria and research founding from Pfizer; consultancy from BMS. The other authors have nothing to declare.

P147

A comparison of FLAMSA vs Thio/Bu/Flu (TBF) vs Flu/Treo (FT) conditioning for allogeneic transplantation in active acute myeloid leukemia. A study from the ALWP-EBMT

Francesco Saraceni ¹ , Myriam Labopin ² , Arne Brecht ³ , Nicolaus Kröger ⁴ , Matthias Eder ⁵ , Johanna Tischer ⁶ , Hélène Labussière-Wallet ⁷ , Herman Einsele ⁸ , Dietrich Beelen ⁹ , Donald Bunjes ¹⁰ , Dietger Niederwieser ¹¹ , Tilmann Bochtler ¹² , Bipin Savani ¹³ , Mohamad Mohty ¹⁴ , Arnon Nagler ¹⁵

¹ Ravenna Hospital, Ravenna, Italy; ² EBMT Paris Study Office - CEREST-TC, Saint Antoine Hospital, Department of Haematology, Paris, France; ³ Deutsche Klinik fuer Diagnostik, KMT Zentrum, Wiesbaden, Germany; ⁴ University Hospital Eppendorf, Bone Marrow Transplantation Centre, Hamburg, Germany; ⁵ Hannover Medical School, Department of Haematology, Hemostasis, Oncology and Stem Cell Transplantation, Hannover, Germany; ⁶ Klinikum Grosshadern Med. Klinik III, Munich, Germany; ⁷ Centre Hospitalier Lyon Sud, Pavillon Marcel Bérard -Bat 1G, Service Hematologie, Lyon, France; ⁸ Universitaetsklinikum Wuerzburg, Med. Klinik und Poliklinik II, Wuerzburg, Germany; ⁹ University Hospital, Department of Bone Marrow Transplantation, Essen, Germany; ¹⁰ Klinik fuer Innere Medzin III, Universitätsklinikum Ulm, Ulm, Germany; ¹¹ University Hospital Leipzig, Division of Haematology & Oncology, Leipzig, Germany; ¹² University of Heidelberg, Medizinische Klinik u. Poliklinik V, Heidelberg, Germany; ¹³ Vanderbilt University Medical Center, Nashville, TN, United States; ¹⁴ Saint Antoine Hospital, Department of Haematology, Paris, France; ¹⁵ Chaim Sheba Medical Center, Department of Bone Marrow Transplantation, Tel Hashomer, Israel

Background: The choice of conditioning regimen is particularly crucial in the setting of active acute myeloid leukemia (AML),as the need for a powerful anti-leukemic activity should not negate a good toxicity profile of the protocol.The sequential FLAMSA regimen (fludarabine-cytosine arabinoside-amsacrine combined with TBI or busulfan and cyclophosphamide),TBF (thiotepa-busulfan-fludarabine) and FT (fludarabine-treosulfan) are three possible options.

Methods: We analysed transplantation outcomes of 856 adult pts with AML in active disease phase (primary refractory or first or second relapse) that underwent a first SCT from matched sibling (MSD) or unrelated donor (UD) between 2005 and 2016 and reported to the ALWP of the EBMT.Six hundred and thirty-one pts received FLAMSA,112 received TBF,while 113 pts were transplanted following the FT protocol.According to institutional policy,FLAMSA was busulfan or TBI -based in 32% and 68% of the cases, respectively.Pts who received oral busulfan,manipulated grafts,or transplant from < 9/10 HLA-matched donor were excluded.

Results: Median follow-up was 53 months in FLAMSA and FT groups and 16 months in the TBF group.Median age at transplant was 52 years in FLAMSA and TBF and 58 years in the FT groups, respectively (p=0.001).At 100 days, cumulative incidence of grade II-IV aGVHD was 28%, 29% and 24% (p=0.7),while incidence of grade III-IV aGVHD was 11%, 12% and 10% for FLAMSA, TBF and FT, respectively (p=0.9).The 2-year cumulative incidence of cGVHD and severe cGVHD was 26%, 26%, 33% (p=0.4) and 11%, 19%, 13% (p=0.5) for FLAMSA, TBF and FT, respectively. Complete remissions cumulative incidence for pts that reached day 100 was 88%, 80%, 92% for FLAMSA, TBF and FT, respectively (p=0.13). No significant difference was observed between the three conditioning regimens in 2 year non-relapse mortality (NRM), being 20%, 24% and 26% in FLAMSA,TBF and FT, respectively (p=0.24) (Figure 1).Similarly, the 2-year relapse incidence (RI) rates were 53%, 54%, 46%, respectively (p=0.33).No significant difference was observed in 2 year leukemia-free survival (LFS),with rates of 27%, 22%, 29% for FLAMSA, TBF and FT, respectively (p=0.28). OS following TBF conditioning (24%) was comparable to FLAMSA (34%) and FT (37%, p=0.10).Finally, the composite endpoint graft-vs-host disease-free, relapse-free survival (GRFS) at 2 years was 20%, 13%, 23% for FLAMSA, TBF and FT, respectively (p=0.15).In multivariate analysis no significant difference was observed between the three preparative regimens in terms of transplant outcome.Karnofsky performance score (KPS) above 80% was independently associated with better OS (HR 0.7, p=0.01) and GRFS (HR 0.7, p=0.01).Importantly, in-vivo T-cell depletion (TCD) was associated with lower incidence of grade III-IV aGVHD (HR 0.4, p=0.018), chronic (HR 0.4, p=0.006) and severe cGVHD (HR 0.4, p=0.005).

Conclusions: These results suggest that the three preparative regimens namely FLAMSA, TBF and FT provide similar anti-leukemic effect in AML pts with active disease.Transplant related complications,TRM and outcome did not differ, as well. Karnofsky performance score emerged as a strong predictor of GRFS and overall survival. In-vivo TCD was associated with lower incidence of acute and chronic GVHD.In conclusion all three conditioning protocols are valid options for transplantation in pts with relapsed/refractory AML to be compared in future prospective studies.

Conflict of interest: [All authors]: nothing to declare

[P147 Figure]

$$$$

P148

Abstract previously published

P149

Alemtuzumab is feasible as substitute for ATG in the conditioning regimen FBM (Fludarabine, BCNU, Melphalan) - a single center study

Merle Weniger ¹ , Natalie Schub ¹ , Thomas Spanberger ¹ , Martin Gramatzki ¹ , Andreas Günther ^1,2

¹ Division of Stem Cell Transplantation and Immunotherapy, Medical Department 2, Kiel, Germany; ² Helios Kliniken Schwerin, Hematology, Oncology and Stem Cell Transplantation, Schwerin, Germany

Background: In recent years conditioning regimens with reduced or intermediate intensity were developed to extend the curative approach of allogenic stem cell transplantation to elderly and unfit patients. The combination of fludarabine, BCNU and melphalan (FBM)¹ has shown activity in acute myeloblastic leukemia (AML), myelodysplastic syndrome (MDS) and multiple myeloma (MM). However, this conditioning regimen can be combined with different methods of in vivo T cell depletion and immunosuppression. In our center two different approaches for T cell depletion are in use according to physician's discretion: the monoclonal CD52 antibody alemtuzumab or rabbit anti T-lymphocyte globulin (Neovii). This retrospective single center study analyzes the impact of the used in vivo T cell depletion on acute toxicity of the conditioning regimen as well as on the long term outcome.

1 Marks et al. Blood. 2008;112:415-25

Methods: All consecutive patients transplanted in our center between 2010 and 2016, where FBM was used as conditioning regimen, were included in the study. Most of the 163 patients were treated with ATG, while 36 received alemtuzumab. According to our institutional standard patients transplanted from a non-related HLA matched donor (NRMD, 9/10 or 10/10) were treated with ATG in a dose of 40 g/kg, while in cases with HLA matched related donors (MRD) the dose was reduced to 30 g/kg. In contrast, the equivalent dose for alemtuzumab calls for 30 mg (absolute) for transplants from NRMD and 20 mg in MRD transplantation. To address acute toxicity of the conditioning regimen clinical symptoms were recorded according to the common terminology criteria for adverse events. The endpoints for long term outcome were overall survival, rate of relapse and acute GvHD.

Results: The median age in both patient groups was comparable (alemtuzumab: 60 years, ATG: 58 years), but the patients treated with alemtuzumab had a significantly higher comorbidity score compared to the ATG cohort (median HCT-CI: 3 versus 1). All patients in the alemtuzumab group were diagnosed with AML or MDS, whereas ATG was also used in MM and ALL cases. Interestingly, the tolerability of alemtuzumab in regard of hypotension, tachycardia, rash, fever and chills was better than with ATG. Notably, with a median follow up of 584 months we observed no significant difference in overall survival in both groups (alemtuzumab: 56 %, ATG: 59%). The rate of acute GvHD was significantly lower in the alemtuzumab group (44% vs 68%, p=0.01). In AML and MDS patients the rate of relapse tended to be higher in alemtuzumab treated patients than in the ATG cohort (39 versus 29 %), although no statistical significance was achieved.

Conclusions: The conditioning regimen FBM showed impressive activity in our single center study with both drugs, ATG and alemtuzumab. The more tolerable drug was alemtuzumab, although the achieved level of immunosuppression seemed to be deeper at the chosen dose levels of both drugs. Since the outcome of both cohorts was comparable, alemtuzumab is found to be a feasible substitute for ATG in patients, where acute toxicity of the conditioning regimen should be avoided.

Conflict of interest: No

P150

Allogeneic stem cell transplantation for elderly patients (60-70 years of age): RIC, chronic GVHD and treatment response improve survival

Kari Remes, Suvi Jokela, Maija Itälä-Remes, Mervi Putkonen, Marjut Kauppila, Tommi Salmi, Soile Salomäki, Urpu Salmenniemi

Turku University Hospital, Dept of Clinical Haematology & Stem Cell Transplantation Unit, Turku, Finland

Background: Due to poor results with chemotherapy in elderly patients and improvements in the control of complications of allogeneic stem cell transplantation (ASCT) the use of ASCT has extended to patients over 60 years of age, and the upper age limit has increased to 70-75 years of age. The assessment of transplant outcomes is critical to test the feasibility of this treatment modality in the older age groups.

Methods: Our institutional transplant registry data were retrospectively analysed for patients aged 61-70 years, transplanted between 1.1.2010 and 30.6.2015 (N=59). The analysis is as of 30.6.2016 when a minimum follow-up of one year was fulfilled. The primary end point is overall survival (OS). Kaplan-Meier curves were produced, and Cox regression model was used to identify factors affecting OS.

Results: The median age is 63 (range: 61-69) years; 17 (29 %) were in the age group 66-70 years. 66 % of patients had AML/MDS, 21% lymphatic/plasma cell malignancies, 8 % other myeloid malignancies, and 5 % aplastic anemia. The disease status at transplant was CR in half of the patients, and the Gratwohl scoring identified an early disease in 46 %, intermediate in 15 % and advanced in 39 % of the patients.

The donor was unrelated in 80 %; source of stem cells was blood in 85 % and marrow in 15 %. RIC conditioning was used for 59 % and MA for 41 % of patients. HLA-matches were full in 29 %, DPB1-mm in 35 %, and HLA-A, C or DRB3 mm in 12 %. Severe acute GVHD (Gr3-4) occurred in 18 (30 %), and extensive chronic GVHD in 7 (12.5 %) patients. There is no significant difference in OS for patients aged 61-65 vs 66-70 years even if the survival curve seems better for those of 61-65 years (Fig). In univariate analysis significant factors for OS were conditioning (RIC > MA; P=0.001), occurrence of chronic GVHD (yes > no; P=0.014) and best response to ASCT (CR > less than CR; P=0.037).

Conclusions: ASCT is a feasible treatment modality for patients aged 61-70 years with relatively good OS. Patients over 65 years may have a worse outcome but their small number excludes meaningful statistics. However, in univariate analysis the age seems not to be a significant factor for OS in opposite to intensity of conditioning, chronic GVHD, and best response to ASCT.

Conflict of interest: All authors: nothing to disclose.

[P150 Figure]

$$$$

P151

BCNU-EAM vs non-BCNU alternative alkylator -EAM conditioning regimens for autologous stem cell transplantation in lymphoma patients

Deniz Goren Sahin ^1,2 , Guldane Seval Cengiz ³ , Ender Soydan ⁴ , Emine Tulay Ozcelik ^1,2 , Ayla Gokmen ⁴ , Zafer Gokgoz ⁴ , Pervin Topçuoğlu ³ , Onder Arslan ³ , Gunhan Gurman ³ , Muhit Ozcan ³ , Mutlu Arat ²

¹ Istanbul Bilim University, Hematology Department, Istanbul, Turkey; ² Sisli Florence Nightingale Hospital, Hematopoietic Stem Cell Transplantation Unit, Istanbul, Turkey; ³ Ankara University School of Medicine, Ankara, Turkey; ⁴ Medicana Ankara International Hospital, Ankara, Turkey

Background: Autologous hematopoietic stem cell transplantation (HSCT) is a potentially curative therapeutic approach for relapsed/refractory lymphomas. Although carmustine, etoposide, cytarabine, and melphalan (B-EAM) is the most commonly used conditioning regimen for high-dose therapy (HDT) before auto-HSCT, to date, there is no published prospective randomized study comparing different types of conditioning regimen. The cost and availability of BCNU forced many centres to search for alternatives. Therefore, gold standard regimens have not been established and clinical rehearsal differs substantially among institutions. Here we aim to compare between B-EAM and alternative alkylator -EAM backbone based regimen on the potential toxicities and survival in lymphoma patients.

Methods: Patients with relapsed refractory lymphoma, who underwent autologous HSCT between 2012-2017, were evaluated from three different highly active transplant institutions in Turkey. One hundred eighty adult lymphoma patients (age: 18-65 years) with a follow up time at least 6 months were included. Patients with B-EAM (n=139) high dose chemotherapy were compared with patients having received non-BEAM regimen; T- thiotepa (T) -EAM (n= 16), Bendamustine (Be)-EAM (n=12), Lomustine (L) -EAM (n=13) using the following matching factors: age, gender, lymphoma subtype, disease status at transplant, toxicity, overall survival, and causes of death.

Results: Totally 180 patients (109 males, 71 females) were identified: 139 patients and 41 patients were treated with BEAM and non-BEAM regimen, respectively. Patients' baseline characteristics are detailed in Table 1.

There was no statistically significant difference between mean age and gender between groups. With regards to toxicities among groups; BEAM group had a significantly lower rate of mucositis and gastrointestinal toxicity than non-BEAM group (p=0.01 and p=0.002, respectively). Renal toxicity did not show difference among groups (p=0.92). Although BEAM group showed a lower rate of transplant related mortality at 100^th day (5.5% vs. 11.4%), the difference did not reach to statistical significance. Estimated mean survival time for BEAM and Non-BEAM was 52.1±2.5 months and 18.8±1.5 months, respectively (p=0.2, log-rank test) (Figure 1).

The significant difference in median follow/up times has to be taken into account.

Conclusions: Our study showed that EAM backbone (non-BEAM) conditioning regimens with alternative alkylating agents have similar results compared to B-EAM regimen in terms of efficacy, and toxicity. There is still need for an optimal conditioning regimen with lower toxicity, better efficacy and less cost than BEAM. Prospective randomized controlled studies with large number of patients are warranted to find out the optimal backbone combination, which may establish as a standard regimen for HSCT in lymphoma patients.

Conflict of interest: Nothing to disclose

Patients	B-EAM	Non-BEAM	p-value
Median follow-up (months)	29.2	6.5	<0.0001
Age at HSCT, median (SD) (years)	46.6 (±15.1)	47.7 (±14.3)	0.67
Gender male (n) %	87 (62.6%)	22 (53.7%)	0.36
Entities (n) (%) NHL/HL	94 (67.6%)/45 (32.4%)	30 (73.2%)/11 (26.8%)	0.56
Disease status at transplant, n (%) CR1/CR2/PR/Refractory	20 (15.9%)/24 (19.0%)/53 (42.1%)/29 (23.0%)	6 (14.6%)/6 (14.6%)/20 (48.8%)/9 (22.0%)	0.74
Mucositis, n (%)	69 (49.6%)	31 (75.6%)	0.01
Renal, n (%)	6 (4.3%)	2 (4.8%)	0.92
Gastrointestinal system, n (%)	71 (51.0%)	32 (78.0%)	0.002
Transplant related mortality at 100th day Yes (n) (%)	6 (5.5%)	4 (11.4 %)	0.25

[[P151 Table] Table 1]

[P151 Figure]

$$$$

P152

BEAM AND FEAM CONDITIONING FOR HODGKIN DISEASE: A RETROSPECTIVE MULTICENTER STUDY FROM FONDAZIONE ITALIANA LINFOMI (FIL)

Matteo Pelosini ¹ , Jacopo Olivieri ² , Federico Mosna ³ , Angelo Fama ⁴ , Simone Ferrero ⁵ , Giuseppe Carli ⁶ , Margherita Giannoccaro ⁷ , Nicola Sgherza ⁸ , Sara Rattotti ⁹ , Dario Marino ¹⁰ , Giacomo Loseto ¹¹ , Anna Maria Mazzone ¹² , Teresa Calimeri ¹³ , Maria Chiara Tisi ^6,14 , Gabriella Tomei ¹⁵ , P Chiusolo ¹⁴ , G Perali ² , A. Billio ³ , Enrico Orciuolo ¹⁶ , K Codeluppi ¹⁷ , P.M. Stefani ¹⁸ , P. Ghione ¹⁹ , F. Gherlinzoni ¹⁸ , Attilio Olivieri ²⁰ , I. Federici ²⁰ , R Centurioni ²¹ , R. Matera ⁷ , S Finotto ²² , N Cascavilla ⁸ , Renato Fanin ² , Francesco Zaja ²

¹ Azienda Ospedaliera Universitaria Pisana, U.O. Ematologia, Pisa, Italy; ² A.O.U. Santa Maria della Misericordia di Udine, Ematologia, Udine, Italy; ³ Ospedale Regionale 'S. Maurizio', Ematologia e Centro Trapianto di Midollo Osseo, Bolzano, Italy; ⁴ IRCCS, Ematologia ASMN, Reggio Emilia, Italy; ⁵ AOU Città della Salute e della Scienza di Torino, • Ematologia 1, Dipartimento di Biotecnologie Molecolari e Scienze per la Salute, Università di Torino, Torino, Italy; ⁶ Ospedale San Bortolo, Ematologia, Vicenza, Italy; ⁷ ASL Le/1 P.O. Vito Fazzi, Ematologia, Lecce, Italy; ⁸ IRCCS Ospedale Casa Sollievo Sofferenza, Ematologia, San Giovanni Rotondo, Italy; ⁹ Fondazione IRCCS Policlinico San Matteo, Ematologia, Pavia, Italy; ¹⁰ IOV, IRCCS, UOC di Oncologia Medica 1, Padova, Italy; ¹¹ IRCCS - Ospedale Oncologico, Ematologia, Bari, Italy; ¹² Ospedale 'S. Giuseppe Moscati, Ematologia, Taranto, Italy; ¹³ IRCCS Ospedale San Raffaele, Divisione di Ematologia, Milano, Italy; ¹⁴ Università Cattolica del Sacro Cuore, Istituto di Ematologia, Roma, Italy; ¹⁵ Ospedale di Ivrea-ASL TO4, Ematologia, Ivrea, Italy; ¹⁶ Azienda Ospedaliera Universitaria Pisana, U.O. Ematologia Universitaria, Pisa, Italy; ¹⁷ ASMN - IRCCS, Ematologia, Reggio Emilia, Italy; ¹⁸ Azienda ULSS 9, Ematologia, Treviso, Italy; ¹⁹ AOU Città della Salute e della Scienza di Torino, Ematologia 1, Dipartimento di Biotecnologie Molecolari e Scienze per la Salute, Università di Torino, Torino, Italy; ²⁰ Azienda Ospedali Riuniti di Ancona, Ematologia, Ancona, Italy; ²¹ ASUR - AV3, UOC Medicina Interna ed Ematologia, Civitanova Marche, Italy; ²² IOV, IRCCS, UOC di Oncologia Medica 1, I, Padova, Italy

Background: High dose chemotherapy with stem cell rescue is the standard of care for relapse refractory Hodgkin disease (HD) patients even if newer treatment are nowadays available. The most used conditioning regimen is the BEAM scheme. In last years carmustine shortage has been an issue worldwide and in Italy as well. Fotoemustine emerged as a possible alternative with the FEAM scheme.

Methods: We retrospectively analyzed data of 1038 consecutive patients transplanted in 16 different Italian centers from 2008 to 2015 with BEAM and FEAM regimen. Here we report data about HD patients (n =278).

Results: Median age was 39 for both groups. There were 42,7% female patients in BEAM and 45,6% in FEAM. Patient characteristics at diagnosis were comparable with stage II (BEAM 36% vs 41,6% FEAM), stage III (BEAM 28% vs 23,8 % FEAM) and stage IV (33,5 % BEAM vs 32% FEAM). B symptoms were present respectively in 48,2% (BEAM) and 56,6% (FEAM). First line treatment was ABVD in 92% in both groups. Radiotherapy before transplant (as consolidation or in 2nd line) was used in 36,6% case in BEAM cohort vs 14 % in FEAM. Disease assessment before SCT showed in the BEAM CR in 63%, PR 26,8% and resistant disease (RD) in 9,1%, metabolic CR (PET negative at transplant) in 59,8% patients. In the FEAM group CR 53,5%, PR 34,1%, RD 11,%. metabolic CR in 47,3 %. BEAM patients had an higher Sorror score at transplant (Sorror >3 14,6 % vs 2,63). Neutrophil engraftment was similar in both groups (10 days). Grade III-IV mucositis occurred in 28% of BEAM and 34 % of FEAM. Febrile neutropenia grade III-IV were observed respectively in 46% (BEAM) and 57 % (FEAM). Response status did not differ at day 100 post ASCT (CR 85% BEAM vs 75,% FEAM) and neither at last follow-up (continuous CR 67,7 % BEAM vs 62,2% FEAM). Non-relapse mortality at 1 year (BEAM 0,6% vs FEAM 2,6%) was comparable. Overall survival (OS) at 2 years was not different (BEAM 92 % vs FEAM 88,6%, p=0.16). Death occurred respectively in 16,4% and 16,6 % for BEAM and FEAM due to disease progression in the majority of patients. However, the FEAM group had a trend to and inferior progression-free survival at 1 year (BEAM 74,6 % vs FEAM 65%,4 p=0.075) and higher relapse incidence (18,6 % BEAM vs 26,4% FEAM, p=0,051)

Conclusions: Substitution with Fotemustine did not increase general toxicities or infectious episodes; however we observed a sligher higher rate of gastrointestinal and infectious toxicities in FEAM patients that did not lead to higher NRM neither significantly worsened OS. Although the retrospective design of the study should caution against definitive conclusions we observed a trend to a better PFS and relapse incidence in BEAM patients, that could be related a better disease control status at the time of transplant.

Conflict of interest: None

P153

BEAM versus LEAM conditioning chemotherapy for autologous transplantation in lymphoma . A retrospective study from the BSBMT

Nick Morley ¹ , John Snowden ¹ , Rachel Pearce ² , Julia Perry ² , Ruth Paul ² , Keiren Kirkland ² , Jonathan Lambert ³ , Ram Malladi ⁴ , Andy Peniket ⁵

¹ Sheffield Teaching Hospitals NHS Foundation Trust, Sheffield, United Kingdom; ² British Society of Blood and Marrow Transplantation, London, United Kingdom; ³ University College London Hospitals NHS Foundation Trust, London, United Kingdom; ⁴ University Hospitals Birmingham NHS Foundation Trust, Birmingham, United Kingdom; ⁵ Oxford University Hospitals NHS Foundation Trust, Oxford, United Kingdom

Background: Since Linch et. al. published the results of a BNLI trial comparing BEAM with autologous rescue with mini-BEAM in the Lancet in 1993, BEAM has been considered the 'gold standard' conditioning regimen for patients with relapsed Hodgkin lymphoma in the UK. Not long afterwards the Parma Study also reported significant benefit from the Carmustine containing high dose chemotherapy (the BEAC regimen) for patients with Non-Hodgkin Lymphoma. Over recent years there have been intermittent shortages of the drug Carmustine leading to increasing use of the drug Lomustine as a substitute.

Methods: In order to compare the safety and efficacy of BEAM with LEAM conditioning chemotherapy we conducted a retrospective registry survey of patients with Hodgkin and Non-Hodgkin Lymphoma for first autologous transplant in the UK with a Day 0 between January 2009 and December 2013 and at least 100 days follow up available. Data to be sourced through Med A/B forms. Overall outcome to be assessed through Overall Survival (OS). Efficacy through progression free survival (PFS) and relapse rate (RR) and Safety through non-relapse mortality (NRM).

Results: Overall 1060 patients were identified from 30 centres with a median follow up of nearly 4 years. 650 (61%) patients had received BEAM and 410 (39%) LEAM conditioning. Both groups were well matched for year of transplant, gender, age, diagnosis, presence of comorbidity, lines of prior therapy, disease status at transplant and performance status. Median time to neutrophil (Median 11 days) and platelet (Median 14-15 days) engraftment were equivalent, as was disease response to transplant (CR 74-78%). Assessing efficacy, PFS was equivalent at 1 year (86% v 91%) and 5 years (51% v 55%), p=0.12, as was RR at both 1 year (20% v 16%) and 5 years (41% v 39%), p=0.246. In terms of safety, NRM was equivalent at 100 days (3% v 2%) and 5 years (6% v 5%), p=0.605. OS was also equivalent at 1 year (86% v 91%) and 5 years (64% v 66%), p=0.121.

Conclusions: Comparison of BEAM with LEAM conditioning chemotherapy for first autologous transplant in patients with Hodgkin and Non-Hodgkin Lymphoma showed equivalent safety, efficacy and overall survival. Lomustine can be safely substituted in place of Carmustine chemotherapy.

Clinical Trial Registry: BSBMT CTCR 15-01

Conflict of interest:

J. Snowden: Honoraria for speaking/chairing Sanofi and Jazz

N. Morley: Conference support and advisory board fees, Roche

[P153 Figure]

$$$$

P154

BEAM versus single agent High Dose Melphalan (HDM) conditioning regimen for autologous hematopoietic stem cell transplant (ASCT): a retrospective matched analysis in relapse/refractory Hodgkin lymphoma

M S Rauf ¹ , Panayotis Kaloyannidis ² , Irfan Maghfoor ¹ , Solaf Kafnar ² , John Apostolidis ² , Khalid Al Anezi ² , Harbi Salman ² , Shaibani Eshrak ² , Jenifer Bacal ² , Tusneem A.M. Elhassan ¹ , Saad Akhtar ¹ , Hani Al Hashmi ²

¹ King Faisal Hospital and Research Center, Oncology Center, Riyadh, Saudi Arabia; ² King Fahad Specialist Hospital, Adult Hematology and Stem Cell Transplantation Department, Dammam, Saudi Arabia

Background: The ideal conditioning regimen still remains a challenge for patients undergoing autologous stem cell transplantation (ASCT) for relapsed/refractory Hodgkin Lymphoma (RR-HL). The commonly used regimen is BEAM but single agent high-dose Melphalan (HDM) has also been used, however so far, there are limited experience and data comparing BEAM vs. HDM.

Methods: Medical records of 112 RR-HL patients, from two different institutions, who evaluated with chemosensitive disease after salvage chemotherapy and underwent ASCT from 2008 onwards, were reviewed retrospectively after Institutional Review Board Approval. A relatively smaller cohort of patients conditioned with HDM (n=28) was compared in a matched-paired analysis (1:3) with a larger cohort of patients who received BEAM (n=84). The patient groups (BEAM vs. HDM) had similar median age (30ys) and sex (M:F 1.7:1 vs. 1.8:1) and were matched for disease status before salvage (late relapse: 36 vs. 12 and early relapse/primary refractory: 48 vs.16) and disease status pre-ASCT [complete remission (CR): 39 vs. 13 and partial remission (PR): 45 vs.15). BEAM-regimen was given in the standard infusion instructions and doses over 6 days, while HDM (200mg/m²) was given in a single day infusion. All patients received prophylaxis against microbial, fungal and viral infections; GCSF was routinely administered at the dose of 5mcg/kg at +1 day (BEAM-group) and at +5 day (HDM-group). Kaplan-Meier method was used to calculate overall survival (OS) and progression free survival (PFS).

Results: The engraftment was successful; the median day for neutrophils >500/mm3 was +11 for both groups whilst for platelets >20000/mm3 a faster recovery was noticed for HDM-group: +13 vs. +22 days (p< 0.001). We had previously reported (Leuk Lymphoma.2008;49:769-78) platelet recovery at 15 days with BEAM, now 22 days because of logistic reasons and extra platelet transfusions for inpatient perma-catheter removal.​ After a median follow-up of approximately 2 years, 64 patients are alive (59 disease-free) in BEAM-group while 24 (19 disease-free) in the HDM-group. For the whole cohort of patients the 5ys-OS was 68% vs. 80% and 5ys-PFS was 52% vs. 65% for the BEAM and HDM-group respectively. The 100-day transplant related mortality was similar and acceptable for both groups: 2/84 (2.3%) in BEAM-group vs.1/28 (3.5%) in the HDM-group (MERS-CoV outbreak). From 1998-2015, 13/334 (3.8%) patients who received BEAM at KFSHRC died of TRM.

Conclusions: This study, though retrospective, demonstrates that for RR-HL patients, the conditioning regimen consisting of single-dose HDM may have similar efficacy as compared to the BEAM-regimen. Importantly, difference in the duration of chemotherapy administration (6 days for BEAM vs. 1 day for HDM), results in a shorter hospitalization period, and the shorter period of GCSF administration post-ASCT, may contribute to a better-cost effectiveness for the HDM-regimen. However, additional studies with larger cohort and longer follow-up are needed. Long term disease outcome, toxicity and meticulous comparative cost analysis are needed.

Conflict of interest: Nothing to disclose

[P154 Figure]

$$$$

P155

BEAM vs FEAM: a retrospective comparative study of conditioning regimens in lymphoma patients undergoing autologous stem cell transplant

Francesco Marchesi ¹ , Saveria Capria ² , Diana Giannerelli ³ , Silvia Maria Trisolini ² , Michela Ansuinelli ² , Maria Denise Caputo ² , Alessandra Serrao ² , Svitlana Gumenyuk ¹ , Daniela Renzi ¹ , Livio Pupo ⁴ , Francesca Palombi ¹ , Ida Provenzano ⁴ , Alice Di Rocco ² , Francesco Pisani ¹ , Atelda Romano ¹ , Antonio Spadea ¹ , Elena Papa ¹ , Marco Canfora ⁵ , Maria Cantonetti ⁴ , Andrea Mengarelli ¹

¹ Regina Elena National Cancer Institute, Hematology and Stem Cell Transplant Unit, Rome, Italy; ² University La Sapienza, Department of Cellular Biotechnologies and Hematology, Rome, Italy; ³ Regina Elena National Cancer Institute, Biostatistical Unit, Rome, Italy; ⁴ Tor Vergata University, Hematology, Rome, Italy; ⁵ Regina Elena National Cancer Institute, Scientific Direction, Rome, Italy

Background: BEAM is the standard conditioning for lymphoma patients undergoing ASCT. However, in Italy, BCNU is often substituted by Fotemustine (FEAM) due to difficult supplying. Comparison studies are lacking.

Methods: From Jan-2007 to Jan-2017, 362 consecutive lymphoma patients undergoing ASCT in two JACIE accredited Institutions were reviewed. Selection for BEAM (200) or FEAM (162) was based on BCNU availability. Median age was 48 years (13-69). ASCT was performed for aggressive NHL in 225 cases (62%), indolent NHL in 37 (10%) and HL in 100 (28%), and after first line in 72 cases (20%), after first salvage treatment in 214 (59%) and in more advanced phase in the remaining 76 (21%). Pre-ASCT response status was CR in 64% (first CR 11%, ≥ second CR 53%), PR in 32% and refractory disease in 4% of cases. The two groups of patients were well-balanced for all analyzed parameters, expect for median follow-up (BEAM: 72 months, FEAM: 25 months; P< 0.0001).

Results: After FEAM, a significant higher incidence of grade 3-4 oral mucositis (52% vs 30%; P< 0.0001) and documented infections (46% vs 36%; P=0.05) were observed. Non-relapse mortality and pulmonary toxicities were similar in the two groups. Response status at day-100 post-ASCT was significantly better for BEAM group (CR: 87% vs 75%; P=0.005). This difference was evident both in DLBCL (CR: 83% vs 64%; P=0.008) and HL patients (CR: 87% vs 76%; P=0.13). At multivariate analysis, conditioning was the only independent factor able to affect post-transplant CR (BEAM vs FEAM ORs: 2.26, 95%CI: 1.28-3.99; P=0.005), together with pre-transplant disease status (CR vs no CR ORs: 4.72, 95%CI: 2.66-8.35; P< 0.0001). PFS and OS were not significantly different (2yPFS BEAM: 76% vs FEAM: 74%; P=0.211; 2yOS BEAM: 87% vs FEAM: 84%; P=0.578).

Conclusions: Despite limitations due to the non-randomized nature of this study, BEAM conditioning regimen seems better tolerated. Although PFS and OS are similar at this time, considering the higher proportion of CR post-ASCT and the longer median follow-up, BEAM should remain the regimen of choice in these patients. A longer follow-up of FEAM patients will allow to evaluate whether the survival curves will differ in the next future.

Conflict of interest: None of the authors has any conflicts of interest to report.

P156

BeEAC conditioning regimen for autologous hematopoietic stem-cell transplantation in malignant lymphoma patients

Yulia Dubinina, Vladislav Sarzhevskiy, Anastasiya Samoylova, Nikita Mochkin, Vladimir Melnichenko, Dina Kolesnikova, Vladimir Bogatyrev, Elena Smirnova, Diana Ivanova, Anna Bannikova

State Budgetary Healthcare Institution, National Medical Surgical Center N.A. N.I. Pirogov, Ministry of Health of Russia, Moscow, Russian Federation

Background: Nowadays there are no randomized trials for comparison the effectiveness and tolerability of different conditioning regimens in malignant lymphoma (ML) patients.

Aim of the study is to assess the safety and toxicity profile, overall (OS) and progressive-free survival (PFS) in patients who received BeEAC, the novel conditioning regimen for autoHSCT in primary-refractory and relapsed ML patients (ClinicalTrials.gov NCT03315520).

Methods: Between January 2016 and August 2017 93 patients received autoHSCT: Hodgkin lymphoma (HL) - 60, non-Hodgkin lymphoma - 33 patients; 42 women and 51 men. Median follow-up was 12 months. The median age was 33 years (18-65). Before autoHSCT 43 patients (45,3%) had CR, 47 (50,5%) - PR, 3 patients (3,2%) were in a stable disease (SD). Median number of chemotherapy lines was 3 (1-9). The conditioning regimen BeEAC: increasing doses of bendamustine 160 mg/m*2, 180 mg/m*2, 200 mg/m*2 D-6, D-5, combined with fixed-dose cytarabine 400 mg/м2 D-4 - D-1; etoposide 400 mg/м2 D-4 - D-1; cyclophosphamide 140 mg/м2 totally, divided in 4 days (D-4 - D-1). In a phase I study patients were divided into 3 cohorts with 3 patients in each, where bendamustine dose increased from 160 to 200 mg/m*2 D-6D-5.

Results: A Phase I study showed no dose limited toxicity was described. The following patients (84) received bendamustine 200 mg/m*2.

Hematological toxicity is presented in table 1.

Cardiotoxic effects were detected in 7 patients (7.5 %) and were following: hydropericardium, dry pericarditis, post-cytostatic cardiomyopathy with acute biventricular heart failure and paroxysmal atrial fibrillation. 41 (44%) patients had no signs of oral mucositis (OM, WHO classification), 22 (23.65%) had grade 1 OM, 24 (25.8%) - grade 2, 5 (5.37%) - grade 3 and 1 (1.1%) patient had grade 4. Intestinal mucositis (NCI-CTC) was determined in 35 patients: grade 1 - 16 (17.2%) patients, grade 2 - 6 (6.4%), grade 3 - 13 (14%).

22 (23.65%) patients developed liver toxicity (CTCAE 4.03), but in majority of cases it was grade 1.

5 (5.37%) patients developed renal toxicity (CTCAE 4.03), 1 patient grade 2.

The CR rate increased from 45.3% to 59% during 100 days after HSCT. 18 (20%) patients remained in PR, 1 (1.05%) - SD.

37 patients relapsed or progressed after HSCT. The frequency of relapse depended on the number of chemotherapy lines (CL). 3 (8.1%) patients had 1 CL, 11 (29.7%) - 2 lines, 11 (29.7%) - 3 lines and 12 (32.4%) patients had 4 CL and even more. 2 patients (2,1%) with NHL B-cell lymphoma relapsed in CNS.

Overall mortality was 12,9% (12 patients). 4 of them died from progressive disease, 3 - from cardiotoxicity, 1 - from sepsis, 2 - from bleeding, 1 - from pulmonary edema during radiation therapy and 1 from postoperative complications not-related with ML. Between them 2 patients (2,1%) died in the early post-transplant period (before D+30): 1 from sepsis and 1 from acute cardiotoxicity.

Median PFS and OS at 1 year were 54% and 86%, respectively

Conclusions: Data analysis showed that BeEAC is a safety conditioning regimen. Further investigation needs for assessment of long-term survival.

Clinical Trial Registry: ClinicalTrials.gov NCT03315520

Conflict of interest: No disclosure

[P156 Figure]

$$$$

P157

Bendamustine as an alternative to BCNU as conditioning regimen for autologous stem cell transplantation in patients with aggressive lymphoma: Results phase 2 study from GELTAMO

Alba María Redondo ¹ , D Valcárcel ² , A.P González ³ , M Suárez-Lledó ⁴ , J.L Bello ⁵ , M Canales ⁶ , J Gayoso ⁷ , E Conde ⁸ , I Jarque ⁹ , R Del Campo ¹⁰ , R Arranz ¹¹ , M.J Terol ¹² , J.J Rifón ¹³ , M.J Rodríguez ¹⁴ , M.J Ramírez ¹⁵ , N Castro ¹⁶ , A Sánchez ¹⁷ , J López-Jiménez ¹⁸ , J Briones ¹⁹ , A López ²⁰ , L Palomera ²¹ , M Baile ¹ , Dolores Caballero ¹ , Alejandro Martín ¹

¹ Hospital Universitario de Salamanca/IBSAL, Hematology, Salamanca, Spain; ² Hospital Vall d'Hebron, Hematology, Barcelona, Spain; ³ Hospital Central de Asturias, Hematology, Oviedo, Spain; ⁴ Hospital Clinic, Hematology, Barcelona, Spain; ⁵ Complejo Hospitalario Universitario de Santiago de Compostela, Hematology, Santiago de Compostela, Spain; ⁶ Hospital Universitario La Paz, Hematology, Madrid, Spain; ⁷ Hospital General Universitario Gregorio Maranon, Hematology, Madrid, Spain; ⁸ Hospital Universitario Marqués de Valdecilla, Hematology, Santander, Spain; ⁹ Hospital Universitario la Fe, Hematology, Valencia, Spain; ¹⁰ Hospital Son LLátzer, Hematology, Palma, Spain; ¹¹ Hospital la Princesa, Hematology, Madrid, Spain; ¹² Hospital Universitario de Valencia, Hematology, Valencia, Spain; ¹³ Clínica Universidad de Navarra, Hematology, Pamplona, Spain; ¹⁴ Hospital Universitario de Canarias, Hematology, San Cristóbal de La Laguna, Spain; ¹⁵ Hospital de Jeréz, Hematology, Jerez de la Frontera, Spain; ¹⁶ Hospital 12 de Octubre, Hematology, Madrid, Spain; ¹⁷ Hospital Virgen de la Arreixaca, Hematology, Murcia, Spain; ¹⁸ Hospital Ramón y Cajal, Hematology, Madrid, Spain; ¹⁹ Hospital Santa Creu I Sant Pau, Hematology, Barcelona, Spain; ²⁰ Hospital Arnau de Villanoba, Hematology, Valencia, Spain; ²¹ Hospital Clínico de Zaragoza, Hematology, Zaragoza, Spain

Background: BEAM is the most employed conditioning regimen for lymphoma patients in Europe. However, BCNU is hardly available and it has been related to pulmonary complications. We conducted a phase 2 trial to evaluate the safety and efficacy of bendamustine instead of BCNU in the BEAM regimen as conditioning for autologous stem-cell transplantation (ASCT) in patients with aggressive lymphomas.

Methods: Inclusion criteria were: histologic diagnosis of i) relapsed or refractory diffuse large B-cell lymphoma (DLBCL) or grade 3B follicular lymphoma (FL) in partial response (PR) or complete remission (CR) after salvage therapy, or ii) transformed DLBCL or peripheral T-cell lymphoma (PTCL) in first or subsequent PR or CR. Conditioning regimen consisted of bendamustine (200 mg/m², days -7 and -6), etoposide (200 mg/m², days -5 to -2), cytarabine (400 mg/m², days -5 to -2), and melphalan (140 mg/m², day -1) (BendaEAM regimen). Primary endpoint was progression-free survival (PFS) at 3 years. Secondary endpoints were toxicity, response to transplant, and overall survival (OS).

Results: Sixty patients (median age 55 [28-71] years, 50% male) from 22 Spanish hospitals were included from May 2011 to November 2012. Histologies were as follows: 40 DLBCL, 3 grade 3B FL, 13 transformed DLBCL and 7 PTCL. 82% of patients had received two or more lines of treatment prior to ASCT. Thirty-seven patients (62%) were in metabolic CR (assessed by PET/CT) at the time of transplant and 23 (38%) in PR.

All patients (except one who died early) engrafted after a median of 11 (9-72) and 14 (4-53) days, respectively, to achieve >0.5 x 10⁹/l neutrophils and >20 x 10⁹/l platelets. A total of 39 serious adverse events were reported before day +100, including 14 infectious episodes, 2 of them resulting in respiratory failure and death of the patient (3.3% of transplant-related mortality), and 5 episodes of renal failure after bendamustine administration, reversible in all cases. Non-relapse mortality after day +100 was 3.3% (1 patient from Wernicke´s encephalopathy and 1 from infectious complications).

Regarding response to transplant, 45 patients (75%) achieved CR and 6 (10%) PR. With a median follow-up of 67 months (40-77), 24 patients had relapsed disease, 3 had secondary neoplasms (2 myelodysplastic syndrome and 1 cholangiocarcinoma) and 20 patients died. Estimated 5-y PFS and OS in the global series were 51% and 65%, respectively. Patients with PET+ disease at study entry had significantly worse PFS (26 vs 67% at 5 years, p=0.001) and OS (56 vs 75% at 5 years, p=0.036) than patients who underwent the ASCT in metabolic CR, and this was the only prognostic factor affecting both PFS (RR 0.27 [0.12-0.56]) and OS (RR 0.40 [0.17-0.97]) in the multivariate analysis.

Conclusions: BendaEAM conditioning is a safe, feasible and active regimen in patients with aggressive lymphomas. Infectious and renal toxicities should be carefully monitored. Our long-term results indicate that efficacy is similar to that previously reported with other regimens most commonly used like BEAM. PET status at the time of transplant is the key prognostic factor for survival.

Clinical Trial Registry: ClinicalTrials.gov Identifier: NCT01296256.

Conflict of interest: No conflicts of interest related to the present abstract in the last 12 months.

P158

Busulfan and melphalan versus high-dose melphalan as a conditioning regimen for autologous stem cell transplantation in patients with multiple myeloma

Sung-Hoon Jung ¹ , Je-Jung Lee ¹ , Jin Seok Kim ² , Chang-Ki Min ³ , Kihyun Kim ⁴ , Yunsuk Choi ⁵ , Hyeon-Seok Eom ⁶ , Young Don Joo ⁷ , Sung-Hyun Kim ⁸ , Jae-Yong Kwak ⁹ , Hye Jin Kang ¹⁰ , Jae Hoon Lee ¹¹ , Ho Sup Lee ¹² , Yeung-Chul Mun ¹³ , Joon Ho Moon ¹⁴ , Sang Kyun Sohn ¹⁴ , Seong Kyu Park ¹⁵ , Yong Park ¹⁶ , Ho-Jin Shin ¹⁷ , Sung-Soo Yoon ¹⁸

¹ Chonnam National University Hwasun Hospital, Hwasun, Korea, Republic of; ² Severance Hospital, Yonsei University College of Medicine, Seoul, Korea, Republic of; ³ Seoul St. Mary's Hospital, Catholic University of Korea, Seoul, Korea, Republic of; ⁴ Samsung Medical Center, Seoul, Korea, Republic of; ⁵ Ulsan University Hospital, Ulsan, Korea, Republic of; ⁶ National Cancer Center, Goyang-si, Korea, Republic of; ⁷ Inje University Busan Paik Hospital, Busan, Korea, Republic of; ⁸ Dong-A Medical Center, Busan, Korea, Republic of; ⁹ Chonbuk National University Medical School, Jeonju, Korea, Republic of; ¹⁰ Korea Cancer Center Hospital, Korea Institute of Radiological and Medical Sciences, Seoul, Korea, Republic of; ¹¹ Gachon University Gil Hospital, Incheon, Korea, Republic of; ¹² Kosin University Gospel Hospital, Busan, Korea, Republic of; ¹³ Ewha Woman's University School of Medicine, Seoul, Korea, Republic of; ¹⁴ Kyungpook National University Hospital, Daegu, Korea, Republic of; ¹⁵ Soonchunhyang University Bucheon Hospital, Bucheon, Korea, Republic of; ¹⁶ Korea University College of Medicine, Seoul, Korea, Republic of; ¹⁷ Pusan National University Hospital, Busan, Korea, Republic of; ¹⁸ Seoul National University Hospital, Seoul, Korea, Republic of

Background: Recent studies have suggested that intravenous busulfan and melphalan (BUMEL) might be an effective and tolerable conditioning regimen for autologous stem cell transplantation (ASCT) in patients with multiple myeloma (MM). This study compared the efficacy and toxicity of BUMEL and high-dose melphalan (HDMEL) as conditioning regimens in transplant- eligible patients with MM.

Methods: We retrospectively analyzed the records of 270 patients with newly diagnosed MM who received BUMEL or HDMEL as conditioning regimens followed by early ASCT (less than 12 months from the initial diagnosis). To reduce selection bias and its effect on the treatment results, we performed a 1:1 propensity score-matched analysis according to age, sex, International Staging System disease stage, lactate dehydrogenase level, and inclusion of cyclophosphamide in the induction regimen.

Results: Seventy-six patients received BUMEL, which consisted of intravenous busulfan administered at 3.2 mg/kg once a day on days -6 to -4 followed by melphalan administered at 70 mg/m² on days -3 and -2. The same number of patients received HDMEL consisting of melphalan at 100 mg/m² on days -3 and -2. All patients were treated with thalidomide-based regimens as the induction therapy. No statistically significant difference in the complete response (CR) and overall response rate (ORR) after ASCT was observed between BUMEL and HDMEL (CR 38.2% vs. 50.0%, P = 0.191; ORR 92.1% vs. 94.7%, P = 0.745). After a median follow-up of 25.2 months in the BU-MEL group and 60.8 months in the HDMEL group, the progression-free survival was 34.3 (20.9-47.8) months and 25.1 (12.3-38.0) months, respectively (P = 0.729). With respect to non-hematologic toxicities, infection was more frequently reported in the BUMEL group (P < 0.001), but no statistically significant differences in the incidence of grade 3 infections was observed (BUMEL 22.4% vs. HDMEL 15.8%, P=0.459). Venous-occlusive disease occurred in three patients who received BUMEL and recovered with supportive therapy.

Conclusions: In conclusion, BUMEL is an effective alternative conditioning regimen with acceptable toxicity for early ASCT in patients who were treated with thalidomide-based induction therapy.

Conflict of interest: The authors declare that they have no conflict of interest

P159

Abstract previously published

P160

Busulfan PK-guided dose-adjustment in adult allogeneic stem cell transplantation - single-center experience

Claudia Langebrake ^1,2 , Christine Wolschke ¹ , Maximilian Christopeit ¹ , Alexander Müller ³ , Nicolaus M. Kröger ¹

¹ University Medical Center Hamburg-Eppendorf, Department of Stem Cell Transplantation, Hamburg, Germany; ² University Medical Center Hamburg Eppendorf, Hospital Pharmacy, Hamburg, Germany; ³ University Medical Center Hamburg Eppendorf, Department of Legal Medicine, Hamburg, Germany

Background: It is well known, that there is a high inter-individual variety between dosing and exposition of busulfan (Bu) on the one hand and the correlation between exposition and clinical outcome on the other hand. Therapeutic drug monitoring (TDM) has been considered to optimise patient outcome. However, due to technical and logistical reasons, pharmacokinetics (PK) -guided dose-adjustment has not been established in the majority of haematopoietic stem cell (HSCT) centres in Europe yet.

Methods: Single-centre retrospective analysis of adult patients undergoing allogeneic HSCT with initial Bu dosing of 3.2 mg/kg Bu intravenously once daily (based on adjusted ideal body weight [AIBW] for obese patients). Busulfan concentration was measured with gas chromatography-mass spectrometry (GC-MS) and AUC was calculated using Bayesian curve fitting. If the calculated AUC was outside the pre-defined target AUC, dose adjustment was undertaken for the subsequent doses.

Results: In a 14 months period a total of 95 patients (male/female: 44/51) received Bu TDM with varying conditioning regimens, e.g. Bu/fludarabine, Bu/cyclophosphamide, thiotepa/Bu/±fludarabine with either antithymocyte globuline (ATG) (n=82) or post-transplantation cyclophosphamide (n=13).

Median age was 58.4 years (range 19.7-73.1), median body mass index was 25.1 kg/m² (range: 15.6-36). AIBW for initial dose calculation was used in 33 (34.5%) patients. Mean AUC at the first day of quantification was 16.4 h*mg/l (± 4.0, range: 9.7 - 31.9 h*mg/l). The calculated busulfan AUC before dose adjustment was in median 20.5 % (mean 18.1 ± 20.5 %) below the target AUC. This resulted in a dose adjustment in 71 (74.7%) patients with an increase of busulfan dose for the subsequent days in 61 (64.9%) and a dose decrease in 10 (10.6%) patients, respectively. After dose adjustment, the median deviation from the received AUC to the target AUC was -2.4 % (mean -5.1 % ± 10.4%), with 73 (76.8 %) of patients being within in the predefined cumulative AUC target range (± 10%).

Conclusions: Our data confirm the importance of patient individual dose adjustment of Bu during conditioning in adult allo SCT patients. The implementation of Bu-TDM within clinical routine care is feasible and results in a better target attainment of cumulative Bu exposition.

Conflict of interest: No conflicts of interest.

P161

Chemo-conditioning before allo-HSCT for children below two years of age with acute leukemia - an EBMT-PDWP study

Andre Willasch ¹ , Christina Peters ² , Petr Sedláček ³ , Jean-Hugues Dalle ⁴ , Stelios Graphakos ⁵ , Akif Yesilipek ⁶ , Jacek Wachowiak ⁷ , Arjan Lankester ⁸ , Arcangelo Prete ⁹ , Amir Ali Hamidieh ¹⁰ , Marianne Ifversen ¹¹ , Jochen Buechner ¹² , Gergely Kriván ¹³ , Rose-Marie Hamladji ¹⁴ , Cristina Diaz de Heredia ¹⁵ , Elena Skorobogatova ¹⁶ , Gérard Michel ¹⁷ , Franco Locatelli ¹⁸ , Alice Bertaina ¹⁸ , Paul Veys ¹⁹ , Sophie Dupont ²⁰ , Reuven Or ²¹ , Tayfun Güngör ²² , Olga Aleinikova ²³ , Sabina Sufliarska ²⁴ , Mikael Sundin ²⁵ , Jelena Rascon ²⁶ , Ain Kaare ²⁷ , Damir Nemet ²⁸ , Franca Fagiloi ²⁹ , Thomas Klingebiel ¹ , Jan Styczynski ³⁰ , Marc Bierings ³¹ , Kálmán Nagy ³² , Manuel Abecasis ³³ , Boris Afanasyev ³⁴ , Marc Ansari ³⁵ , Arnaud Dalissier ³⁶ , Myriam Labopin ³⁷ , Eric Beohou ³⁶ , Peter Bader ¹

¹ EBMT Paediatric Diseases Working Party and Division for Stem Cell Transplantation and Immunology, Department for Children and Adolescents, University Hospital, Goethe University, Frankfurt am Main, Germany; ² St Anna Children's Hospital, Vienna, Austria; ³ University Hospital Motol, Department of Pediatric Hematology and Oncology, Prague, Czech Republic; ⁴ Hôpital Robert Debré and Paris-Diderot University, Department of Pediatric Hemato-Immunology, Paris, France; ⁵ Stem Cell Transplant Unit, Aghia Sophia Children's Hospital, Thivon and Papadiamantopoulou, Athens, Greece; ⁶ School of Medicine, Bahcesehir University, Department of Pediatric Hematology-Oncology, Istanbul, Turkey; ⁷ University of Medical Sciences, Department and Clinic of Pediatric Oncology, Hematology and Transplantology, Poznan, Poland; ⁸ Leiden University Medical Center, Department of Pediatrics, Leiden, Netherlands; ⁹ Hematology-Oncology Unit 'Lalla Seràgnoli', Department of Pediatrics, University of Bologna, Bologna, Italy; ¹⁰ Hematology-Oncology and Stem Cell Transplantation Research Center, Tehran University of Medical Sciences, Tehran, Iran, Islamic Republic of; ¹¹ Copenhagen University Hospital, Rigshospitalet, Department of Pediatrics, Copenhagen, Denmark; ¹² Oslo University Hospital Rikshospitalet, Department of Pediatric Medicine, Oslo, Norway; ¹³ United St. Istvan and St. László Hospital, Department of Paediatric Haematology and Stem Cell Transplantation, Budapest, Hungary; ¹⁴ Centre Pierre et Marie Curie, Service Hématologie Greffe de Moelle, Alger, Algeria; ¹⁵ Servicio de Hematologia y Oncologia Pediátricas, Hospital Universitario Vall d'Hebron, Barcelona, Spain; ¹⁶ Russian Children's Hospital, BMT Department, Moscow, Russian Federation; ¹⁷ Timone Enfants Hospital, AP-HM and Aix-Marseille University, Department of Pediatric Hematology and Oncology and Research Unit EA 3279, Marseille, France; ¹⁸ IRCCS Ospedale Pediatrico Bambino Gesù, Università di Pavia, Dipartimento di Onco-Ematologia Pediatrica e Medicina Trasfusionale, Rome, Italy; ¹⁹ Great Ormond Street Hospital for Children NHS Foundation Trust, London, United Kingdom; ²⁰ Cliniques Universitaires Saint-Luc, Department of Pediatric Hematology and Oncology, Brussels, Belgium; ²¹ Hadassah-Hebrew University Medical Center, Department of Bone Marrow Transplantation, Jerusalem, Israel; ²² Division of Stem Cell Transplantation, Children’s Research Center (CRC), University Children's Hospital, Zurich, Switzerland; ²³ Republic Clinical Research Centre for Pediatric Oncology and Hematology, Minsk, Belarus; ²⁴ BMT Unit, Department of Pediatric Hematology and Oncology, Comenius University Medical School, Limbová, Bratislava, Slovakia; ²⁵ Pediatric Blood Disorders, Immunodeficiency and Stem Cell Transplantation, Astrid Lindgren Children's Hospital, Karolinska University Hospital, Stockholm, Sweden; ²⁶ Center of Pediatric Oncology and Hematology, Bone Marrow Transplantation Unit, Children's Hospital, Affiliate of Vilnius University Hospital Santariskiu Klinikos, Vilnius, Lithuania; ²⁷ Tartu University Hospital, Tartu, Estonia; ²⁸ Internal Clinic, University Hospital Centre, Department of Haematology, Zagreb, Croatia; ²⁹ Paediatric Onco-Haematology, City of Science and Health of Turin, Regina Margherita Children's Hospital, Turino, Italy; ³⁰ Nicolaus Copernicus University Torun, Department of Pediatric Hematology and Oncology, Collegium Medicum, Bydgoszcz, Poland; ³¹ BMT-Unit, University Medical Centre Utrecht Pediatrics, Utrecht, Netherlands; ³² Child Welfare Center, Borsod County Teaching Hospital, Department of Hematology, Miskolc, Hungary; ³³ Bone Marrow Transplant Program, Instituto Portugues Oncologia, Lisbon, Portugal; ³⁴ Raisa Gorbacheva Memorial Research Institute for Pediatric Oncology, Haematology and Transplantation, Saint Petersburg State Medical I.P. Pavlov University, Saint Petersburg, Russian Federation; ³⁵ Geneva University Hospital, Geneva University, Department of Pediatrics, Onco-Hematology Unit, Geneva, Switzerland; ³⁶ EBMT Paediatric Diseases Working Party, Paris, France; ³⁷ EBMT Paris Study Office, Paris, France

Background: Paediatric patients (pts) below two years (y) of age with acute leukemia commonly receive chemotherapy for myeloablative conditioning (MAC) before allogeneic haematopoietic stem cell transplantation (allo-HSCT). It is unknown, which combination of chemotherapeutics offers the best outcome.

Methods: To investigate the outcomes of different chemotherapy-based conditioning-regimens, we performed a retrospective Paediatric Diseases Working Party (PDWP)-EBMT-registry based study. Children below two y of age after MAC for first allo-HSCT of bone marrow (BM), peripheral blood SC (PBSC) or cord blood (CB) from matched siblings (MSD) or unrelated donors (UD) in first remission (CR1) between 2000 and 2014 were included. Propensity score weighting was used to consider the bias due to confounding variables, e.g. different size and composition of subgroups. Univariate analysis and multivariable Cox model were weighted on diagnosis, age at HSCT, y of HSCT, interval between diagnosis and HSCT, donor type and SC source. Cox model was clustered on EBMT centers.

Results:

In total 400 pts (AML: 218 (54%), ALL: 182 (46%)) were included. 215 pts (54%) received BM, 128 pts (32%) CB and 57 (14%) PBSC from UD (254 (64%)) or MSD (146 (36%)). Thirty-six different conditioning-regimens were used. Busulfan/Cyclophosphamide+/-Melphalan+/-ATG (Bu/Cy+/-Mel+/-ATG) was the most frequently applied regimen (255 (64%)). The remaining conditionings were summarized as “others” (145 (36%)). Patients treated with these two conditioning groups had a median age of 1.2 y (interquartile range (IR) 0.8-1.6 y) and 1.1 y ((IR) 0.8-1.4 y) and a median follow-up of 3.8 and 2.4 y.

No significant differences between the two conditioning cohorts Bu/Cy+/-Mel+/-ATG and “others” were detected for LFS (5-y-LFS, 61.6 vs. 64.7%), relapse incidence (5-y-RI, 27.0 vs. 24.7%), OS (5-y-OS, 70.5 vs. 66.8%), non-relapse mortality (5-y-NRM, 11.3 vs. 10.6%), aGvHD grade III-IV (day 100) 12.6 vs. 13.0%, cGvHD (5-y) 16.2 vs. 16.9% and for GvHD-free/relapse-free survival (5-y-GRFS, 54.4 vs. 58.0%).

In multivariable Cox model, no significant differences between the two conditioning cohorts were found as well.

Conclusions: Allo-HSCT for children below two years of age with high relapse-risk acute leukemia offers an impressive chance of cure. The most frequently applied regimen Bu/Cy+/-Mel+/-ATG did not demonstrate superiority compared to other conditionings. This retrospective data is currently re-evaluated in a prospective, randomized, international trial (ALL SCTped 2012 FORUM).

Conflict of interest: The authors declare no relationships to disclose.

P162

CMV status predicts survival in refractory/relapsed myeloid patients after a clofarabine-based sequential regimen: a retrospective study on behalf of the SFGM-TC

Amandine Le Bourgeois ¹ , Myriam Labopin ² , Felipe Suarez ³ , Regis Peffault De Latour ⁴ , Didier Blaise ⁵ , Sylvain Chantepie ⁶ , Stephanie Nguyen ⁷ , Natacha Maillard ⁸ , Reza Tabrizi ⁹ , Ibrahim Yakoub-Agha ¹⁰ , Anne Huynh ¹¹ , Tony Marchand ¹² , Karin Bilger ¹³ , Patrice Ceballos ¹⁴ , Amandine Charbonnier ¹⁵ , Pascal Turlure ¹⁶ , Marie-Therèse Rubio ¹⁷ , Mohamad Mohty ² , Patrice Chevallier ¹

¹ Nantes Hospital, Hematology, Nantes, France; ² Saint Antoine Hospital, Hematology, Paris, France; ³ Necker Hospital, Hematology, Paris, France; ⁴ Saint Louis Hospital, Hematology, Paris, France; ⁵ Institut Paoli Calmettes, Hematology, Marseille, France; ⁶ Caen Hospital, Hematology, Caen, France; ⁷ Salpétrière Hospital, Hematology, Paris, France; ⁸ Poitiers Hospital, Hematology, Poitiers, France; ⁹ Bordeaux Hospital, Hematology, Bordeaux, France; ¹⁰ Lille Hospital, Hematology, Lilles, France; ¹¹ Toulouse Hospital, Hematology, Toulouse, France; ¹² Rennes Hospital, Hematology, Rennes, France; ¹³ Strasbourg Hospital, Hematology, Strasbourg, France; ¹⁴ Montpellier Hospital, Hematology, Montpellier, France; ¹⁵ Amiens Hospital, Hematology, Amiens, France; ¹⁶ Limoges Hospital, Hematology, Limoges, France; ¹⁷ Nancy Hospital, Hematology, Nancy, France

Background: Recently, we have published the prospective results of a sequential regimen using a clofarabine (clo), cytosine arabinoside and reduced-intensity conditioning (RIC) regimen in 24 cases with primary refractory acute myeloid leukemia (AML) showing encouraging results with a very low non relapse mortality (Mohty, 2017). Here we report the outcomes after such sequential regimen in a larger cohort of patients.

Methods: This was a retrospective study including all patients reported within the SFGM-TC registry having received a clo-based sequential conditioning regimen before allotransplant for active myeloid disease. Data were obtained through PROMISE, an internet-based system shared by all European transplantation centers. All patients gave informed consent allowing to collect their personal data from this data base.

The sequential regimen consisted of clo (30 mg/m2/day) and cytosine arabinoside (1 g/m2/day) for 5 days, followed, after a 3-day rest, by RIC combining cyclophosphamide (60 mg/kg), iv busulfan (3.2 mg/kg/day) for 2 days and anti-thymocyte globulin (2.5 mg/kg/day) for 2 days.

The primary objective of the study was to report the main outcomes at 1 year: overall survival (OS), disease free survival (DFS), relapse incidence (RI), NRM and GVHD Relapse free survival (GRFS). Secondary objectives were to identify prognostic factors for patient's survival.

Results: Between January 2007 and December 2016, 131 patients (males n=75, median age: 52.6 years, range: 18-71) met the inclusion criteria. There were 111 AML patients, including 9 secondary AML, and 20 cases with myelodysplastic or myeloproliferative syndrome. Status at transplant was as follows: primary refractory n=81, relapse 1 or 2 n=46, missing n=4. The majority of patients received peripheral blood stem cell as source of graft (n=127, bone marrow n=4). All patients received a graft from a matched donor (sibling n=64, unrelated n=67). Donor (D)/recipient (R) CMV status was as follows: D-/R-: n=55, D+/R-: n=14, D-/R+: 24, D+/R+: 38.

Engraftment was observed in 105/122 assessable cases (86%) and patients achieving complete remission (CR) after transplant were 63% (n=72/114 evaluable cases). With a median follow-up of 12 months (range: 2.7-74.3) for alive patients, 1-year OS, DFS, RI, NRM and GRFS were 39.2% [30.2-48.2], 28.1% [19.8-36.5], 41% [32.1-49.8], 30.8% [22.7-39.3], and 22.2% [14.3-30], respectively.

In multivariate analysis, D/R CMV negative status was associated with lower OS (HR: 1.74, 95%CI: 1.10-2.76, p=0.016), DFS (HR: 1.70; 95%CI: 1.09-2.66, p=0.018) and GRFS (HR: 1.76; 95%CI: 1.15-2.71, p=0.009) and higher RI (HR: 2.48; 95%CI: 1.37-4.50, p=0.002). D/R CMV status was not associated with NRM.

Conclusions: Despite high CR achievement, this large cohort of patients confirmed the relatively poor outcome of patients with active myeloid disease at transplant and receiving a clofarabine-based sequential regimen. The favorable impact on survivals of positive D or R CMV status may be related to a potential graft-vs-leukemia effect of CMV reactivation (thanks to NK or T cells stimulation) after transplant. The validation of this hypothesis is on-going.

Conflict of interest: no conflict of interest

P163

Combined use of Treosulfan and epigenetic agent in high-dose conditioning regimen in patients with very high-risk solid tumors

Igor Dolgopolov, George Mentkevich, Nataly Subbotina, Vasily Boyarshinov, Vidmante Daylidite

Institute for Pediatric Oncology & Hematology, Pediatric BMT, Moscow, Russian Federation

Background: In attempt to improve the treatment results in patients with very high-risk Ewing's sarcoma (ES) and neuroblastoma (NB) we initiated a study of HDCT with autologous SCT including epigenetic therapy (in order to increase tumor cell chemosensitivity) and split conditioning regimen based on a near maximum of tolerated dose of Treosulfan (Treo).

Methods: Two patients with solid tumor with multiple bone involvement were included. One with stage 4 ES after 5 courses and one with stage 4 NB after 7 courses of induction therapy. Both patients were over age 2 and in VGPR. Split conditioning regimen consisted of treosulfan 10000 mg/m²/d on d-11,-10, and -4, -3 (40000 mg/m²/course), melphalan 70 mg/m²/d, on d-3,-2 (140 mg/m²/course) and 5-azacitidine 75 mg/m²/d, given on d-14 through -12 and d -9 through -3 (8 injections, 600 mg/m²/course). On d0 patients were transplanted with autologous PBSC with 5.2 and 5.0x10⁶ CD34+ cells/kg. G-CSF was given from d0.

Results: Regimen related toxicity was moderate and manageable. Both pts developed grade 2-3 GI (stomatitis with painful erythema requiring parenteral nutrition and narcotic drugs, diarrhea with 4-6 stool/day, moderate dyspepsia and/or dysphagia) and febrile neutropenia requiring IV antibiotics. In both cases transient liver toxicity of grade 2-3 with AST/ALT elevation and hemorrhage of grade 2 were observed. Patients received 1 and 3 RBC, and 3 and 14 PLT transfusions, respectively. Both patients engrafted with WBC>1.0x10⁹/l on d+10 and +16 and PLT>20x10⁹/l on d+14 and +30, respectively. Patient with NB is alive without disease 10 mos. after HDCT. Patient with ES relapsed 19 mos. after HDCT and stated on a research protocol.

Conclusions: We can speculate that Treo based split regimen with addition of 5-azacitidine is feasible with acceptable toxicity profile. This regimen seems to be reasonable for solid tumors with low cell growth fraction. More pts and longer follow-up period are needed to make further conclusions.

Conflict of interest: No conflict of interest.

P164

Comparable Outcomes after Allogeneic Stem-cell Transplantation with Myeloablative BuFlu or BuCy Conditioning Regimens in Acute Myeloid Leukemia - Analysis of the Polish Adult Leukemia Group

Anna Lojko-Dankowska ¹ , Barbara Nasilowska-Adamska ² , Patrycja Mensah-Glanowska ³ , Agnieszka Piekarska ⁴ , Magdalena Matuszak ¹ , Beata Piatkowska-Jakubas ³ , Kazimierz Halaburda ² , Tomasz Wrobel ⁵ , Mieczyslaw Komarnicki ¹ , Anna Czyz ⁵

¹ University of Medical Sciences, Poznan, Poland; ² Institute of Hematology and Transfusion Medicine, Warsaw, Poland; ³ Jagiellonian University Medical Collage, Krakow, Poland; ⁴ Medical University of Gdansk, Gdansk, Poland; ⁵ Wroclaw Medical University, Wroclaw, Poland

Background: Busulfan (Bu)-based myeloablative conditioning regimen is widely used for patients with acute myeloid leukemia (AML) treated with allogeneic hematopoietic stem cell transplantation (alloHCT). Busulfan is most frequently used in combination with either cyclophosphamide (Cy) or fludarabine (Flu). The aim of this multicenter retrospective analysis was to compare the outcome of alloHCT in adults patients with AML in first complete remission (CR1) or beyond CR1 who were conditioned with myeloablative BuCy or reduced toxicity BuFlu regimen.

Methods: Adult patients with AML (n=128) treated with alloHCT from either HLA-identical sibling (n=37) or unrelated donor (n=91), in CR1 (n=99) or beyond CR1 (n=29), who were conditioned with iv busulfan given in myeloablative dose (9,6-12,8 mg/kg) in combination with Flu (n=72) or Cy (n=56), between year 2000 and 2017, were included in the analysis. Patients treated with BuFlu were older (median 44 y. vs. 34 y., p< 0.001) and received more frequently stem cells collected from peripheral blood (94% vs. 43%, p< 0.001).

Results: Engraftment was observed in all patients. The median time to neutrophil count (0.5 G/L) and platelet count (20 G/L) recovery was shorter after BuFlu in comparison with BuCy (18 days vs 21 days; p < 0.001 and 11 days vs. 18 days; p< 0.001). Febrile neutropenia occurred in 67% of patients in the BuFlu group and 84% in the BuCy group (p=0.06). Non-infectious complications grade 2-4 were observed in 11% and 9% of patients, respectively (p=0.13). Acute GvHD grade II-IV was present in 14% of patients in the BuFlu group and 20% patients in the BuCy group (p= 0.58). The respective frequency of extensive moderate/severe chronic GvHD was 24% and 14% (p= 0.13). The median follow up time among survivors is 22 months (range, 2-80). The 2-year OS for patients conditioned with BuFlu was 65% and for those conditioned with BuCy 67% (log-rank p= 0.61). The respective 2-year LFS was 62% and 64% (p= 0.56). In addition, the OS and LFS for patients transplanted in CR1 and beyond CR1 did not differ significantly between the BuFlu and BuCy group. Similarly, the 1-year cumulative incidence (CI) of non-relapse mortality and the 2-year CI of relapse were comparable between two groups (11% vs 9% and 21% vs 22%, respectively) (p ns). In multivariate analysis faster lymphocyte recovery favorably influenced OS (HR 0.47, p=0.026).

Conclusions: Myeloablative reduced toxicity conditioning regimen based on iv busulfan combined with fludarabine appears as effective as the BuCy combination in adult patients with AML, both in first and subsequent CR. The results of our study indicate that Cy may be replaced by fludarabine in busulfan-based myeloablative conditioning regimen without negative impact on disease control.

Conflict of interest: None of the authors has anything to declare.

P165

Comparison between Equine and Rabbit anti-thymocyte globulin (ATG) used in conditioning for pediatric patients undergoing Hematopoietic stem cell transplantation (HSCT) for thalassemia and aplastic anaemia

Sunil Bhat ¹ , Ruchi Chaudhary ¹ , Shobha Badiger ¹ , Pooja Mallya ¹ , Sohini Chakraborty ¹ , Archana M V ¹ , Ravi Joshi ¹ , Nataraj K S ² , Sharat Damodar ²

¹ Mazumdar Shaw Cancer Center, Narayana Health City, Pediatric Hematology Oncology and Bone Marrow Transplant, Bangalore, India; ² Mazumdar Shaw Cancer Center, Narayana Health City, Bangalore, India

Background: HSCT related complications and disease relapse are major obstacles in successful transplantation. Addition of anti-thymocyte globulin (ATG) has reduced rejection and GVHD incidence significantly. This study was conducted to look for relative efficacy of rabbit and equine ATG in pediatric HSCT conditioning.

Methods: Single centre retrospective analysis of thalassemia major and severe aplastic anemia (SAA) patients who underwent HSCT (January 2013-May 2017) and received ATG during conditioning regimen.

Results: Rabbit or Equine ATG were used as a part of conditioning regimen in 79 thalassemia (Pesaro class I/class II/class III: 2/61/16) and 21 patients of SAA (frontline transplant/after failure of immunosuppression: 18/3). Equine ATG group was labeled as group-1 and rabbit ATG as group-2.

Median CD34 dose (10⁶/kg) used in group-1 and 2 were 6 and 5.55 in thalassemia while 2.95 and 5.54 in SAA respectively. CSA-MTx was used as GVHD prophylaxis. Median total dose of equine ATG was 90mg/kg and rabbit ATG 7.5mg/kg. Flu/Bu/CY/ATG was the conditioning regimen in class I and II of thalassemia while treosulphan based conditioning was used in class III. Both groups were analyzed for transplant related complications and outcome. Outcomes are shown in table-1 and were comparable except Group-1 had more cases of proven bacterial infection (p=0.07) and veno-occlusive disease (p=0.03). Acute GVHD was seen in 40% patients in both groups (p=0.985) and chronic GVHD in 8.5% in group-1 and 6.6% cases of group-2(p=1). Median chimerism at day+30 and day+90 in group-1 were 96% and 97% while 96% and 95.5% in group-2(p value 0.92& 0.86). Overall survival (OS) were 85% and 84% in group-1 and group-2 over a median period of follow up 29months (as shown in figure-1) and event free survival (EFS) 80.8% in group-1 and 78% in group-2 respectively(p-value 0.9 & 0.77).

In SAA, Flu/Bu/ATG was used for conditioning and CSA-MTx as GVHD prophylaxis. Bone marrow was used as stem cell source in 10 patients and PBSC in 11. Group-1 had early neutrophil and platelet engraftment with less number of proven bacterial infections. Median chimerism at day+30 and day +90 were 90% and 93% in group-1 while it was 93% in group-2(p=0.91) at both time intervals. Acute GVHD was seen in 22% in group-1 and in 25% cases in group-2. OS was 67% in both groups.

Conclusions: Effectiveness of rabbit and horse ATG were similar in thalassemia patients while in SAA equine ATG use was associated with early neutrophil engraftment and less incidence of bacterial infection but OS in both groups were comparable.

Conflict of interest: None

	Thalassemia			Aplastic anemia
	Rabbit(N=32)	equine(N=47)	P value	Rabbit(N=12)	equine(N=9)	P Value
Type of transplant MSD/MRD/HAPLO/MUD	20/2/3/7	38/2/1/6	0.23	7/2/3	7/2/0	0.63
Neutrophil/ Platelet engraftment(median)	14/20	13.5/18.5	0.24/0.24	15/22	13/14	0.75/ 0.68
Proven bacterial infection	16	5	0.07	6	2	0.21
CMV reactivation	30	21	0.1	6	6	0.46
Invasive fungal infection	3	3	0.62	1	2	0.38
Acute GVHD Grade I - II Grade III - IV	10 9	4 9	0.958	0 3	2 0	0.19
Chronic GVHD	1	2	1	2	0	0.21

[ [P165 Table] Table 1]

[P165 Figure]

$$$$

P166

Comparison of non-myeloablative and reduced-intensity conditioning prior to allogeneic stem cell transplantation in older patients with myelodysplastic syndrome

Madlen Jentzsch, Christian Pfrepper, Richard Linke, Christine Döhring, Juliane Grimm, Stefanie Beinicke, Janine Häntschel, Andrea Hille, Wolfram Pönisch, Vladan Vucinic, Georg-Nikolaus Franke, Gerhard Behre, Dietger Niederwieser, Sebastian Schwind

Universitätsklinik Leipzig, Hematology and int. Oncology, Leipzig, Germany

Background: Myelodysplastic syndrome (MDS) is a clinically heterogeneous disease. Allogeneic stem cell transplantation (HSCT) remains the only curative treatment. The most appropriate conditioning intensity remains to be determined. In younger MDS patients a lower non relapse mortality (NRM) but higher cumulative incidence of relapse (CIR) was shown in reduced-intensity (RIC) compared to myeloablative conditioning HSCT. Dose reduced regimens, including RIC and non-myeloablative (NMA) regimens, allow HSCT in older individuals representing the majority of MDS patients. Data comparing RIC and NMA conditioning for HSCT in older MDS patients are lacking.

Methods: We retrospectively analyzed 117 MDS patients older than 50 (median 61.3, range 50.1-74.9) years who received NMA (3x30mg/qm fludarabine with 2Gy [n=70] or 3Gy [n=1] total body irradiation [TBI]) or RIC (5x30mg/qm fludarabine with 8mg/kg oral [n=41] or 6.4mg/kg intravenous [n=5] busulfan, including 3x2mg/kg thymoglobulin in unrelated HSCT) HSCT. While patients < 65 years received NMA (51%) or RIC (49%) conditioning, patients >=65years predominantly received NMA (85%). Thus, outcomes for patients younger and older than 65years at HSCT were analyzed separately. Karyotype analyses were performed centrally at our institution. IPSS-R was 0% very low, 8% low, 24% intermediate, 28% high, 40% very high.

Results: Overall, NMA patients were older (P=.007), but age did not differ between RIC and NMA conditioned patients when patients younger (P=.96) or older (P=.55) than 65 years were compared separately. The percentage of patients receiving therapy (demethylating agents and/or chemotherapy) prior to HSCT did not differ between RIC and NMA conditioning (68% vs. 56%, P=.24), however due to predominant application of NMA conditioning before 2012, NMA patients less often received demethylating agents alone (27% vs. 64%, P=.005). The proportion of patients with blast excess (P=.27) and complex karyotype (25% vs. 24%, P=1) as well as scores for IPSS-R risk (P=.57), IPSS-R genetic risk (P=.64) and HCT-CI (P=.80) did not differ between NMA- or RIC-conditioned patients. NMA-conditioned patients had more graft rejections (11% vs. 0%, P=.02) or chronic graft-versus-host disease by trend (cGvHD, 87% vs. 49%, P=.08). In patients younger than 65years (n=84), NMA associated with higher NRM (P=.02, Figure 1A) and shorter overall survival (OS, P=.04, Figure 1B). In contrast, there was no difference in CIR (P=.22) or secondary acute myeloid leukemia (CIsAML, P=.93) between NMA or RIC conditioning. In patients older than 65years at HSCT (n=33), there were no differences in NRM (P=.96), OS (P=.49), CIR (P=.25) or CIsAML (P=.59) with regard to the conditioning regime.

Conclusions: Patients aged 50-65years receiving NMA-HSCT had more graft rejection, more cGVHD by trend, higher NRM and shorter OS. Our study is limited by its retrospective nature and a possible selection bias towards NMA in older or less fit patients. While age and the comorbidity index HCT-CI did not differ between NMA and RIC, some comorbidities might not be appropriately reflected in the HSCT-CI scoring. Patients younger than 65years might benefit from the more intensive RIC or intensification of NMA (e.g. 3 Gy TBI) conditioning. However, NMA conditioning in older patients above 65years allows curative treatment with long-term survival.

Conflict of interest: G. Franke: Takeda: consultancy; Pfizer: honoraria; Novartis: honoraria, research funding.

S. Schwind: Novartis: Consultancy, research funding.

All other authors have nothing to disclose.

[P166 Figure]

$$$$

P167

Comparison of Safety and Efficacy of Bendamustine versus Carmustine in the conditioning regimen for autologous stem cell transplantation in patients with relapsed/refractoy lymphoma

Munira Moosajee, Samad Jehangir, Sobia Sawani, Wajeeha Tariq, Natasha Ali, Mohammed Usman Sheikh, Salman Adil

Aga Khan University, Karachi, Pakistan

Background: Bendamustine and carmustine have shown have both been used as part of conditioning regimen for autologous stem cell transplant (SCT) for relapsed/ refractory lymphoma. However, bendamustine has a higher toxicity profile as compared to carmustine in some studies. Carmustine availability is sporadic in many countries and therefore bendamustine is being used more frequently. We report our experience with bendamustine as part of conditioning regimen (BeEAM) because of non availability of carmustine in Pakistan.

Methods: This is a retrospective study to compare the safety and efficacy of bendamustine at 150 mg/m2 on day -5 and -4 coupled with etoposide 200 mg/m2 day -5 to -2, cytarabine 200 mg/2 on day -5 to -2 and melphalan 140mg/m2 on day -1 (BeEAM) versus carmustine 300 mg/m2 on day -6 coupled with the same agents (BEAM) as conditioning regimen to autologous SCT.

Results: Total 45 patients were included in the study. 14 patients in the BeEAM group and 31 patients in the BEAM group. The median age was 28 years in both groups. 49 % (n=22) had Hodgkin's lymphoma and 28% (n= 17) were diagnosed with non hodgkins lymphoma (n= 15), predominantly diffuse large B cell and mantle cell lymphoma. 57 % vs 6 % of the patients were in CR at time of transplant in the BeEAM and BEAM groups respectively. The median number of CD 34+ cells infused was 5.5 x 10(6) CD 34+ cells/ kg and was similar in both groups. Median time to absolute neutrophil count engrafted of 17 days (BeEAM) vs 13 days (BEAM). Median time to platelet engraftment of 25 days (BeEAM) versus 16 days (BEAM). The 100 day transplantation mortality was 28.6 % (4 patients) vs 19.3% (6 patients) . 8 patients (57.2%) vs 5 patients (16%) had GII - III acute kidney injury, 5 patients (35.7 %) vs 9 patients(29%) had grade II-III hyperbilirubinemia and 7 patients (50%) vs 7 patients (22%) had GII diarrhea. At the end of 3 months, 35.7% (5 patients) vs 15 patients (48%) were in complete response. The median disease free survival with 3 months vs 15.4 months (p=0.0420) and overall survival was 30 months vs. 44 months (p=0.02) in the BeEAM and BEAM cohorts respectively.

Conclusions: Our study shows that bendamustine is a potentially toxic agent in the conditioning regimen when compared to BCNU for autologous SCT with significant liver, kidney and GI toxicity. In addition, the overall survival and disease free survival is inferior to the BEAM conditioning regimen. Prospective trials need to be done to determine the efficacy and safety of bendamustine in conditioning regimen for ASCT.

Conflict of interest: None of the authors has anything to disclose.

P168

Cyclophosphamide, bendamustine and etoposide (CBeV), a new CBV conditioning regimen for lymphoma autotransplants

Gregorio Jaimovich ¹ , Belen Rosales Ostriz ² , Martin Castro ² , Nicolas Fernendez Escobar ³ , Agustina Cia ³ , Leandro Riera ⁴ , Patricio Duarte ⁴ , Cecilia Foncuberta ⁵ , Adriana Vitriu ⁵ , Alejandra Banchieri ⁶ , Hector Longoni ⁶

¹ Favaloro University Hospital / Sanatorio Anchorena, BMT, Buenos Aires, Argentina; ² Sanatorio Anchorena, BMT, Buenos Aires, Argentina; ³ Favaloro University Hospital, BMT, Buenos Aires, Argentina; ⁴ Cemic, BMT, Buenos Aires, Argentina; ⁵ Fleming Institute, BMT, Buenos Aires, Argentina; ⁶ Clinicas Hospital Buenos Aires, BMT, Buenos Aires, Argentina

Background: CBV (cyclophosphamide, carmustine [BCNU] and etoposide) is commonly used for pretransplant conditioning for lymphoma autotransplants. However, carmustine is associated with lung toxicity, high cost and variable availability.

Methods: We developed a new CBV regimen (CBeV) swapping bendamustine for carmustine and conducted a multi-center phase-2 study evaluating the safety and efficacy of CBeV (cyclophosphamide, 4.8-6.0 g/mE+2, bendamustine, 400 mg/mE+2, etoposide 1.0-2.4 g/mE+2) in 100 consecutive subjects receiving autotransplants for lymphomas. Median age was 49 years (range, 18-74 years). 61 were male. 36 had Hodgkin lymphoma, 24, diffuse large B-cell lymphoma (DLBCL), 18, mantle cell lymphoma (MCL) and 22, other lymphoma diagnoses. 78 subjects failed ≥2 treatment lines pretransplant. At transplant 33 were in 1^st complete remission (CR), 42 in 2^nd CR and 25 in partial remission (PR). Subjects received a median of 3.9 X 10E+6 CD34-postve cells/kg (range, 1,2-27 x10E+6/kg). Median follow-up is 13 months (range, 3-34 months).

Results: All evaluable subjects had bone marrow recovery after a median of 11 days (range, 9-23 days) for neutrophils >0.5 X 10E+9/L and 14 days (range, 7-40 days) days for platelets >20 X 10E+9/L. Grade- 3- 4 toxicities were fever (N=75), gastrointestinal (N=36), heart (N=10),renal (N:4) and pulmonary (N= 6), but in contrast to the lung toxicity of carmustine toxicity after CBeV was reversible. 1 subject died of renal and multi-organ failure. 90 subjects are alive; 10 died including 9 from lymphoma progression. Overall survival at 1.5 years was 88 % (confidence interval (CI) 95% (81.4 - 96.1) including 92% (CI: 86, 99%) for subjects transplanted in 1^st or 2^nd CR and 77% (CI: 59, 100%); for those transplanted in PR, p=0.168. Cumulative incidence of relapse at 1.5 years was 25% (CI: 12, 37%) for subjects transplanted 1^st or 2^nd CR and 32% (CI: 3, 52%); for those transplanted in PR, p=0.42.

Conclusions: The safety and efficacy profiles of CBeV are similar to that of CBV but CBeV has the advantage of lower cost and more consistent availability. Preliminary efficacy results are encouraging but longer follow-up is needed to assess safety and efficacy of this regimen. A randomized trial comparing CBeV to CBV would be needed to definitively address this question.

Conflict of interest: None of the authors has anything to disclose.

[P168 Figure]

$$$$

P169

Excellent outcome for 91 Fanconi Anemia patients undergoing matched related transplants using Cyclophosphamide 60mg/kg in Curitiba, Brazil

Carmem Bonfim ^1,2,3 , Samantha Nichele ¹ , Lisandro Ribeiro ^1,3 , Gisele Loth ¹ , Daniela Pilonetto ¹ , Ana Luiza Fabro ² , Cilmara Kuwahara ² , Fernanda Lara Benini ² , Daniela Marinho ¹ , Marco Antonio Bitencourt ¹ , Rebeca Toassa Gomes Mousquer ¹ , Noemi Farah Pereira ¹ , Ricardo Pasquini ^1,3

¹ BMT Unit Federal University of Parana, Curitiba, Brazil; ² Hospital Infantil Pequeno Principe, Curitiba, Brazil; ³ Hospital Nossa Senhora das Graças, Curitiba, Brazil

Background: Fanconi anemia(FA) is a rare inherited disorder characterized by congenital abnormalities, progressive bone marrow failure and cancer predisposition. Survival after Hematopoietic Cell transplantation (HCT) improved dramatically over the past decade due to the use of better preparatory regimens and supportive care.

Methods: In this retrospective study we will analyze data on clinical presentation and outcome of 91 patients(pts) with FA transplanted between 06/1999 and 02/2017. Median age at transplantation was 9ys (range: 7 - 23). 57% were male. 87% were CMV positive. 77pts (85%) received a median of 7 blood transfusions (range: 1-151) before transplant. All pts were transplanted in marrow failure and received bone marrow from matched siblings (MSD) (n=74) or other matched related donors (MRD) (n=17). All pts received a preparatory regimen with Cyclophosphamide 60mg/kg(CY60) without (n=82) or with rabbit-ATG 5mg/kg (n=9) and GVHD prophylaxis with cyclosporine and methotrexate.

Results: One patient died before D+21 and was not evaluable for engraftment. Primary graft failure occurred in one patient in the MSD group and this patient was successfully rescued after a 3^rd transplant using a haploidentical donor with post-transplantation CY. Ninety pts engrafted and secondary graft failure occurred in 5pts (MSD=3pts; MRD=2pts) between 49 and 742 days after transplant (M: 156). Four out of these 5 pts are alive and well with full donor chimerism after a 2^nd or 3^rd HCT. Mucositis grade II-III occurred in more than 70% of pts. Acute GVHD occurred in 10/89 evaluable pts; grade II (n=5); grade III (n=2) and grade IV (n=3) while chronic GVHD occurred in 21/89 evaluable pts; NIH global severity score was mild (n=10); moderate (n=5) and severe (n=6). Seven pts died at a median of 906 days after transplant (range: 12- 5088). Causes of death include oral squamous cell carcinoma (n=2); GVHD (n=3); central nervous system bleeding (n=1) and sepsis (n=1). Transplant related mortality at one year was very low (n=3). Eighty-four pts are alive with a median follow-up of 7 ys (Range: 10 months - 18 years) and an overall survival of 95% at 5y. There was no difference in survival according to the type of donor MSD(91%) vs MRD(94%). In the univariate analysis survival was much better for pts without acute-GVHD compared to pts with this complication (98.7% vs 50% p< 0.001). 14 pts did not receive any transfusions before transplant and all of them are alive and well, none developed acute-GVHD and 2 had mild chronic GVHD.

Conclusions: In this study, the use of CY60mg/kg with or without ATG was associated with an excellent survival, good engraftment and very low mortality rate. This highly successful and low-cost preparatory regimen can be used all over the world to treat patients with FA in aplastic phase.

Conflict of interest: All authors declare no conflict of interest

P170

Favorable efficacy of TBI-based compared with high-dose chemotherapy conditioning for auto-SCT in patients with peripheral T-cell lymphomas: An analysis by the Polish Lymphoma Research Group

Tomasz Czerw ¹ , Andrzej Smagur ¹ , Anna Czyz ² , Maria Sadus-Wojciechowska ¹ , Jacek Najda ¹ , Wlodzimierz Mendrek ¹ , Malgorzata Sobczyk-Kruszelnicka ¹ , Joanna Romejko-Jarosinska ³ , Grzegorz Helbig ⁴ , Agnieszka Piekarska ⁵ , Lidia Poplawska ³ , Beata Piatkowska-Jakubas ⁶ , Dorota Hawrylecka ^6,7 , Barbara Nasilowska-Adamska ⁸ , Anna Lojko-Dankowska ⁹ , Anna Kopinska ⁴ , Jan Walewski ³ , Mieczyslaw Komarnicki ⁹ , Leszek Miszczyk ¹⁰ , Jerzy Holowiecki ¹ , Sebastian Giebel ¹

¹ Maria Sklodowska-Curie Memorial Cancer Center and Institute of Oncology, Gliwice Branch, Department of Bone Marrow Transplantation and Oncohematology, Gliwice, Poland; ² Wroclaw Medical University, Wroclaw, Poland; ³ Maria Sklodowska-Curie Cancer Centre and Institute of Oncology, Warsaw, Poland; ⁴ Silesian Medical University, Katowice, Poland; ⁵ Medical University of Gdansk, Gdansk, Poland; ⁶ Jagiellonian University Medical College, Cracow, Poland; ⁷ Oncology Center, Brzozow, Poland; ⁸ Institute of Hematology and Blood Transfusion, Warsaw, Poland; ⁹ Poznan University of Medical Sciences, Poznan, Poland; ¹⁰ Maria Sklodowska-Curie Memorial Cancer Center and Institute of Oncology, Gliwice Branch, Radiotherapy Department, Gliwice, Poland

Background: High-dose consolidation therapy with autologous stem cell support (autoSCT) is considered beneficial for transplant-eligible patients with newly diagnosed and relapsed peripheral T-cell lymphomas (PTCL). The evidence is based on retrospective analyses which included heterogeneity of mainly chemotherapy-based conditioning regimens. Achieving complete remission (CR) before autoSCT is one of the strongest predictors of the outcome. There are few data regarding the comparison of Total Body Irradiation (TBI) and non-TBI based approaches, which was the goal of this study.

Methods: The current analysis included patients with PTCL who were autografted in CR. We retrospectively analyzed the outcome of 18 patients conditioned with 12 Gy TBI (given in 3 fractions of 4 Gy on 3 consecutive days) combined with 120mg/kg cyclophosphamide with 35 patients who received high-dose chemotherapy regimens (BEAM (BCNU, etoposide, cytarabine, melphalan), n=25; CBV (BCNU, etoposide, cyclophosphamide), n=8; other, n=2). All patients who received TBI-based regimen were transplanted in a single institution (Department of BMT in Gliwice) whereas the remaining autoSCTs were performed in the other Polish Lymphoma Research Group registered centers. The groups characteristics were as follows (TBI and non-TBI, respectively): the median age at transplantation: 42 years (22-65) and 44 (15-63); PTCL subtype (PTCL NOS-6; AITL-2; ALCL ALK(-)-4, ALK(+)-6 and PTCL NOS-21; AITL-5; ALCL ALK(-)-4, ALK(+)-1, ALK status unknown-4); stage at diagnosis (II-5; III-10; IV-3 and II-5; III-10; IV-20); the median number of pre-transplant chemotherapy lines: 1 (1-2) and 2 (1-3); year of transplantation: 2010-2016 and 1998-2011; the median observation time 43 months (8-85) and 61 (9-199). Patients in the non-TBI group received more pre-transplant lines of therapy and were transplanted in the earlier years. All the remaining factors were comparable between the studied groups.

Results: All patients in the TBI group engrafted whereas one patient in the chemotherapy group did not engraft. There were no early and late non-relapse mortality cases noted in the TBI group. In the chemotherapy group four patients died due to non-relapse causes (2-sepsis; 1-secondary neoplasm; 1-heart failure). The probability of overall survival at 3 years and 5 years after autoSCT was significantly higher for those conditioned with TBI-based regimen compared with chemotherapy-based approach (100% vs 74% and 100% and 66%; p=0.025, respectively). The probability of progression-free-survival at 3 years and 5 years was 94% vs 67% and 81% vs 63%; p=0.15, respectively.

Conclusions: TBI seems to be an effective part of conditioning before autoSCT for PTCL allowing potentially to obtain better outcomes compared with high-dose chemotherapy regimens for patients transplanted in CR.

Conflict of interest: The authors declare no potential conflicts of interest.

P171

Favorable outcome of post-transplantation cyclophosphamide haploidentical peripheral blood stem cell transplantation with a targeted busulfan-based myeloablative conditioning regimen using intensive pharmacokinetic monitoring for pediatric patients

Kyung Taek Hong ¹ , Hyoung Jin Kang ¹ , Jung Yoon Choi ¹ , Che Ry Hong ¹ , Sang Hoon Song ² , Kyung-Sang Yu ³ , In-Jin Jang ³ , Hee Young Shin ¹

¹ Seoul National University Hospital, Seoul National University College of Medicine, Pediatrics, Seoul, Korea, Republic of; ² Seoul National University Hospital, Seoul National University College of Medicine, Laboratory Medicin, Seoul, Korea, Republic of; ³ Seoul National University Hospital, Seoul National University College of Medicine, Pharmacology and Clinical Pharmacology, Seoul, Korea, Republic of

Background: The use of a haploidentical donor for hematopoietic stem cell transplantation has shown promisssing results for hematologic malignancies and nonmalignant diseases, particularly in cases without HLA-matched donor. Unfortunately, data on haploidentical hematopoietic stem cell transplantation (haplo-HSCT) with post-transplantation cyclophosphamide (PTCy), in children, are rare and conflicting. Wer had been used an intensive daily pharmacokinetic (PK) monitoring method for busulfan dosing, and optimized intensity of conditioning regimen by calculating the total exposure of busulfan. Here we applied this targeted-busulfan method using daily PK monitoring to haplo-HSCT using PTCy in pediatric patients, and the safety and outcomes were evaluated.

Methods: We retrospectively studied 34 patients who underwent haplo-HSCT (23 malignant diseases and 11 nonmalignant diseases) at the Seoul National University Children's Hospital, from February 2014 to April 2017. The conditioning regimen of the haplo-HSCT was composed of busulfan, fludarabine (40 mg/m^2 once daily via IV from days -8 to -4), and cyclophosphamide (14.5 mg/kg once daily via IV from days -3 to -2. Busulfan (120 mg/m^2 for patients aged ≥1 year and 80 mg/m2 for patients aged < 1 year) was administered as a starter dose on day -8 and administered once daily, thereafter, and a subsequent targeted dose of busulfan was analyzed according to the daily intensive therapeutic drug monitoring results from days -7 to -5. PTCy (50 mg/kg) was given on day 3 and 4.

Results: The median follow-up period was 23 months (range, 1-47 months), and the median value of the total infused busulfan AUC was 74,078 µg x h/L (range, 67,302-78,478). The median days of neutrophil and platelet engraftment were 15 days and 29 days, respectively, and the cumulative incidence (CI) rates of engraftment were 97.1% and 96.6%, respectively. The CI rates of grade II-IV, grade III-IV acute and extensive chronic graft versus host disease were 38.2%, 5.9%, and 9.4%, respectively. The overall survival rate, event-free survival rate, and treatment-related mortality rate at the time of median follow-up were 84.6%, 79.4%, and 2.4%, respectively. Based on the subgroup analysis of patients with malignancies (n=23), the relapse incidence rate at the time of median follow-up was 21.7%.

Conclusions: Haplo-HSCT using PTCy, along with a targeted busulfan-based myeloablative conditioning regimen and peripheral blood as a stem cell source, was a safe and promising therapeutic option for children with hematologic malignancies and nonmalignant diseases.

Conflict of interest: The authors declare no competing financial interests

P172

Fit Elderly Patients with aggressive lymphoma do not need dose reduction of conditioning regimen

Osman Ilhan, Güldane Cengiz Seval, Atilla Uslu, Sinem Civriz Bozdag, Selami Koçak Toprak, Meltem Kurt Yuksel, Pervin Topcuoglu, Onder Arslan, Muhit Ozcan, Taner Demirer, Hamdi Akan, Meral Beksac, Gunhan Gurman

Ankara University School of Medicine, Department of Hematology and Stem Cell Transplant Unit, Ankara, Turkey

Background: High-dose chemotherapy followed by autologous hematopoietic cell transplantation (auto HSCT) may provide survival benefit in patients with non-hodgkin lymphoma (NHL). Retrospective analyses suggest that the benefit of HDT extends to elderly patients with NHL, which is an important finding considering that the median age at diagnosis is 67 years (range: 65-74) for NHL in United States, using 2010-2014 US SEER data. We aimed to define the efficacy and toxicity of auto HSCT in patients >60 years with NHL.

Methods: From January 2005 through August 2017, data from 26 patients with aggressive NHL above 60 years of age who were eligible for auto HSCT according to geriatric assessment (GA) were evaluated. All these data were obtained from the Ankara University Faculty of Medicine, Department of Hematology and Bone Marrow Transplant Unit. Their diagnosis were as following; 16 diffuse large B cell lymphoma (primary refractory or relapsed disease), 7 mantle cell lymphoma (first complete remission), 2 follicular lymphoma and 1 anaplastic large cell lymphoma. We compared the toxicity profile and outcome between the research group: patient aged 60 years and above and the control group: patient < 60 years.

Results: All of the patients were stage III or IV at diagnosis; ten out of 26 elderly patients had active disease at the time of auto HSCT. The median follow-up was 20.5 months (range, 1-60 mos). Prior to transplantation majority (85%) of the elderly patients received BEAM protocol as conditioning treatment. Bone marrow stem cell was used in only 1 patient None of the patient had mobilization failure, the median peripheral CD34 level was 5.24x10⁶/kg. Forty-eight percent of the patients experienced grade 3-4 mucositis and 77% of the patients had microbiology-documented infection. Sixty-two percent of the patients had diarrhea with median duration of 8 days (range, 5-20 days). Renal toxicity was occurred in 7 (27%) patients while hepatic toxicity in 1(10%) patients. Median time to neutrophil recovery was 10 days (range, 8-18 days) and platelet recovery 11 days (range, 10-32 days). Overall response was obtained from all patients (23% CR). At the time of data collection, 4 patients (15%) of patients' ≥ 60 years have deceased. Relapse (n=3) was the main course of death. The probability of 4-year progression free survival (PFS) and estimated overall survival (OS) in elderly patients were 44.4% and 39.4%, respectively.

Conclusions: Based on this single center study, auto-HSCT is safe and efficacious in the treatment of elderly lymphoma patients. We emphasize the need for further research in order to determine the risk-benefit threshold for HSCT based on age coupled with comorbidity and fragility.

Conflict of interest: None of the authors has anything to disclose

P173

Fludarabin/Busulfan/Thiotepa/ATG is as an effective myeloablative conditioning protocol in adults undergoing unrelated donor stem cell transplantation

Edmund Watson ¹ , Katalin Balassa ^1,2 , Giancarlo Fatobene ³ , Marcin Lubowiecki ¹ , Rebecca Cash ¹ , Daja Barton ¹ , Mandy Ellis ¹ , Lara Rowley ¹ , Denise Wareham ¹ , Claudia Costa ¹ , Rachel Miller ¹ , Robert Danby ¹ , Vanderson Rocha ^1,2,3 , Andy Peniket ¹

¹ Oxford University Hospitals NHS Foundation Trust, Clinical Haematology, Oxford, United Kingdom; ² NHS Blood and Transplant, Oxford, United Kingdom; ³ Universidade de São Paulo, Department of Haematology, Sao Paulo, Brazil

Background: TBI-based myeloablative conditioning (MAC) is well-established in the treatment of haematologic malignancies. However, radiotherapy can cause logistical problems for transplant scheduling and is associated with significant long- and short-term toxicity. Non-TBI based MAC protocols are now available but there are limited outcome data. We report the use of a Fludarabin/Busulfan/Thiotepa/ATG regimen - originally developed for umbilical cord transplantation - in adult unrelated donor haematopoietic stem cell transplantation (HSCT).

Methods: We performed an analysis of fourteen consecutive patients undergoing Thiotepa-based allogeneic HSCT at two institutions. Primary diagnoses were AML (n = 9), MDS (2), ALL (1), extranodal NK/T cell lymphoma (1), and ALCL (1). At conditioning, eleven patients were in complete remission, one had primary refractory AML, one had AML in partial remission, and one had MDS RAEB1. The recipients' age range was from 17 to 52 years (median 27.5); eight were male.

All patients received the myeloablative combination of Thiotepa (10 mg/kg), Busulfan (9.6 mg/kg), Fludarabine (150 mg/m²) and Rabbit ATG (8 mg/kg), adapted from Sanz et al (2012). Thirteen allografts were from a matched unrelated donor (MUD) and one was from a mismatched related donor. Ten allografts were 10/10 HLA allele matched and four were 9/10. Twelve of the recipients were seropositive for EBV prior to transplant.

Results: Median follow-up for live patients was 40 months (range 6 to 59). Neutrophil engraftment occurred in thirteen patients, with a median of nineteen days (from 15 to 30); two patients experienced delayed engraftment (at 29 and 30 days). The patient with primary refractory AML died prior to engraftment.

Overall survival and progression-free survival at three years were 61 and 62%, respectively. Transplant-related mortality was 21% at 100 days and 29% at one year. Five patients died in follow-up: three from sepsis (at 21, 64 and 212 days from transplant), one from disease relapse (at 497 days) and one from EBV-related PTLD (at 76 days). Acute GvHD occurred in six patients (57.1%), with five episodes of Grade I and one of Grade III. Chronic GvHD was reported in two patients (14.3%).

Of the twelve EBV-positive recipients, six (50%) experienced asymptomatic EBV reactivation of at least 10³ and those whose titre reached 10⁶ were all successfully treated with Rituximab (n = 3). An additional three patients (25%) had EBV reactivation and post-transplant lymphoproliferative disorder (PTLD). Two of these achieved CR with radio- or immunochemotherapy, whereas one was diagnosed four days prior to death and was not fit for treatment; this patient had underlying EBV positive NK/T cell lymphoma.

Seven recipients were seropositive for CMV prior to transplant and five (71.4%) experienced CMV reactivation. All CMV seronegative recipients received cells from seronegative donors.

Conclusions: This Thiotepa-based MAC regimen is an effective alternative to TBI-based conditioning in allogeneic HSCT and is associated with successful engraftment and a low rate of GvHD. We describe a high incidence of EBV reactivation and advise careful monitoring of EBV titres in patients treated with this regimen. Overall, our outcomes suggest this is a promising protocol for use in patients requiring myeloablative conditioning.

Conflict of interest: None of the authors has anything to disclose.

P174

Fludarabine with either 8Gy or 12Gy total body irradiation as conditioning regimen before allogenic hematopoietic cell transplantation in patients with AML. A single center experience

Malgorzata Sobczyk-Kruszelnicka ¹ , Malgorzata Ociepa ¹ , Tomasz Czerw ¹ , Wlodzimierz Mendrek ¹ , Jacek Najda ¹ , Maria Sadus-Wojciechowska ¹ , Katarzyna Michalak ¹ , Andrzej Frankiewicz ¹ , Magdalena Glowala-Kosinska ¹ , Agata Chwieduk ¹ , Wojciech Fidyk ¹ , Slawomir Blamek ² , Leszek Miszczyk ² , Jerzy Holowiecki ¹ , Sebastian Giebel ¹

¹ Maria Sklodowska-Curie Institute - Cancer Center, Gliwice Branch, Department of Bone Marrow Transplantation and Oncohematology, Gliwice, Poland; ² Maria Sklodowska-Curie Institute - Cancer Center, Gliwice Branch, Radiotherapy Department, Gliwice, Poland

Background: Although acute myeloid leukemia (AML) is the most frequent indication for allogenic hematopoietic cell transplantation (alloHCT) the optimal conditioning has not been established so far. Among regimens based on total body irradiation (TBI), the combination of TBI 8Gy + fludarabine(Flu) was prospectively compared with TBI 12 Gy + cyclophosphamide, showing similar efficacy while reduced toxicity. In our center patients with comorbidities or older age were treated with TBI 8Gy/Flu while for younger, „fit” patients the dose of TBI in combination with Flu was escalated to 12 Gy. The letter regimen has not been reported. The aim of the stydy was to evaluate its toxicity and efficacy.

Methods: The analysis included 54 patients (women - 31, men - 23), 27 treated with TBI 8Gy/Flu and 27 treated with TBI 12Gy/Flu, between years 2011-2016. The median age was 54 (20 - 67) and 36 (20 - 61), respectively. The disease stage before the procedure was as follows: CR1-44, CR2-6, CR3-1, PR-1 and NR-2. Patients were treated with HSCT from either HLA-matched siblings (n=22) or unrelated donors (n=32). TBI was administered in 4Gy once daily fractions on days (-3), -2, -1. The immunosuppressive therapy consisted of cyclosporine + methotrexate +/- ATG. Peripheral blood was used as a source of stem cells.

Results: Median neutrophil recovery time (ANC> 0.5G / l) was 17 days in the 8Gy group and 16 days in the 12Gy group. The incidence of grade 2-4 acute GvHD was 11% and 15%, while extensive chronic GvHD 0% and 4%, respectively. The rate of grade 3 non-hematological toxicities was 11% and 18%, respectively. Grade 4 adverse events were not reported. The probability of OS after 2 years was 62% (+/- 10%) in the 8Gy group and 80% (+/- 11%) in the 12Gy group (p=0.14). The PFS probability was 58% (+/- 10%) and 77% (+/- 11%) respectively (p=0.16). The rates of relapse were 32% and 7% (p=0.11), while non-relapse mortality was 10% and 13% (p=NS), respectively.

Conclusions: The combination of fludarabine with TBI is characterized by good tolerance and enhancing efficacy. Unless there are any contraindications, the TBI 12Gy seems to be the preferred dose. However, the efficacy and safety of the TBI 12Gy conditioning regimen with fludarabine requires verification in prospective clinical trial.

Conflict of interest: nothing to disclose

P175

Haploidentical hematopoietic stem cell transplantation with post-transplant high-dose cyclophosphamide in high-risk children: A single-center study

Vedat Uygun ¹ , Gülsün Karasu ¹ , Seda Öztürkmen ¹ , Hayriye Daloğlu ¹ , Suar Kılıç ² , Volkan Hazar ² , Akif Yeşilipek ¹

¹ MedicalPark Antalya, Pediatric Bone Marrow Transplantation Unit, Antalya, Turkey; ² MedicalPark Göztepe, Pediatric Bone Marrow Transplantation Unit, İstanbul, Turkey

Background: Graft-versus-host disease (GVHD) remains a major impediment to the widespread applicability of allogeneic blood and marrow transplantation. Haploidentical hematopoetic stem cell transplantations are increasing steadily due to immediate availability of a donor and better results. In this study, we present the results of 74 transplantations of 72 children who received haploidentical HSCT in our center with unmanipulated bone marrow or peripheral blood stem cells and GVHD prophylaxis with posttransplant cyclophosphamide.

Methods: Most patients received a busulfan-based myeloablative conditioning regimen with weight adjusted doses. G-CSF-primed, unmanipulated BM with additional next-day collected PBSC was the most common stem cell source. Posttransplant CY (50 mg/kg IV) was given on days three and five post-HSCT in case the transplantation performed on the same day. However CY was given on the fourth and fifth day if the stem cell infusion was performed on day 0 and 1. CsA or tacrolimus with MMF or MP was also used for GVHD prophylaxis.

Results: The majority of patients were transplanted for malignancy (84%) and the median age was eight yr (range: 0.4-20.1 yr). Seventy (95%) of the 74 patients engrafted, and 67 (95%) of 70 evaluable subjects achieved full donor chimerism by day+30. Only one patient experienced with Grade IV GVHD. Grades II-III acute GVHD developed in thirty-four patients (46%), and chronic GVHD was found in fourteen patients (19%). Fourteen patients died within the first 100 days mostly due to infections (TRM 19%). Forty patients are currently alive, with a median follow-up of 27 months (range 6-56 months). The 12-month OS and DFS were 58% (95% CI: 45.2-68.0) and 52% (95% CI: 39.8-62.8), respectively.

Conclusions: Our results with these high-risk patients are encouraging for haploidentical HSCT in pediatric patients. Given the growing data on the similarity of outcomes after HLA-matched and haploidentical HSCT, further studies are required to determine whether others factors may be more important for donor selection than HLA-matching in pediatric population.

Conflict of interest: Nothing to disclose

P176

Abstract previously published

P177

Impact of CD34⁺ and CD3⁺ cell doses on outcomes in alemtuzumab-based reduced intensity transplantation for high-risk MDS and AML: outcomes of a single UK centre

Nuno M. Borges, Louise Duncan, Erin Hurst

Freeman Hospital, Haematology, Newcastle upon Tyne, United Kingdom

Background: The impact of cellular composition on transplant outcomes remains controversial. Others have assessed the role of cell doses on outcome in mixed populations of myeloablative, reduced intensity and T-cell depleted or non-depleted transplants. In this study we have assessed the impact of CD34⁺ and CD3⁺ doses in a well-defined cohort of MDS/AML patients, the commonest transplant indications. Patients were transplanted using alemtuzumab-based reduced intensity conditioning (RIC) protocols. To our knowledge, this is the first time exclusively alemtuzumab T-cell depleted transplant population outcomes in relation to cellular composition have been evaluated.

Methods: Outcomes pertaining to consecutive MDS/AML patients transplanted with alemtuzumab-based RIC protocols over a 2-year period in a single centre were retrospectively analysed, including demographics, cell doses, engraftment, chimerism, donor lymphocyte infusions (DLI), and graft-versus-host disease.

Results: Between May 2015 and May 2017, 44 patients underwent RIC HSCT for MDS/AML. Twenty-two (50%) were male, and the average age at transplant was 56 years (31-71). Conditioning consisted of fludarabine and melphalan in 22 (50%) of patients, with the remainder receiving busulfan instead of the latter. All patients received T-cell depletion with alemtuzumab, 30mg in sibling (SIB) transplants, 60mg in matched unrelated donor (MUD) transplants, and 90mg in mismatched (< 10/10) unrelated donor transplants, and received ciclosporin for GVHD prophylaxis. Thirty-two (73%) patients received MUD, and 12 (27%) received sibling donor cells. The average infused cell doses were 6x10⁶ (2.2-11.6) and 264x10⁶ (86-660) for CD34⁺ and CD3⁺ cells, respectively. The average time to engraftment was 15.3 days (9-31). Twenty-two (50%) patients received DLI (average cell dose 1.3x10⁶, range 0.1-17), and each patient received on average 1.6 fractions (1-4), either prophylactically (73%) or therapeutically (27%). Fourteen (32%) patients developed acute GVHD grades II-IV and 27 (61%) chronic GVHD (including progressive and de novo). In the acute group, the incidence of grades I, II, and III were 11 (44%), 11 (44%), and 3 (12%), respectively (no patients developed grade IV). In the chronic group, the incidence of NIH severity grades mild, moderate, and severe were 12 (44%), 9 (33%), and 6 (22%), respectively. Overall mortality was 17/44 (39%). Relapse mortality was 6/44 (14%) and non-relapse mortality was 11/44 (25%). There was no significant association between CD34⁺ or CD3⁺ cell dose on outcomes including engraftment, number of DLI infusions, survival or the development of acute or chronic GVHD. There was a trend towards higher GVHD rates in MUDs compared to Sib's, (p=0.08 for acute GVHD, and p=0.74 for chronic GVHD).

Conclusions: In a uniformly treated MDS/AML cohort, we have demonstrated no impact of cell doses on transplant outcomes for the first time in an exclusively RIC alemtuzumab-based transplant group.

Conflict of interest: None of the authors has anything to disclose.

P178

Impact of reduced-intensity conditioning on engraftment and transfusion needs in ABO-incompatible bone marrow allogeneic transplantation

Gala Aglaia Méndez ¹ , Cristina Amunárriz ² , José Luis Arroyo ² , Iñigo Romón ¹

¹ Hospital Universitario Marqués de Valdecilla, Santander, España, Hematology, Santander, Spain; ² Banco de Sangre y Tejidos de Cantrabria, Liencres, Spain

Background: Current evidence on reduced intensity conditioning (RIC) regimens in ABO-mismatched allogeneic transplantation is limited. It is hypothesized that this kind of conditioning provoke a longer survival of receptor´s plasma cells, thus, producing isohemagglutinins against donor erythrocytes, which may cause delayed erythrocyte recovery, higher transfusion needs and pure red cell aplasia (PRCA). Although some studies have supported these associations, there is still controversy surrounding it [1].

We have carried out this study, therefore, to describe our experience in major ABO-mismatched bone marrow (BM) allogeneic transplantation in the setting of myeloablative (MA) and RIC conditionings.

Methods: We retrospectively studied major and bidirectional ABO-incompatible allogeneic BM transplants performed in or center from 2006 to 2017. We analyzed neutrophil and platelet engraftments, red blood cell (RBC) recovery, RBC units transfused until day +100, platelet units transfused until day +60, and PRCA.

Data was analyzed with IBM SPSS software®, version 24. Descriptive statics are reported as median (and range). Shapiro Wilks test was used for testing normality and mean comparisons were performed with Mann Whitney's U test.

Results: A total of 63 transplants were included: 65% MA and 35% (n=22) RIC. The latter were mostly Fludarabine-Melphalan (63.6%) conditionings and, in the remaining, Busulphan-Fludarabine (18%) or schemes based on cyclophosphamide (18%). Diagnoses were mainly AML/MDS (43,4%), ALL (14,3%) and NHL (11.1%). They were HLA identical (n=13), unrelated (n =46) and haploidentical (n 4). An average of 0,13 ml/kg recipient body weight (0.03-0.37) ABO-mismatched erythrocytes were infused, without differences between groups.

Median RBC recovery was 30 days (40 in RIC and 29 in MA, p=0.6 NS). We identified 6 cases of APSR, all of them on group O recipients from a group A donor but, contrary to expectations, mainly in myeloablative group.

Medians to reach neutrophil and platelet engraftments were 19 and 24 days respectively, with 5 platelet units transfused until day +60 and 6 RBC units transfused until day +100. There weren't any differences between both groups.

Conclusions: In our series, we have observed a tendency to a delayed RBC recovery in the RIC group, although not statistically significant. We haven't found differences in terms of transfusion requirements nor in PRCA, unlike some of previous reports. This could be due to the fact that in our reduced intensity conditioning group, it predominated regimens based on alkylating agents, which may present higher activity against host plasma cells than other agents as cyclophosphamide, as previously hypothesized [2].

In conclusion, studies with more homogeneous groups of patients are required to identify which RIC conditionings are more likely to cause immunohematological complications in ABO- incompatible transplants.

References:

1. Hefazy M et al. Transfusion 2016 Feb;56(2):518-27.

2. Kanda et al. Br J Heamatol 2002; 118, 128-131.

Conflict of interest: None of the authors has disclosed any conflicts of interest.

[P178 Figure]

$$$$

P179

Incorporation of GSTA1 genetic variants into a population pharmacokinetic model for IV Busulfan in paediatric hematopoietic stem cell transplantation

Tiago Nava ^1,2 , Nastya Kassir ³ , Mohamed Azis Rezgui ⁴ , Chakradhara Rao S. Uppugunduri ^1,5 , Patricia Huezo-Diaz Curtis ^1,5 , Yves Théoret ⁴ , Liane Esteves Daudt ⁶ , Catherine Litalien ⁴ , Marc Ansari ^1,5 , Maja Krajinovic ⁴ , Henrique Bittencourt ⁴

¹ CANSEARCH Research Laboratory, University of Geneva, Faculty of Medicine, Geneva, Switzerland; ² University Hospital of Geneva/University of Geneva, Department of Pediatrics, Geneve, Switzerland; ³ Certara Strategic Consulting, Montreal, Canada; ⁴ Centre Hospitalier Universitaire Sainte-Justine, Université de Montréal, Montreal, Canada; ⁵ University Hospital of Geneva/University of Geneva, Pediatrics, Geneve, Switzerland; ⁶ Hospital de Clínicas de Porto Alegre, Porto Alegre, Brazil

Background: Busulfan (Bu) is an alkylating agent used before hematopoietic stem cell transplantation (HSCT). In children, it is associated with high pharmacokinetic (PK) variability, in spite of using the intravenous (IV) formulation. As such, population PK (PopPK) models using anthropometric data such as weight and/or age are needed for Bu dosing in children and adolescents. Genetic functional variants in GSTA1, gene coding for (the main metabolizing enzyme of Bu) have been consistently associated with Bu PK and were recently successfully introduced into an adult PopPK model. Hence, the investigators aim was to develop a paediatric PopPK model for IV Bu to investigate the effect of the GSTA1 genotype on PK parameters.

Methods: Bu concentration-time data was collected from 112 patients from 0.1 to 20 years of age (median 5.4 years) who received IV Bu during the conditioning regimen prior to HSCT in a single center. Both q6h and q24h administration schedules were included. Weight, sex, baseline disease (malignant vs. non-malignant), age, conditioning regimen (≥2 alkylating agents vs. others) and GSTA1 haplotypes, grouped as diplotypes, were evaluated as covariates of PK parameters by using non-linear mixed effects analysis. The ability to achieve the target AUC_24h (3,600 - 6,000 µM.min or its equivalent for q6h infusions) after the first dose was assessed by estimating the first dose based on the present PopPK model. Results were compared to other available models in children, including current dosing recommendations from regulatory agencies.

Results: A total of 115 transplantations were included. Raw Bu concentrations from six- (q6h) and eight-timepoint (q24h) PK profiles were available for the analysis, from the first dose (100%) up to the fourth day of drug administration (22%). A one-compartment model with first-order elimination best described the data. Allometric scaling of weight and an age-based factor of Bu metabolism maturation (Fmat) were included in the base model. GSTA1 diplotypes were found to be the only significant covariate of Bu clearance (CL) and explained 27% of the between-subject variability. Bu CL was 7% faster and 12% slower for rapid and poor metabolizers, respectively, in comparison with normal metabolizers. Bu doses calculated using the parameters of the proposed PopPK model were estimated to achieve the target AUC in 85.2% of the cases, against 75% with the currently available most performing models and 37.4% and 42.6% when doses were calculated according to the current recommendations from European (EMA) and American (FDA), respectively.

Conclusions: The present study provides the first PopPK model for IV Bu in children and adolescents that incorporates a genetic background (GSTA1 genetic variants) explaining the variability in PK parameters. The present model, by adapting the treatment to an individualized metabolic potential, intends to minimize fluctuations of the drug exposure, which may decrease the occurrence of treatment-related toxicities in more susceptible individuals. A prospective study is required to validate the current findings.

Conflict of interest: The authors declare no conflict of interest

P180

Intensified conditioning with fludarabine/melphalan plus total body irradiation for acute myeloid leukemia with minimal residual or active disease

Sarah Lindner, Fabian Lang, Tobias Berg, Julia Riemann, Zuzana Jedlickova, Saskia Gueller, Rosa Toenges, Juliane Steinmann, Hans Martin, Hubert Serve, Gesine Bug

University Hospital Frankfurt, Goethe University, Frankfurt, Germany

Background: In acute myeloid leukemia (AML), outcome after allogeneic hematopoietic stem cell transplantation (HSCT) in patients with active or minimal residual disease (MRD) at transplantation remains poor, but may be improved by providing highly efficient antileukemic drug(s) shortly before a reduced intensity conditioning regimen.

Methods: We retrospectively analyzed the impact of melphalan (140mg/m²) followed by fludarabine (120mg/m²) and total body irradiation (TBI, 4-8 Gy; FLU/MEL/TBI, n=27) compared to fludarabine (100-120 mg/m²), cytarabine (4000-8000mg/m²), amsacrine (400mg/m²) in addition to fludarabin (50-60 mg/m²) and intravenous busulfan (6.4-9.6mg/kg; FLAMSA-RIC, n=12) on the outcome of 39 consecutively transplanted AML patients with MRD (FLU/MEL/TBI: n=5) or active disease (FLU/MEL/TBI: n=22; FLAMSA-RIC: n=12) between 01/2011-06/2017. Disease status prior to transplant was primary refractory (n=27) or relapsed AML (n=12). MRD was determined by quantitative polymerase chain reaction (qPCR) using mutated NPM1 (n=4) and CBFB-MYH11 (n=1). Active disease was defined as ≥5% bone marrow (BM) blasts or extramedullary disease. All patients received a T cell-replete, HLA-matched donor graft, mostly peripheral blood stem cells (95%). Graft-versus-host-disease (GvHD) prophylaxis consisted of cyclosporine A and mycophenolate mofetil; additional antithymocyte globulin was applied in the unrelated donor setting.

Results: In the FLU/MEL/TBI group, patients were significantly younger (48 vs. 67 years, p=0.008) and had a significantly lower median Hematopoietic Cell Transplantation-Comorbidity Index (0 vs. 3, p=0.001). Patients in the FLU/MEL/TBI group were more likely to have an adverse-risk karyotype (45% vs. 17%, p=0.049), whereas more patients in the FLAMSA-RIC group presented with secondary AML (58% vs. 10%, p=0.004). Median BM blast count (23 vs. 28%, p=0.907) and time from diagnosis/relapse to transplant (2.5 vs. 2.4 months, p=0.871) did not differ between groups. Six patients (15%) were grafted from a matched related donor, 29 patients (74%) from a 10/10 matched unrelated donor (MUD), 4 patients (10%) from a 9/10 MUD. All patients engrafted. With a median follow-up of 19 months (range, 2-67), 2-year overall survival (OS) was 81±7% and significantly superior in the FLU/MEL/TBI vs. FLAMSA-RIC group (86±6% vs. 66±14%, p=0.043, Figure 1), relapse-free survival was 82±9% vs. 58±14% (p=0.015). Relapse after FLU/MEL/TBI occurred in 4 patients with high-risk genetics (complex aberrant karyotype with loss of TP53 (n=2), RUNX1 mutation (n=1), FLT3-ITD (n=1)) and 1 patient with extramedullary disease. No relapse was seen in patients with MRD positive disease at transplantation. There was no non-relapse mortality after 2 years in both groups. Cumulative incidences of acute GvHD grade 3-4 at day +100 and chronic severe GvHD at 1 year did not differ between groups (3±0% vs. 8±1%, p=0.895 and 11±0% vs. 8±0%, p=0.589).

Conclusions: Our data suggest that addition of TBI to FLU/MEL is well tolerated and associated with favorable OS in patients with active or MRD positive AML, who face a similarly high risk of relapse. Thus, an intensified conditioning regimen such as FLU/MEL/TBI may also be considered for MRD positive AML.

Conflict of interest:

F. Lang received support from the Frankfurter Förderung “Nachwuchswissenschaftler” and the EUTOS funding program, had advisory roles at Novartis, Ariad, Sanofi Aventis, and Bristol-Myers Squibb and received funding from Novartis.

T. Berg received funding by Alexion, Abbvie, Astellas, Celgene and had advisory roles at Riemser.

G. Bug received honoraria from Eurocept and funding by the Alfred und Angelika Gutermuth-Stiftung and the Verein Knochenmarktransplantation/Gentherapie Frankfurt (KGF).

The other authors have nothing to disclose.

[P180 Figure]

$$$$

P181

IV BUSULFAN THERAPEUTIC AUC IN CHILDREN UNDERGOING ALLOGENIC HAEMATOPOIETIC STEM CELL TRANSPLANTATION

Hemalatha Doss ¹ , Adam Gassas ² , Patricia Breslin ² , Michelle Cummins ² , Colin Steward ² , Christina Moris ² , Jessica Fuidge ² , Stephen Palmer ¹ , Ponni Sivaprakasam ²

¹ University of Bristol, Haem/Onc/BMT, Bristol, United Kingdom; ² University of Bristol, Bristol, United Kingdom

Background: In children, age or weight based Busulfan (Bu) dosing is used with drug monitoring to achieve therapeutic area under the curve (AUC). Achieving a specific target is important for the success of allogeneic HSCT and also to reduce hepatic toxicity.

Objectives: 1.This study looks at the accuracy of achieving target AUC with the standard weight based intravenous (IV) Bu dose. 2. The accuracy of achieving target AUC after dose adjustment. 3. Toxicity, in particular veno-occlusive disease (VOD), in relation to Bu AUC.

Methods: Weight based dosing for Busulfan were used as follows: 2mg/kg for < 9kg; 2.4mg/kg for 9-< 16 kg; 2.2mg/kg for 16-23kg; 1.9mg/kg for >23-34kg and 1.6mg/kg for >34kg. Bu was administered every 12 hours as a 3 hour infusion with a maximum of 8 doses. AUC levels were measured after the first dose and Bu dose was adjusted, if necessary. If dose adjustment was made, a repeat level after dose 5 was measured in some patients. The maximum allowed increment was 50% of the first dose. The different AUC target ranges were 70-100mg/L.hr for myeloablative conditioning (MAC), 60-65mg/L.hr for reduced intensity conditioning (RIC) and 45-65mg/L.hr for children with haemophagocytic lymphohistiocytosis (HLH). Study design was a retrospective analysis.

Results: Between October 2014 to November 2017, we identified 34 children with a median age of 8.4 yrs (range: 0.7 yrs -16.2 yrs) who received Bu-based conditioning for allogeneic HSCT. Fourteen out of the 34 patients (41%) achieved target AUC without requiring dose modification. All of these children had no VOD.

Dose modification was done in the remaining 59% children. Out of the 20 children who needed dose modification, 17 required dose increment and had received MAC. 13/17 had repeat AUC to confirm the achievement of final AUC. Nine of the 13 achieved the target AUC based on the repeat level at dose 5. Four out of these 9 patients developed VOD requiring treatment with Defibrotide and 1 patient died from VOD. The remaining 4 out of the 13 patients (30%) didn't achieve the target in spite of maximum dose increment and none of them developed VOD. All patients engrafted except for one child with early death due to severe, progressive VOD at D+18.

Conclusions: Only 41% of the patients achieved target AUC based on the standard weight dosing. Those who required dose increase and achieved AUC target at dose 5 had a high incidence of VOD (45%). A different model of Bu dose adjustment that also includes genetic factors such as the glutathione S-transferase (GST) gene variation may be required for accurate Bu dosing without increasing toxicity.

Conflict of interest: nothing to declare

P182

Long Term Outcome of Patients Receiving Nonmyeloablative Conditioning and Allogeneic Hematopoietic Cell Transplantation for Advanced Haematologic Malignancies

Si Ying Tang ^1,2 , Yeh Ching Linn ³ , Bee Choo Tai ⁴ , Michelle LM Poon ¹ , Lip Kun Tan ¹ , Belinda Tan ¹ , Yelly Lim ¹ , Teck Guan Soh ¹ , Davanaliz Gonzales Ramos ¹ , William YK Hwang ³ , Liang Piu Koh ^1,2

¹ National University Cancer Institute, Department of Haematology-Oncology, Singapore, Singapore; ² National University Hospital, Department of Medicine, Singapore, Singapore; ³ Singapore General Hospital, Department of Haematology, Singapore, Singapore; ⁴ National University of Singapore, Saw Swee Hock School of Public Health, Singapore, Singapore

Background: Nonmyeloablative (NM) hematopoietic cell transplantation (HCT) has extended the potential curative treatment option of allografting to patients in whom it was previously contraindicated due to advanced age or comorbidity.

Methods: From 2001 to 2011, 62 patients aged 16 to 66 years (median age 45 years) with haematologic malignancies, who were poor candidates for conventional HCT, received NM conditioning with fludarabine 90mg/m² and total body irradiation (TBI) 200c-Gy, followed by filgrastim-mobilized peripheral blood stem cell transplant from HLA identical (n=57), matched unrelated (n=2), 5/6 mismatched sibling (n=3) donors. A calcineurin inhibitor and mycophenolate mofetil with/without methotrexate were used for postgrafting immunosuppression. Diagnoses include acute leukemia (n=18), myelodysplastic syndrome and myelofibrosis (n=19), lymphoma (n=7), CML (n=8) and myeloma (n=10).

Results: With a median follow up of 10 years (range 0.9 to 13.4 years) in surviving patients, the 10-year overall survival (OS) was 43%, Based on disease risk index (DRI), the OS was 61%, 47% and 11% for patients with low risk, intermediate risk and high/very high risk group respectively. The day 120 cumulative incidence of acute graft versus host disease (GVHD) grade 2-4, and grade 3-4 were 40% and 19%, respectively. The day 60 cumulative incidence of chronic GVHD was 41%. At 5 years, non relapse mortality (NRM) rate was 19%, and relapse mortality rate (RRM) was 37%. Most NRM was due to GVHD. Patients with high or very high risk DRI have the highest risk of RRM (HR 6.96; 95%CI 1.13-43.1; p=0.037). In multivariate analysis, improved OS was seen in patients with low risk group according to the DRI (HR 0.2, 95%CI 0.12-0.68; p=0.005).

Conclusions: Allogeneic NM conditioning using fludarabine and low dose TBI, relying almost exclusively on graft-vs-tumour effect, can result in long term remission but it is still limited by high incidence of relapse, especially in high-risk diseases. Novel approaches should be explored to reduce early disease progression after HCT without increasing regimen-related toxicity.

Conflict of interest: [S. Ying Tang]: nothing to disclose

P183

Long Term Survival Outcomes of AML Patients Undergoing Hematopoietic Cell Transplantation Using FTBI-based Myeloablative Conditioning: A Single Institution Experience

Amandeep Salhotra ¹ , Susanta Hui ² , Dongyun Yang ³ , Joycelynne Palmer ³ , Sally Mokhtari ³ , Haris Ali ¹ , Ibrahim Aldoss ¹ , Ahmed Aribi ¹ , Matt Mei ¹ , Sandhu Karamjeet ¹ , Monzr Al Malki ¹ , Sameer Khaled ¹ , David Snyder ¹ , Margaret O'Donnell ¹ , Joseph Rosenthal ¹ , Vinod Pullarkat ¹ , Jeffrey Wong ² , Ryotaro Nakamura ¹ , Guido Marcucci ¹ , Stephen Forman ¹ , Anthony Stein ¹

¹ City of Hope National Medical Center, Hematology/HCT, Duarte, CA, United States; ² City of Hope National Medical Center, Radiation Oncology, Duarte, CA, United States; ³ City of Hope National Medical Center, Biostatistics and Population Sciences, Duarte, CA, United States

Background: Fractionated total body irradiation (FTBI) combined with Cyclophosphamide/etoposide or intravenous (IV) busulfan combined with Fludarabine/ Cyclophosphamide, are currently used as myeloablative conditioning regimens for allogeneic hematopoietic cell transplantation (alloHCT) for patients with acute myeloid leukemia (AML). Recent data has suggested that IV busulfan may be a preferred myeloablative regimen in patients with my