European LeukemiaNet-defined primary refractory acute myeloid leukemia: the value of allogeneic hematopoietic stem cell transplant and overall response

We sought to appraise the value of overall response and salvage chemotherapy, inclusive of allogeneic hematopoietic stem cell transplant (AHSCT), in primary refractory acute myeloid leukemia (prAML). For establishing consistency in clinical practice, the 2017 European LeukemiaNet (ELN) defines prAML as failure to attain CR after at least 2 courses of intensive induction chemotherapy. Among 60 consecutive patients (median age 63 years) correspondent with ELN-criteria for prAML, salvage was documented in 48 cases, 30/48 (63%) being administered intensive chemotherapy regimens and 2/48 consolidated with AHSCT as first line salvage. 13/48 (27%) attained response: CR, 7/13 (54%), CRi, 2/13 (15%), MLFS, 4/13 (31%). The CR/CRi rate was 9/48 (19%), with CR rate of 7/48 (15%). On univariate analysis, intermediate-risk karyotype was the only predictor of response (44% vs 17% in unfavorable karyotype; P = 0.04). Administration of any higher-dose (>1 g/m2) cytarabine intensive induction (P = 0.50), intensive salvage chemotherapy (P = 0.72), targeted salvage (FLT3 or IDH inhibitors) (P = 0.42), greater than 1 salvage regimen (P = 0.89), age < 60 years (P = 0.30), and de novo AML (P = 0.10) did not enhance response achievement, nor a survival advantage. AHSCT was performed in 12 patients with (n = 8) or without (n = 4) CR/CRi/MLFS. 1/2/5-year overall survival (OS) rates were 63%/38%/33% in patients who received AHSCT (n = 12) vs 27%/0%/0% in those who achieved CR/CRi/MLFS but were not transplanted (n = 5), vs 14%/0%/0% who were neither transplanted nor achieved CR/CRi/MLFS (n = 43; P < 0.001); the median OS was 18.6, 12.6 and 5.6 months, respectively. Although CR/CRi/MLFS bridged to AHSCT (n = 8), appeared to manifest a longer median OS (20 months), vs (13.4 months) for those with no response consolidated with AHSCT (n = 4), the difference was not significant P = 0.47. We conclude AHSCT as indispensable for securing long-term survival in prAML (p = 0.03 on multivariate analysis), irrespective of response achievement.


INTRODUCTION
The 2017 European LeukemiaNet (ELN) characterizes primary refractory acute myeloid leukemia (prAML) as the absence of complete remission with count recovery (CR), after at least 2 courses of intensive induction chemotherapy [1]. Current literature on prAML is confounded by non-uniform disease definitions and study populations inclusive of relapsed AML [2][3][4][5][6][7]. AHSCT is proclaimed the best and only curative treatment option [2][3][4][5][6]. In the absence of controlled evidence for guidance, current prAML treatment is focused on achieving CR or CR without count recovery (CRi) first, through administering a panoply of salvage regimens, possibly affecting transplant candidacy. A 2020 report from MD Anderson Cancer Center (MDACC), cited rates of CR/CRi for AML refractory to one course of intensive higher dose (>1 g/ m 2 ) induction, as 11%, 5%, and 2%, after treatment with one, two, and three salvage regimens, respectively [8]. In general, response in prAML has been influenced by younger age (<60 years), favorable karyotypes, de novo context, blast percentage prior to salvage, and intensive salvage regimens [9][10][11]. In the current retrospective case series of 60 consecutive patients with ELNdefined prAML, we clarify the value and futility of salvage chemotherapy; identifying AHSCT as invaluable for procuring longterm survival regardless of response category prior to transplant.

METHODS
After approval from our institutional review board, the Mayo Clinic, Rochester, MN, AML database (n = 1800; 2004-2021) was queried to identify 60 patients conforming to the ELN 2017 definition of prAML [1]. All patients displayed persistent leukemia with >5% bone marrow blasts after 2 courses of ELN-classified intensive induction chemotherapy. Supplementary Table 1 delineates these intensive chemotherapy induction regimens.
Therapeutic response categories included CR, CRi, and morphologic leukemia-free state (MLFS) pre-transplant or in those not proceeding to transplant, assessed until death. CR entailed <5% bone marrow (BM) blasts, absence of blasts with Auer rods, absence of extramedullary disease, absolute neutrophil count (ANC) ≥ 1000/μl, and platelet count ≥100,000/μl. CRi required all criteria for CR, but with either ANC < 1000/μl or platelet count <100,000/μl (not both). MLFS indicated BM < 5%, absence of extramedullary disease, absence of blasts with Auer rods, no hematologic recovery required, and BM cellularity of at least 10% or enumeration of 200 cells. No response indicated the absence of CR, CRi, or MLFS pre-transplant or in those not bridged to transplant, until death. Among patients who achieved morphologic remission (<5% BM blasts), minimal residual disease (MRD) was assessed by multiparameter flow cytometry and all MRD positive patients (n = 6) had MRD greater than 0.1% (n = 3/6, had MRD ≥ 1%). Date of last follow-up or death was updated until April 2021. Progression-free survival was measured from the date of CR/CRi/MLFS to the date of relapse or date of death/last follow up. Overall survival (OS) indicated date of AML diagnosis to death (regardless of cause) or last follow up. Statistical analysis was with JMP Pro 16.0.0 (SAS Institute, Cary, NC, USA). Categorical variables were assessed by chi-squared; continuous variables with Wilcoxon rank-sum test. The Kaplan-Meier method computed OS, with Cox proportional hazard models estimating covariate impact upon survival. P < 0.05 was considered significant.
Patients receiving only less intensive salvage (16/48, 33%) had unfit status or prolonged complications from chemotherapyinduced cytopenias, such as transfusion dependence and neutropenic infections. In excluding the 2 patients bridged directly to transplant, on univariate and multivariate, median survival after any intensive salvage chemotherapy was 8.5 months vs. 7.4 months for only less intensive/targeted therapy (P = 0.08). The majority with only less intensive/targeted therapy received only 1 salvage regimen 14/16 (88%) vs 10/30 (33%) having intensive salvage (P = 0.001) In comparison to patients   unfavorable cytogenetics (P = 0.47), FLT3/NPM1 status (P = 0.22), age ≥60 years (P = 0.31), primary vs. secondary AML at diagnosis (P = 0.14), % BM blasts prior to salvage (P = 0.24), and attainment of response CR/CRi/MLFS (P = 0.72) with greater than 1 salvage (P = 0.30). MRD evaluation after response in this cohort was limited, due to earlier period of AML diagnoses when MRD statuses were not routinely performed. MRD was assessed in 7/13 (54%) patients who achieved either CR (n = 5)/CRi (n = 1) /MLFS (n = 1). Only 1 of these 7 patients was CR-MRD Negative and bridged to transplant, the remaining 6 were MRD Positive (n = 2 were bridged to transplant). Irrespective of bridging to AHSCT, the univariate analysis did not display a significant difference in median survival P = 0.26. Median survival of MRD Positive patients was 14.8 months, the median survival of the n = 1 MRD Negative patient has not yet been reached. Similarly, there was no difference in median overall survival between MRD Positive >1% (n = 3, median 15 months) or MRD Positive >0.1% and <1% (n = 3, median 18 months), P = 0.59; or progression-free survival (6.6 months for MRD Positive >1% vs 12 months for MRD Positive >0.1%), P = 0.87.
Progression-free survival after CR/CRi/MLFS (3-12 months and greater than 12 months), on univariate analysis: was determinant on FLT3 WT and NPM1 WT status (P = 0.09), aggressive salvage (P = 0.01), and SCT (P = 0.01). Intermediate-risk cytogenetics, de novo AML, MRD positivity, and receipt of targeted therapy were insignificant. Multivariate analysis identified SCT as the most robust predictor of an enriched and enhanced response (P = 0.02).
Transplanted patients in groups A and B manifested similar age distribution (≥60 years, n = 3 in group A vs 3 in group B, P = 0.21) while de novo AML was overrepresented in group A (75% vs 0% in group B, P = 0.005). Nine patients had cytogenetics evaluated prior to AHSCT, and 8/9 (89%) were classified as being intermediate-risk. The relapse rate after attaining any response (CR, CRi, or MLFS) was 31% (4/13), n = 3 had MRD status checked and were MRD positive. MRD status was assessed in 3/8 (37.5%) patients who had achieved a response prior to AHSCT, and all had CR. Interestingly, survival after AHSCT was adversely affected by female gender (n = 3/12, P = 0.17) which approached significance on univariate and multivariate; response status, time to transplant, BM blasts prior to salvage, greater than one salvage regimen, intermediate cytogenetic risk, de novo AML context, and age ≥60 years, additionally were insignificant factors.
Four patients were bridged to AHSCT without any response achievement. 2/4 (50%) were transplanted directly (n = 1 had an aplastic marrow and died 15 days post-transplant from cytopenias and multiorgan failure; n = 1 had low burden 5% BM blasts and a persistent clonal abnormality of trisomy 8. This latter patient relapsed 79 months after AHSCT in the form of isolated central nervous system myeloid sarcoma and remains alive). The remaining 2/4 (50%) were transplanted with a BM blast burden greater than 30% after being randomized to the investigational Iomab-B (anti-CD45) trial. [n = 1 had a high BM blast burden of 80%, and after Iomab-B had persistent active disease with greater than 40% blasts (approximating 50% cytoreduction) prior to AHSCT. This patient relapsed 1 month after AHSCT; n = 1 had failed two aggressive chemotherapy salvage attempts and had a BM blast burden 40% before Iomab-B, reduced to 30% BM blasts pre-AHSCT. AML relapse was noted 3 months post-transplant].

DISCUSSION
The current study addresses a frequent and practical question in the care of patients with prAML; is salvage chemotherapy and overall response worth pursuing, and what are the conditions for success or failure in that regard? Our observations suggest that salvage chemotherapy in prAML is clearly justified in the setting of consolidation with AHSCT. Although the number of informative cases was too small to allow definitive conclusions, we were encouraged that AHSCT retained its value in the absence of CR/ CRi/MLFS, at the time of transplant. The latter point is noteworthy considering the limited merit of administering multiple salvage regimens for achieving CR/CRi/MLFS before transplant. For prAML patients who did not undergo AHSCT, the benefit from salvage chemotherapy was limited to prolongation of survival by a few months, and only in patients with intermediate-risk karyotype.
Consistent with existing literature on prAML, we confirm its dire prognosis with cited median survival ranging from 3 to 3.8 months in patients treated with supportive care alone [9] and 3-19 months for those receiving salvage therapy [9][10][11][12][13][14]. The ELN denotes regimens containing higher doses of cytarabine (≥1000 mg/m 2 ) are generally considered the best option for those without response to a first course of 7 + 3. Achieving CR after a second course of a higher dose cytarabine regimen, may be lower in comparison to a second 7 + 3 induction after the failure of the first [1]. In this study, neither response (P = 0.88) nor survival benefit (P = 0.69) was seen with a higher cytarabine-dose induction compared to a lower intensive cytarabine induction. Of the 38/60 patients receiving higher-dose cytarabine induction either during their first or second course, 3/38 (8%) had been induced with 2 consecutive courses of only higher-dose cytarabine regimens. These 3 patients manifested a median survival of 2.6 months and failed to achieve a response, underlining the futility of higher dose cytarabine at the expense of efficacy and morbidity [1].
We also observed similar response rates and outcomes between intensive vs alternative modes of salvage chemotherapy. This is parallel to a retrospective study from MDACC where salvage regimens (intense vs HMA), not inclusive of transplant strategies, did not affect survival in 197 patients refractory to 1 course of high dose cytarabine induction [8].
AHSCT has been proposed as the only line of salvage that secures long-term survival in prAML with estimated 5-year survival rates of 20-30% [15,16]. In a 2016 retrospective study of 8907 patients with newly diagnosed AML [17], 473 (5%) were identified as being refractory to 2 courses of intensive induction therapy, and their 5-year survival was reported at 8% vs 9% in patients refractory to 1 course of intensive induction vs 40% in those who achieved CR after 1 course of intensive induction; 86 (18%) of the 473 prAML patients proceeded to AHSCT and survival was superior in those who achieved CR after salvage chemotherapy (n = 49), compared to those transplanted in active disease (n = 37): 38% vs 17%, respectively. Though a small population 12/60 (20%), this is in contrast to our findings where CR/CRi/MLFS bridged to AHSCT (n = 8), appeared to manifest a longer median OS (20 months) vs (13.4 months) for those with no response consolidated with AHSCT (n = 4), the survival difference being insignificant, P = 0.47.
A 2020 study by Short et al. [18] at the MD Anderson Cancer Center focused on those undergoing AHSCT after response (CR/CRi/ MLFS) with MRD negative or positive status. The cohort involved 141 patients with relapsed or refractory AML (88/141, 62% were refractory to induction or had remission lasting <1 year) in whom CR/CRi/MLFS was achieved after 1 cycle of salvage therapy and MRD status evaluated. Refractoriness was defined as refractory to 1 cycle of intensive chemotherapy or 2 cycles of less intense chemotherapy. 90 patients (64%) received intensive cytotoxic chemotherapy, and 51 patients (36%) had lower-intensity therapy (mainly HMA). Thirteen patients (9%) had undergone AHSCT. CR-MRD negativity was noted in 43% and manifested lower rates of early relapse, however, the 13 patients undergoing AHSCT (all in second remission) had the best outcomes despite MRD status or type of response. MRD status evaluation was limited in our cohort, with assessment in 7/13 (54%), and only n = 1 manifesting CR-MRD negativity, with a median survival not yet reached. The CR-MRD negative patient was bridged to SCT, and MRD positivity, >1% (n = 2) pre-AHSCT did not impact relapse risk (P = 0.31) or OS (P = 0.48) post-AHSCT. Though a limited sample, our findings corroborate with Short et al, in MRD status not affecting outcome post-SCT.
Limited clinical data exists on the significance of ELN-defined MLFS. Particularly, in our prAML cohort, outcomes were similar in MLFS vs. CR/CRi with or without transplant. This is consistent with a 2015 retrospective study on 270 AML patients, without primary induction failure, undergoing AHSCT [19], where 206 patients were in CR, 45 in CRi, and 19 in MLFS at time of transplant, with respective 3-year survival rates, post-transplant, of 49%, 46%, and 47% (P = 0.88). Survival after AHSCT, in prAML, requires further validation that should consider confounding effects from age, number of chemotherapy cycles before transplant, BM and peripheral blood blast %, and karyotype [20,21].
The impact of intensive salvage chemotherapy for attaining CR/ CRi prior to AHSCT has been studied in 845 AML cases failing one course of intensive induction chemotherapy [10], in which multivariate analysis revealed CR/CRi prior to transplant, predicted the best survival while older age and unfavorable karyotype adversely affected CR/CRi. Other reports have similarly addressed the importance of attaining CR before transplant [11]. On the other hand, AHSCT as first-line salvage, irrespective of achieving remission, has also been investigated in prAML [21][22][23]. In one retrospective cohort from MDACC involving patients refractory to 1 course of intensive induction, median survival was 16 months (3-year survival 39%) in those bridged directly to AHSCT (n = 28) vs 3 months (3year survival 2%) with salvage alone (n = 149) (P < 0.001) [12].
Venetoclax (an oral highly selective Bcl-2 inhibitor) in combination with HMA, as opposed to venetoclax single-agent therapy, has shown promise in refractory/relapsed AML (defined as refractory or relapsed after one induction regimen of either intense or less intense (inclusive of HMA alone), or relapse post-AHSCT) [24][25][26]. Most patients received at least ≥ 2 salvage regimens prior to initiation. Toxicities (particularly grade 3 and 4 cytopenias: neutropenia and thrombocytopenia) complicate its use in refractory or relapsed AML. Nonetheless, activity has been shown in FLT3, TP53 mutations, and unfavorable risk karyotypes. After a median of 1 to 2 cycles of HMA + Venetoclax, overall response rates (CR/CRi) have ranged from 46% (41/90), median OS (7.8 months for all patients and 16.6 months for CR/CRi) [24]; to 33% (14/42), median OS 3 months vs 15 months for CR/CRi [25]; to 42% (23/55), 62% (34/55), median OS 7.8 months with response, and duration of response 16.8 months [26]. In this study, 2 patients received HMA + Venetoclax as a first or second line of salvage. One patient with NRAS, MPL positivity, FLT3/NPM1 WT had received 2 courses of high-dose cytarabine inductions with residual BM blast burden 80% and dural myeloid sarcoma. 1 cycle of decitabine + venetoclax was started, though discontinued due to severe pancytopenias (OS 2.6 months). The second patient was FLT3-TKD positive, NPM1 WT, biallelic WT1, who was started  on gilteritinib as first-line salvage, cycle 2 halted due to elevated liver enzymes. BM blast burden had reduced >50% (from 40% to 11%). 2 months after, with BM blasts rising to 60%, FISH positive for RUNX1 in 53% of nuclei and a cryptic 3:1 translocation, cycle 1 of decitabine + venetoclax was started. CR was achieved after 1 cycle of decitabine + venetoclax with the disappearance of FLT3-TKD positivity and FISH abnormalities. MRD positivity on flow cytometry was observed to be 0.13%. The patient has continued with at least 3 cycles of decitabine + venetoclax, cycle 3 delayed due to a neutropenic PICC line infection.
In summary, our findings highlight the importance of AHSCT for securing long-term survival in a homogeneously exclusive cohort of ELN-defined prAML (primary induction failure), regardless of the degree of hematologic recovery. Currently, available salvage chemotherapy regimens (either intensive or less intensive (inclusive of targeted therapy)) for this prAML population were similar in regard to response or survival benefits. The promising clinical data acquired for HMA + venetoclax in combined refractory and relapsed AML requires further investigation in patients with primary induction failure and molecular mutations.