Impact of disease burden on clinical outcomes of AML patients receiving allogeneic hematopoietic cell transplantation: a study from the Acute Leukemia Working Party of the European Society for Blood and Marrow Transplantation

Pre-transplant detectable measurable residual disease (MRD) is still associated with high risk of relapse and poor outcomes in acute myeloid leukemia (AML). We aimed at evaluating the impact of disease burden on prediction of relapse and survival in patients receiving allogeneic hematopoietic cell transplantation (allo-HCT) in first remission (CR1). We identified a total of 3202 adult AML patients, of these 1776 patients were in CR1 and MRD positive and 1426 patients were primary refractory at time of transplant. After a median follow-up of 24.4 months, non-relapse mortality and relapse rate were significantly higher in the primary refractory group compared to the CR1 MRD positive group (Hazards Ratio (HR) = 1.82 (95% CI: 1.47–2.24) p < 0.001 and HR = 1.54 (95% CI: 1.34–1.77), p < 0.001), respectively. Leukemia-free survival (LFS) and overall survival (OS) were significantly worse in the primary refractory group (HR = 1.61 (95% CI: 1.44-1.81), p < 0.001 and HR = 1.71 (95% CI: 1.51–1.94), p < 0.001, respectively). Our real-life data suggest that patients in CR1 and MRD positive at time of transplant could still be salvaged by allo-HCT with a 2-year OS of 63%, if negative MRD cannot be obtained and their outcomes are significantly better than patients transplanted with active disease.


INTRODUCTION
Acute myeloid leukemia (AML) is a very heterogeneous hematological malignancy, including numerous genetically defined subsets with different prognostic implications. Multiple baseline factors such as age, cytogenetics, and genetic mutations are considered independent risk factors for survival and are incorporated into the 2017 European Leukemia Net (ELN) risk stratification [1]. One emerging robust prognostic variable in AML is the depth of response represented by the assessment of measurable residual disease (MRD) post-treatment [2]. It is currently used to refine the traditional complete remission (CR) status assessed by microscopy to a new well-defined response criterion by ELN consensus "CR without detectable MRD" [1]. There are various ways to detect MRD in AML including multiparametric flow cytometry (MFC), polymerase chain reaction (PCR) and next generation sequencing (NGS) [3]. Among the European Society of Blood and Marrow Transplantation (EBMT) reporting centers, MRD assessment was routinely done in the peri-transplant setting in 92 centers using only PCR in 23 centers, MFC in 13 centers, both PCR and MFC in 39 centers, and all three techniques in 14 centers [4].
Patients with persistent detectable MRD after chemotherapy are offered allogeneic hematopoietic cell transplantation (allo-HCT) as a post-remission strategy. However, pretransplant MRD positive disease is still associated with high risk of relapse and poor outcomes [5,6]. Whereas many studies evaluated the impact of pre-transplant MRD status in patients who were in morphological remission, little is known about MRD positivity outcomes in comparison with active disease at transplant. In one study, patients who were transplanted in CR and were MRD positive had the same post-allo-HCT relapse rate and survival as patients who were transplanted in active disease, with a 3-year overall survival (OS) of 26% and 23%, respectively, compared to 73% in patients who were in CR and were MRD negative [7]. This risk remains high regardless of MRD conversion with conditioning posttransplant [8].
Additionally, the preferred conditioning and donor types are still an ongoing debate in the setting of allo-HCT with pretransplant MRD positive status. Few studies have demonstrated that myeloablative conditioning, not reduced-intensity conditioning, is able to overcome the negative predictive value of pretransplant MRD positivity, whereas other studies did not show any difference based on the intensity of the conditioning regimens [9][10][11][12].
Through the EBMT database, we aim at evaluating the impact of disease burden (CR MRD positive versus active disease) on prediction of relapse and survival in patients receiving allo-HCT in first CR.

MATERIALS AND METHODS Study design and data collection
This is a retrospective, registry-based multicenter analysis. Data were provided and approved by the Acute Leukemia Working Party of the EBMT. The EBMT is a voluntary working group of more than 600 transplant centers which are required to report all consecutive HCTs and follow-ups once a year. Audits are routinely performed to determine the accuracy of the data. Since January 2003, all transplant centers have been required to obtain written informed consent prior to data registration with the EBMT, following the guidelines of the Declaration of Helsinki, 1975.
Eligibility criteria for this analysis included age >18 years, with de novo AML who received their first allo-HCT between 2010 and 2019, with available MRD status either by MFC or PCR. The collected variables included recipient age at transplant, recipient and donor gender, date of diagnosis, karyotype and molecular profile at diagnosis, MRD status at transplant, Karnofsky performance status score at time of transplant, transplant-related factors including conditioning regimen, graft-versushost disease (GvHD) prophylaxis, donor type, stem cell source (bone marrow [BM] or peripheral blood [PB]), patient and donor cytomegalovirus status, development of acute and chronic GvHD, and date of last follow-up or death and cause of death.

Definitions
Primary refractory or active disease at time of allo-HCT is defined by a persisting blast count >5% in BM, persisting blasts in PB or the detection of extramedullary disease, whereas CR1 MRD positive status refers to a morphologic remission with <5% blasts in the BM with detectable MRD regardless of the type of marker or method used.

Clinical endpoints
Clinical endpoints include OS defined as time to death from any cause, leukemia-free survival (LFS), defined as time being alive without evidence of relapse, relapse incidence (RI), non-relapse mortality (NRM) defined as death without evidence of relapse, acute GvHD (aGvHD), chronic GvHD (cGvHD) and refined GvHD-free relapse-free survival (GRFS) defined as time being alive with neither grade III-IV aGvHD nor severe cGvHD nor relapse at any time point after transplant.

Statistical analysis
The probabilities of OS, LFS, and GRFS, were calculated from time of transplant using the Kaplan-Meier estimate. The follow-up time was calculated using the reverse Kaplan-Meier method. GvHD, RI, and NRM were calculated using cumulative incidence curves in a competing risk setting. Univariate comparisons between groups were performed using the Chi-square and Fischer's exact test for categorical variables and the Mann-Whitney test for continuous variables, the Gray's statistic for cumulative incidence functions (GvHD, NRM, RI) and the log-rank test for survival outcomes (OS, LFS, and GRFS).
Multivariate analysis was performed using a Cox proportional-hazards model which included variables differing significantly (p < 0.05) between the groups, factors known to be associated with outcomes, plus a center frailty effect to take into account the heterogeneity across centers. Results were expressed as a hazard ratio (HR) with a 95% confidence interval (CI).
Sensitivity analysis included subgroup analyses based on genetic subtype, and conditioning intensity.
All tests were two-sided. The type-1 error rate was fixed at 0.05 for determination of factors associated with time-to-event outcomes. Statistical analyses were performed with SPSS 27.0 (SPSS Inc., Chicago, IL, USA) and R 4.1.1 (R Development Core Team, Vienna, Austria, URL:https:// www.R-project.org/).

RESULTS
We identified a total of 3202 adult patients diagnosed with AML, of these, 1776 patients were in CR1 with detectable pre-transplant MRD ( . Patients with primary refractory disease were more likely to have poor-risk cytogenetics (43.1% vs 18.1%, p < 0.0001), FLT3 wild-type disease (75% vs 49%, p < 0.0001) and NPM1 wildtype disease (85.6% vs 39.4%, p < 0.0001), and more likely to receive reduced-intensity conditioning (56.3% vs 42.3%, p < 0.0001). Baseline characteristics of all patients and each group are summarized in Table 1. Transplant-related baseline characteristics are summarized in Table 2.

Subgroup analyses
Patients with primary refractory disease at time of transplant had worse outcomes compared to those transplanted in CR1 MRD positive regardless of the type of conditioning regimen used, or underlying biology of the disease. When selecting patients who received myeloablative conditioning, the 2-year LFS and OS were significantly worse in the primary refractory group (30% vs 58%, p = 0.001 and 38.6% vs 68.4%, p = 0.001, respectively). Similarly, for patients who received reduced-intensity conditioning, the 2-year LFS was 32.4% vs 50%, p = 0.001 and the 2-year OS was 37.4% vs 56.8%, p = 0.001, respectively, results not shown.
Based on the cytogenetic risk, patients with primary refractory disease performed worse in all three subgroups (good, intermediate, and poor). The 2-year LFS and OS for patients with primary refractory disease were 36.5% and 54.9% respectively, in the good-risk cytogenetic subgroup, 37.9% and 44.7%, respectively, in the intermediate-risk cytogenetic group, and 22.7% and 28.3% respectively, in the poor-risk cytogenetic subgroup, compared to 70.9% and 79.7%, 56% and 64.7%, 35.8%, and 45.5% in the three respective cytogenetic subgroups for patients transplanted in CR1 MRD positive (p = 0.001 for all comparisons), results not shown.

DISCUSSION
Allo-HCT in high-risk AML is still considered a therapeutic approach of curative intent, even in a subset of patients with active disease [13,14]. The use of myeloablative conditioning, sequential chemotherapy followed by reduced-intensity conditioning, and prophylactic donor lymphocyte infusions (DLI) post  allo-HCT have shown encouraging results in patients with active AML pre-transplant [15][16][17][18]. Within this subgroup of patients transplanted while not in remission, the burden of disease remains one of the main prognostic factors [18]. Better long-term OS has been reported in patients with a blast count <20% (3-year OS, 32% versus 5%) [18]. Available evidence indicates that the burden of disease can be observed as a continuous variable, where any MRD (regardless of methodology) is related to worse outcomes. Patients transplanted in remission but with residual cytogenetic abnormalities had worse survival compared to those transplanted with cytogenetic remission [19]. With the advancement of technology, deeper detection of residual leukemia cells was possible up to 10 −6 in some cases using NGS-based techniques [20]. Nowadays, MRD assessment in AML provides an additional prognostic tool to determine the relapse risk of an individual patient. Patients with detectable MRD pre-transplant regardless of the MRD marker and method are at higher risk of relapse and mortality compared to those transplanted in remission without detectable MRD. It is intuitive to conclude that detectable residual leukemia is a continuous prognostic variable independent from other factors including AML subtype, molecular biology, conditioning regimen and donor type.
However, some data suggest that the outcomes of patients transplanted with detectable MRD are similarly as poor as patients transplanted in active disease [7]. Different groups have reported that CR MRD positive and active disease have the same cumulative incidence of relapse (50-60% relapse rate) regardless of the type of conditioning used, advocating for attainment of MRD negativity as an important objective pre-transplant [21]. With the emergence of effective novel drugs in AML such as FLT3 inhibitors and venetoclax, MRD-directed treatment prior to transplant might have a role in improving allo-HCT outcomes. This strategy is under investigation and is still unsolved [22]. A recent randomized trial investigated the role of sequential conditioning followed by allo-HCT compared to intensive remission induction chemotherapy followed by allo-HCT in relapsed and/or refractory (R/R) AML. It resulted in similar 1-year LFS (71.5% versus 69.9%, respectively p = 0.8) and 1-year OS (69.1% versus 71.9%, respectively p = 0.47) supporting the use of allo-HCT as an effective anti-leukemia treatment without the need of inducing remission prior to transplant [23].
Despite the association of detectable pre-transplant MRD with less favorable outcomes, our real-life data suggest that patients in CR1 with MRD positivity at time of transplant could still be salvaged by allo-HCT with a 2-year OS of 63%, especially if negative MRD cannot be obtained. Their outcomes are significantly better than patients transplanted with active disease. Although the baseline characteristics between both groups (CR1 MRD positive and active disease) were different, as more patients in CR1 MRD positive were younger, had better cytogenetic risks and received myeloablative regimens, the difference in long-term survival outcomes was seen across subgroups regardless of the conditioning regimen used, or cytogenetic risk category. Those patients could probably benefit from post-transplant interventions such as rapid tapering of immunosuppression, prophylactic DLI, pharmacologic maintenance therapies like sorafenib or hypomethylating agents [24][25][26].
The role of post-transplant interventions cannot be confirmed using this database, as variables pertinent to post-transplant   strategies could not be collected and verified. Other limitations of our study include the inability to quantify the burden of disease in both groups (CR MRD positive and active disease). This information could be of interest to confirm the continuous prognostic value of disease burden in AML. In conclusion, we were able to demonstrate, based on realworld evidence that allo-HCT can still be a curative option for a subset of patients with persistent detectable disease pretransplant. Whether MRD converting strategies are to be implemented in the pre-or post-transplant setting, is still a matter of controversy, especially with the increased use of novel effective agents.

DATA AVAILABILITY
Upon request from the corresponding author.