Hematopoietic stem cell transplantation is usually performed without considering the ABO compatibility between donor and recipient. There are few studies analyzing ABO matching impact on transfusion outcome of umbilical cord blood transplantation (UCBT) recipients. The aim of this study was to analyze factors influencing transfusion outcome, highlighting the ABO matching between donor and recipient. This study has reviewed data from 318 patients who underwent single unit UCBT at la Fe University Hospital from January 2000 to December 2014. There were no differences between RBC and platelet (PLT) requirements or RBC and PLT transfusion independence according to ABO matching between donor and recipient. RBC and PLT requirements were statistically correlated (ρ=0,841, P<0.001). A total of 170 and 188 patients achieved RBC and PLT independence, respectively, within 180 days after UCBT. Persistence of recipient isoagglutinins was detected in 6.8% of patients with major ABO incompatibility at median of 176 days (103–269) after UCBT. Autoimmune haemolytic anemia was diagnosed in 15 patients, 12 of them due to cold antibodies. In conclusion, ABO matching has not influenced transfusion requirements of patients undergoing UCBT.
Hematopoietic stem cell transplantation (HSCT) is a common treatment for patients affected by hematologic malignancies and congenital diseases of the immune system.1 Umbilical cord blood transplantation (UCBT) is an HSCT modality that has some differences as compared with other stem cell sources. Delayed platelet (PLT) recovery and more prolonged blood product transfusion requirements have been described in UCBT, as compared with peripheral blood HSCT.2 Blood product transfusion is especially important, therefore, in the supportive treatment during the aplastic period after UCBT.
As opposed to solid organ transplantations, HSCT is usually performed without considering the ABO compatibility between donor and recipient.3, 4 In previous studies, major ABO incompatibility led to greater blood product requirements in recipients of bone marrow and peripheral blood, for both myeloablative and nonmyeloablative transplants.5, 6, 7, 8 Nevertheless, there are few studies analyzing the ABO matching impact on transfusion outcome of UCBT recipients.9, 10
Specific immunohematologic complications have been described in major, minor and bidirectional ABO incompatible hematopoietic transplants.11, 12 Although minor ABO incompatibility causes passenger lymphocyte syndrome (PLS), in major ABO incompatibility, persistence of recipient isogglutinins can occur. These events are the cause of immune hemolysis after transplantation and consequently may increase the transfusion requirements.
The aim of this study is to analyze factors influencing transfusion outcome, highlighting ABO matching between donor and recipient, in the largest series published to date, of adult patients who underwent single-unit UCBT at a single institution. In addition, immunohematologic events and their impact on transfusion were also reviewed.
Materials and methods
This study has reviewed data from 318 patients who underwent single unit UCBT at la Fe University Hospital from January 2000 to December 2014.
Transplantation procedures including conditioning regimes, GvDH prophylaxis and supportive care had been published previously.13 Briefly, two different myeloablative conditioning (MAC) regimens based on the combination of thiotepa, busulphan, cyclophosphamide (CY) or fludarabine and anti-thymocyte globulin were administered. Patients unfit to receive MAC were treated with a reduced intensity conditioning (RIC) regimen including thiotepa, fludarabine or CY, busulphan (BU) and anti-thymocyte globulin (ATG).
Packed RBCs were transfused if the hemoglobin level of patients was <80 g/dL and prophylactic pooled PLTs were transfused at PLT counts <20 × 109/L. All blood products were leukocyte filtered and irradiated with 25 Gy. The ABO type of RBCs and PLTs to be transfused after the UCBT was determined according to the donor and recipient blood groups, as described previously.3 Major incompatibility is defined when the recipient plasma has isohemagglutinins against recipient RBC antigens and minor when the donor has isohemagglutinins directed against recipient RBC antigens. Bidirectional incompatibility shares features of both major and minor ABO incompatibility. For minor ABO-incompatible transplantations donor-type RBCs and recipient-type PLTs were transfused, whereas for major ABO-incompatible recipients recipient-type RBCs and donor-type PLTs were transfused. For bidirectional ABO-incompatible patients, group O RBCs and group AB PLTs were scheduled. Modifications of blood products such as washing or removing plasma were not performed. PLT transfusion independence was defined as the last PLT transfusion, with no PLT transfusions in the following 7 days. RBC transfusion independence was defined as the last transfusion, with no RBC transfusion in the following 30 days. We recorded transfusion requirements at 30, 90, 180 and 365 days after UCBT.
ABO, Rh typing and the indirect antiglobulin test were determined in the automated system ORTHO Autovue Innova (Ortho Clinical Diagnostics, Marlow, UK). Sera of patients with positive results were tested for specificity using panels of RBC with known antigens (Ortho Clinical and Diagnostics, Diamed and Makropanel). Patients who developed antibodies to RBC antigens, considered clinically significant, received blood that was negative for the identified antigen.
Direct antiglobulin test (DAT) was systematically performed as part of the pre-transfusion compatibility tests after transplantation. The DAT was performed on RBC from EDTA acid samples according to standard methods. RBC eluates were prepared by acid elution (Elu-Kit II, Gamma Biologicals, Houston, TX, USA). Eluates and sera of patients with positive results were tested for specificity using panels of RBC with known antigens (Ortho Clinical and Diagnostics, Diamed and Makropanel).
Definitions of immunohematologic events
Transfusion alloimmunization to RBCs, PLTs or hematopoietic progenitors was defined as the appearance of alloantibodies against RBC antigens different to anti-A or anti-B, at least a week after the sensitisation event.
Persistence of recipient isogglutinin was defined as patients having recipient anti-A and/or anti-B in plasma or on RBCs more than 2 months after a major ABO mismatched UCBT. PLS consists of the production of antibodies by the donor passenger lymphocytes transplanted with the graft, against the recipient RBC antigens (mainly anti-A and/or anti-B) during the first month after UCBT.
Autoimmune haemolytic anemia (AIHA) was diagnosed if patients had a positive DAT, a positive indirect antiglobulin test with broad reactivity to RBC in serum and/or eluate, and clinical and laboratory evidence of hemolysis.
Computer software SPSS (version 15, SPSS Inc., Chicago, IL, USA) and EZR, a graphic user interface for R 2.13.0,14 were used to perform the statistical analysis when applicable. Descriptive statistics was presented for variables. The Kolmogorov–Smirnov test was employed to investigate the normality distribution of the variables. Results are expressed as mean±s.d. or median and range for continuous variables, and as numbers with percentages for categorical variables.
Categorical variables were compared by means of the χ2-test or the Fisher’s exact test. Correlation among variables was assessed by Spearman’s coefficient. The Mann–Whitney U-test or the Kruskal–Wallis tests for continuous variables was used to compare the groups when applicable. Stepwise multiple logistic regression was used to investigate the risk factors on transfusion requirement at 90 days and transfusion independence after UCBT. RBC units and PLT concentrates transfused were converted from a continuous variable to a categorical variable for this purpose. The forward variable selection was employed. The cumulative incidence of RBC and PLT transfusion independence was calculated in a competing risk setting, death and relapse being treated as competing events. Multivariable analysis to assess the independent effect of variables on time to RBC and PLT transfusion independence was performed using the Fine and Grey proportional hazard model for competing events. The multivariable analysis included: age, age (<55 years vs 55 years or more), sex (female vs male), ABO matching (identical, minor and major), diagnosis (acute myeloid leukemia vs others), total nucleated cells7/kg, CD345+ cells/kg, HLA matching for class I and II, conditioning regimen (fludarabine vs CY), conditioning regimen (MAC vs RIC), acute GvHD (yes vs no), days until PLT >20 000/μL, days until PMN >1000/μL, immunohematologic events (yes vs no), CMV serostatus (negative vs positive) and EBV serostatus (negative vs positive). All reported values are two-sided and P<0.05 was considered significant.
Patient data and follow-up
The study reviewed 318 patients undergoing UCBT for a 15-year period. Table 1 shows the characteristics of patients according to the ABO matching between donor and recipient. Patients with bidirectional incompatibility (n=11) were included in the major incompatibility group. There was no statistical difference for any variable, except for total nucleated cells × 107 infused that was higher for the ABO identical group. Within the first 30 days after transplantation, 23 patients died (11 ABO identical, 5 minor incompatible and 7 major incompatible). After a median follow-up of 1799 days (range 54–5399), 101 patients are alive (31.6%). Thirty-six patients have relapsed within 180 days after transplantation.
Transfusion requirements at different time points after UCBT are showed in Table 2. There were no differences between RBC and PLT requirements according to ABO matching between donor and recipient. Fresh frozen plasma was transfused only to 28 patients (8.7%) at 30 days after UCBT. Median and range of fresh frozen plasma units transfused were 2 and 1–9, respectively. A total of 170 (86.3%) and 188 (95.4%) patients of 197 surviving patients achieved RBC and PLT independence, respectively, within 180 days after UCBT. Median time and range to RBC and PLT transfusion independence was 39 (0–180) and 43 (14–180) days, respectively. Results showed no differences between ABO matching groups. Figures 1 and 2 show the cumulative incidence of RBC and PLT transfusion independence at 180 days after UCBT, respectively, in patients receiving ABO identical, minor and major ABO incompatible grafts.
Patients who received RIC had less RBC (6 vs 8) and PLT (14 vs 20) transfusion requirement median than those who received the MAC regimen (P<0.05), only at 30 days after UCBT. However, median time to reach neutrophils count >1 × 109/L was 22 days and 19.5 days for patients receiving MAC regimen and RIC regimen, respectively (P=0.075), whereas median time to reach PLT count >20 × 109/L was 50 (10–188) days and 35 (20–184) for patients receiving MAC regimen and RIC regimen, respectively (P=0.024).
Table 3 shows the multivariable analysis of risk factors for RBC and PLT transfusion 90 days after UCBT. RBC and PLT requirements were statistically correlated (ρ=0,841, P<0.001). Table 4 shows the variables that significantly influenced RBC and PLT transfusion independence. There were no differences in hematopoietic engraftment between patients with AML or with the other diseases.
Positive indirect antiglobulin test before UCBT was detected in six (1.9%) patients: one cold antibody, 3three anti-D, one anti-K and one anti-Lea. No RBC alloimmunization was detected in any patient after UCBT.
Of 48 Rh (D)-negative patients, only 5 patients received Rh (D)-negative hematopoietic progenitors. Thirty days after UCB, 40 Rh (D)-negative patients received a median of 10.5 (range 3–102) Rh (D)-positive PLT concentrates. One year after UCBT, 10 Rh (D)-negative patients had received a median of 31 (range 3–87) Rh (D)-positive PLT. Neither alloantibody against the Rh system was developed.
Of the 88 patients with major ABO incompatibility, 6 (6.8%) had detectable anti-A/anti-B (5 anti-A and 1 anti-B) more than 2 months after UCBT. Persistence of recipient isoagglutinins was detected at a median of 176 days (103–269) after UCBT. Median and range of hemoglobin, bilirubin and serum lactic dehydrogenase (LDH) were 6.8 g/dL (6.2–8.1), 3.0 (1.3–8.9) and 1685 (1090–2570), respectively. With a median follow-up of 463 days, two patients were alive. There were no differences in survival (P=0.612), PLT or RBC transfusion independence, or in PLT and RBC transfusion requirements at any time after UCBT between patients who had persistence of recipient isogglutinins and those not having.
AIHA was diagnosed in 15 patients, 12 of them due to cold antibodies. One case shared AIHA and persistence of anti-A. Our group recently published clinical and serological features of these patients.13 Transfusion requirements of patients who developed AHAI were higher only for RBC at 365 after UCBT (35.6±22.8 vs 20.6±19.4, P=0.014).
This is the largest study specifically analyzing transfusion outcomes in patients undergoing UCBT over a long time period. As published to date, transfusion requirements are higher for patients receiving UCBT than for those transplanted with other sources of hematopoietic progenitors.2, 7 Differential characteristics of umbilical cord blood as much less total nucleated cell and CD34+ cell content and, consequently, longer time for hematopoietic engraftment support this fact.15 ABO-mismatched HSCT mostly from peripheral blood has been associated with increased transfusion requirements in many studies,5, 6, 7, 8, 16 whereas few reports have shown similar transfusion needs for ABO-matched and ABO-mismatched hematopoietic transplants.17, 18 This issue, therefore, remains controversial. To our knowledge, there is only one previous report focusing on ABO incompatibility between donor and recipient, and transfusion outcome in patients who underwent UCBT. Tomonari et al.9 collected data from 95 patients and concluded that ABO incompatibility between donor and recipient was associated with more RBC and PLT transfusions, and delayed PLT engraftment after UCBT. Our results for a higher number of patients, surprisingly, show the contrary: ABO matching has no influence on transfusion requirements. Use of different protocols for conditioning regimen and GvHD prophylaxis also could explain the differences between centers. Nevertheless, multivariable analysis revealed some interesting factors that significantly increased both RBC and PLT transfusion needs: days until neutrophils >1 × 109/L and the presence of acute GvHD. Ozkurt et al.17 reported more frequent PLT transfusion requirements in patients who received peripheral blood HSCT and developed acute GvHD and other transplant-related toxicities. Acute GvHD is a severe inflammatory complication of allogeneic HSCT targeting the skin, gastrointestinal tract and liver. The hematopoietic system can also be targeted producing lymphopenia and thrombocytopenia.19
Our analysis shows two protective factors of transfusion: use of fludarabine instead of CY in the conditioning regimen before UCBT and diagnosis of AML. In fact, the use of fludarabine significantly reduced the RBC and PLT transfusion needs of patients. When comparing FLU and CY regimens in each subset MAC and RIC UCBT, the number of days to PMN and PLT engraftment was not statistically significant (data not shown). It is well known that the use of higher dose CY is associated with an incidence of hemorrhagic cystitis exceeding 20%.20 To our understanding, this is the main reason why patients who received CY-containing conditioning regimen received more transfusions. Lower transfusion requirements in patients undergoing HSCT receiving RIC regimen has also been published,6 although this has not been demonstrated in UCBT. Our results show higher RBC and PLT requirements only at 30 days after UCBT for patients receiving the MAC regimen; however, in multivariable analysis this factor did not reach statistical significance. AML as compared with the rest of the diagnosis was the second factor that significantly decreased RBC and PLT transfusions. Our hypothesis explaining this is the better condition of this subset of patients before UCBT. AML diagnosis also increases the probability to reach the RBC and PLT transfusion independence, as shown in Table 4.
When analysing factors influencing transfusion independence, it is important to note that total nucleated cell content and CD34+ content significantly and inversely influenced RBC and PLT transfusion independence, respectively. Our data, therefore, emphasize the importance of cell content when selecting a cord blood unit for transplantation, especially when PLT transfusions have been associated with adverse clinical outcomes after HSCT in previous reports.21 There is a negative correlation between age and RBC and PLT independence. In previous studies, age has also been negatively correlated with overall survival.22 Other factors that have been considered to influence transfusion such as CMV seropositivity did not show an influence on transfusion in our study.9 Differences in transplantation procedures and characteristics of patients can contribute in part to explain these differences among studies.
Patients with ABO mismatched bone marrow or peripheral blood HSCT are at high risk for RBC haemolysis.6, 23 In ABO major incompatible transplants, persistence of recipient isogglutinins has been detected in 13% of patient. Pure red cell aplasia and longer RBC requirements have also been related to this condition.11 To our knowledge, the incidence of this problem in UCBT has never been previously published. We detected 6.8% of cases, less than previously published for other hematopoietic progenitors sources and this condition did not influence transfusion requirements.
PLS occurs in patients receiving hematopoietic stem cells from a minor ABO mismatched donor. This syndrome has been attributed to the lymphocytes, which are infused with the stem cell product, which proliferate and produce antibodies against recipient RBCs.24 PLS is related to the lymphocytes burden infused with the graft. Snell et al.25 reported 15 of 24 ABO minor incompatible peripheral blood HSCT presenting hemolysis between 7 and 13 days posttransplant due to this condition. We have not detected a single case of PLS in ABO minor incompatible UCBT patients. In this context, several considerations should be highlighted: first, that cord blood has a decreased number of lymphocytes and the B-lymphocytes are predominantly naive B phenotype.26 Second, previous publications have shown a lack of isohemagglutinin production following minor ABO incompatible UCBT and thus no evidence of hemolysis has been observed in 14 cases.25 For all that, one can expect fewer immune complications in patients undergoing UCBT due to ABO incompatibility. However, cold antibody AIHA remains as the main immune complication, our group13 having reported a cumulative incidence of 5.4%.
In summary, graft cell count is much more important than ABO matching in determining transfusion outcome of patients undergoing single-unit UCBT. As expected theoretically, immune hemolysis due to ABO incompatibility is a very rare adverse effect, whereas AIHA is the most relevant immune complication after UCBT.
Sanz J, Sanz MA, Saavedra S, Lorenzo I, Montesinos P, Senent L et al. Cord blood transplantation from unrelated donors in adults with high-risk acute myeloid leukemia. Biol Blood Marrow Transplant 2010; 16: 86–94.
Solh M, Brunstein C, Morgan S, Weisdorf D . Platelet and red blood cell utilization and transfusion independence in umbilical cord blood and allogeneic peripheral blood hematopoietic cell transplants. Biol Blood Marrow Transplant 2011; 17: 710–716.
Booth GS, Gehrie EA, Bolan CD, Savani BN . Clinical guide to ABO-incompatible allogeneic stem cell transplantation. Biol Blood Marrow Transplant 2013; 19: 1152–1158.
Rowley SD, Donato ML, Bhattacharyya P . Red blood cell-incompatible allogeneic hematopoietic progenitor cell transplantation. Bone Marrow Transplant 2011; 46: 1167–1185.
Watz E, Remberger M, Ringden O, Lundahl J, Ljunqman P, Mattson P et al. Analysis of donor and recipient ABO incompatibility and antibody-associated complications alter allogeneic stem cell transplantation with reduced-intensity conditioning. Biol Blood Marrow Transplant 2014; 20: 264–271.
Wang Z, Sorror ML, Leisenring W, Schoch G, Maloney DG, Sandmaier BM et al. The impact of donor type and ABO incompatibility on transfusion requirements after nonmyeloablative hematopoietic cell Transplantation. Br J Haematol 2010; 149: 101–110.
Blin N, Traineau R, Houssin S, Peffault de Latour R, Petropoulou A, Robin M et al. Impact of donor-recipient major ABO mismatch on allogeneic transplantation outcome according to stem cell source. Biol Blood Marrow Transplant 2010; 16: 1315–1323.
Le Viellez A, P´Ng S, Buffery S, Wright M, Cooney J, Cannell P et al. Red cell and platelet transfusion burden following myeloablative allogeneic haematopoietic stem cell transplantation. Intern Med J 2015; 45: 1286–1292.
Tomonari A, Takahashi S, Ooi J, Tsikada N, Konuma T, Kobayashi T et al. Impact of ABO incompatibility on engraftment and transfusion outcome after unrelated cord blood transplantation: a single institute experience in japan. Bone Marrow Transplantation 2007; 40: 523–528.
Konuma T, Kato S, Ooi J, Oiwa-Monna M, Ebihara Y, Mochizuki S et al. Effect of ABO blood group incompatibility on the outcome of single-unit cord blood transplantation after myeloablative conditioning. Biol Blood Marrow Transplant 2014; 20: 577–592.
Petz LD. . Immune hemolysis associated with transplantation. Semin Hematol 2005; 42: 145–155.
Hoffman PC . Immune hemolytic anemia-selected topics. Hematology Am Soc Hematol Educ Program 2009, 80–86.
Sanz J, Arango M, Carpio N, Montesinos P, Moscardo F, Martin G et al. Autoimmune cytopenias after umbilical cord blood transplantation in adults with haematological malignancies: a single-center experience. Bone Marrow Transplant 2014; 49: 1084–1088.
Kanda Y . Investigation of the freely available easy-to-use software ‘EZR’ for medical statistics. Bone Marrow Transplant 2013; 48: 452–458.
Muñoz J, Shah N, Rezvani K, Hosing C, Bollard CM, Oran B et al. Concise review: umbilical cord blood transplantation: past, present and future. Stem Cells Transl Med 2014; 3: 1435–1443.
Canals C, Muñiz-Diaz E, Martinez C, Martino R, Moreno I, Ramos A et al. Impact of the ABO incompatibility on allogeneic peripheral blood progenitor cell transplantation after reduced intensity conditioning. Transfusion 2004; 44: 1603–1611.
Ozkurt ZN, Yegin ZA, Yenicesu I, Aki SZ, Yagci M, Sucak GT. . Impact of ABO-incompatible donor on early and late outcome of hematopoietic stem cell transplantation. Transplantation Proceedings 2009; 41: 3851–3858.
Dahl D, Hahn A, Hoenecke C, Heuft HG, Dammann E, Stadler M et al. Prolonged isolated red blood cell transfusion requirement after allogeneic blood stem cell transplantation: identification of patients at risk. Transfusion 2010; 50: 649–655.
Holtan SG, Pasquini M, Weisdorf DJ . Acute graft-vs-host disease: a brench-to-bedside update. Blood 2014; 124: 363–373.
Ayas M, Iddiqui K, Al-Jefri A, El-Sohl H, Al-Ahmari A, Khairy A et al. Factors affecting the outcome of related allogeneic hematopoietic cell transplantation in patients with Fanconi anemia. Biol Blood Marrow Transplant 2014; 20: 1599–1603.
Christou G, Kekre N, Petrcich W, Tokessy M, Neurath D, Giulivi A et al. Impact of platelet transfusion on toxicity and mortality after hematopoietic progenitor cell transplantation. Transfusion 2015; 55: 253–258.
Berglund S, LeBlanc K, Remberger M, Gertow J, Uzunel M, Svenberg P et al. Factors with an impact on chimerism development and long-term survival after umbilical cord blood transplantation. Transplantation 2012; 94: 1066–1074.
Sokol RJ, Stamps R, Booker DJ, Scott FM, Laidlaw ST, Vandenberghe EA et al. Posttransplant immune-mediated hemolysis. Transfusion 2002; 42: 198–204.
Salmon JP, Michaux S, Hermanne JP, Baudoux E, Gerard C, Sontag-thull D et al. Delayed massive immune hemolysis mediated by minor ABO incompatibility after allogeneic peripheral blood progenitor cell transplantation. Transfusion 1999; 39: 824–827.
Snell M, Chau C, Hendrix D, Fox R, Downes KA, Creger R et al. Lack of isohemagglutinin production following minor ABO incompatible unrelated HLA mismatched umbilical cord blood transplantation. Bone Marrow Transplantation 2006; 38: 135–140.
Theilgaard-Monch K, Jensen KR, Palm H. . Flow cytometric assessment of lymphocytes subsets, lymphoid progenitors, and hematopoietic stem cell in allogeneic stem cell grafts. Bone Marrow Transplant 2001; 28: 1073–1082.
We thank to Mr P Gutteridge for his assistance in the revision of the manuscript.
The authors declare no conflict of interest.
About this article
RBC and platelet transfusion support in the first 30 and 100 days after haploidentical hematopoietic stem cell transplantation
Comparison of transfusion requirements in adult patients undergoing Haploidentical or single‐unit umbilical cord blood stem cell transplantation
European Journal of Haematology (2019)
Red blood cell transfusion burden by day 30 predicts mortality in adults after single-unit cord blood transplantation
Bone Marrow Transplantation (2019)