Acute myeloid leukemia (AML) is a phenotypically and prognostically heterogeneous hematopoietic stem cell disease that may be cured in eligible patients with intensive chemotherapy and/or allogeneic stem cell transplantation (allo-SCT). Tremendous advances in sequencing technologies have revealed a large amount of molecular information which has markedly improved our understanding of the underlying pathophysiology and enables a better classification and risk estimation. Furthermore, with the approval of the FMS-like tyrosine kinase 3 (FLT3) inhibitor Midostaurin a first targeted therapy has been introduced into the first-line therapy of younger patients with FLT3-mutated AML and several other small molecules targeting molecular alterations such as isocitrate dehydrogenase (IDH) mutations or the anti-apoptotic b-cell lymphoma 2 (BCL-2) protein are currently under investigation. Despite these advances, many patients will have to undergo allo-SCT during the course of disease and depending on disease and risk status up to half of them will finally relapse after transplant. Here we review the current knowledge about the molecular landscape of AML and how this can be employed to prevent, detect and treat relapse of AML after allo-SCT.

Despite the curative potential of allogeneic stem cell transplantation (allo-SCT) in patients with acute myeloid leukemia (AML) and myelodysplastic syndromes (MDS), many patients will relapse. Until recently therapeutic options mainly consisted of palliative care, chemotherapy, donor lymphocyte infusions and second transplantation in selected cases. Still many patients either do not tolerate intensive therapies or do not achieve durable remissions and will finally succumb. Given this unmet medical need the hypomethylating agents (HMA), Azacitidine (Aza) and Decitabine (DAC) have been tested as salvage therapy in patients with myeloid malignancies relapsing after allo-SCT. Furthermore, they have also been incorporated into prophylactic and pre-emptive approaches to avoid haematological relapse. In this review, we summarize the evidence from retrospective studies but also from a few prospective trials regarding the use of HMA after transplant. To aid clinicians in their daily clinical practice, we also comment on some practical aspects such as dosing and schedule, the choice of HMA and the use of complementary cellular therapies. Finally, this review also gives an overview on potential mechanisms mediating the efficacy of HMA after transplant as well as ongoing preclinical research and clinical activities aiming to further improve this treatment approach.

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is the only curative treatment for patients with advanced myelodysplastic syndrome (MDS) and secondary acute myelogenous leukemia (sAML), but in the absence of prospective trials the impact of pretransplant cytoreduction is controversially discussed. We retrospectively analyzed the outcome of 165 patients with MDS and excess blasts (n = 126, 76%) and sAML (n = 39, 24%) according to a pretransplant strategy. Sixty-seven patients (41%) were directly transplanted (upfront group), whereas 98 patients (59%) had received pretransplant cytoreductive treatment (induction chemotherapy [CTX], n = 64; hypomethylating agents [HMAs], n = 34) resulting in a significantly higher complete remission rate in the CTX group (59% versus HMA 18%, P < .0001). Estimated rates of 5-year overall survival (OS) and relapse-free survival (RFS) for the entire group were 54% and 39%, respectively. The 5-year OS rates of the upfront, CTX, and HMA groups were 61%, 50%, and 45%, respectively (P = .116), whereas RFS rates were 38%, 41%, and 38% (P = .926). Cumulative incidence of relapse (CIR) and nonrelapse mortality (NRM) did not differ between treatment groups. In the upfront group no difference regarding OS and RFS was seen with respect to pretransplant blast count (>10% versus <10%). In multivariate analyses type of pretransplant strategy did not have an effect on OS, RFS, CIR, and NRM, whereas cytogenetics (OS, RFS, CIR), reduced-intensity conditioning (OS, RFS, CIR), and an unrelated donor (RFS, CIR) were identified as negative predictors. When compared with the upfront group, 5-year OS was significantly lower in patients with CTX-refractory disease (34% versus 64%, P = .0346) and by clear trend in HMA nonresponders (42% versus 61%, P = .073), whereas RFS did not differ significantly. In further support of the concept, that pretransplant therapy may favor the selection of resistant clones, patients in the upfront group had a higher likelihood to respond to HMAs as salvage therapy for relapse in comparison with pretreated patients (complete remission, 58% versus 10%; P = .0005) and a higher 2-year OS rate after relapse (59% versus 19%, P = .0001). These data suggest that an upfront transplant strategy is at least not inferior to pretransplant cytoreduction and may be augmented by HMAs + donor lymphocyte infusion salvage therapy in case of relapse after allo-HSCT.

In contrast to the evidence regarding azacitidine (Aza), there is limited knowledge about the combination of decitabine (DAC) and donor lymphocyte infusions as salvage therapy for relapse after allogeneic stem cell transplantation (allo-SCT) so far. We retrospectively analyzed data of 36 patients with hematological (n = 35) or molecular relapse (n = 1) of acute myeloid leukemia (AML, n = 29) or myelodysplastic syndrome (MDS, n = 7) collected from 6 German transplant centers. Patients were treated with a median of 2 cycles DAC (range, 1 to 11). DAC was the first salvage therapy in 16 patients (44%), whereas 20 patients (56%) had previously received 1 to 5 lines of salvage therapy including 16 of them had been treated with Aza. In 22 patients (61%), a median of 2 DLI per patient (range, 1 to 5) was administered in addition to DAC. As a result, overall response rate was 25% including 6 complete remissions (CR, 17%) and 3 partial remissions (PR, 8%). Three patients within the first-line group achieved CR, while also 3 patients receiving DAC as second-line treatment reached CR including 2 patients with previous Aza failure. Median duration of CR was 10 months (range, 2 to 33) and no patient relapsed so far. The 2-year OS rate was 11% (± 6%) without any difference between first-line and pretreated patients. Incidence of acute and chronic graft-versus-host disease was 19 and 5%. Taken together, DAC exerts clinical efficacy in patients with AML or MDS relapsing after allo-SCT and is able to induce durable remissions in individual patients suggesting that DAC may be an alternative to Aza or even a second choice after Aza failure.

Taken together, the combination of sorafenib, Aza, and DLI shows promising efficacy and deserves further evaluation in larger patient groups.

Overexpression of the Wilms' tumor 1 (WT1) gene is informative in many patients with acute myelogenous leukemia (AML) and myelodysplastic syndromes (MDS) and is measurable in peripheral blood (PB). Despite these advantages, WT1 has not broadly been established as a marker for minimal residual disease (MRD) monitoring after allogeneic hematopoietic stem cell transplantation (allo-HSCT) due to limited patient numbers, differing sample sources, and nonstandardized in-house methods. To estimate the value of WT1 as an MRD marker, we serially quantified PB WT1 expression using a standardized European LeukemiaNet-certified assay in 59 patients with AML and MDS after allo-HSCT. We compared its performance with routine methods such as chimerism, XY-fluorescence in situ hybridization (FISH), disease-specific cytogenetic, and molecular analyses, which were accessible in 100%, 34%, 68%, and 37%, respectively. Twenty-four patients (41%) relapsed within a median of 126 days after allo-HSCT, and 20 of them showed at least 1 elevated WT1 value above the validated cutoff. The other 35 patients (59%) remained in complete remission, and only 1 patient had a transient increase in WT1 expression. This reflects a sensitivity of 83% and a specificity of 97% for WT1 and appears to be favorable compared with the sensitivities and specificities observed for chimerism (33% and 91%), XY-FISH (67% and 73%), cytogenetic (33% and 77%), and molecular (78% and 85%) analyses. Further supporting its predictive impact, elevated WT1 expression prompted an earlier BM biopsy and consecutively the diagnosis of relapse in 62% of patients. The results of this real-life experience imply that PB WT1 expression is measurable by a standardized assay and predicts imminent relapse after allo-HSCT with high sensitivity and specificity in most patients with AML and MDS.

To investigate the efficacy and toxicities of CPX-351 outside a clinical trial, we analyzed 188 patients (median age 65 years, range 26–80) treated for therapy-related acute myeloid leukemia (t-AML, 29%) or AML with myelodysplasia-related changes (AML-MRC, 70%). Eighty-six percent received one, 14% two induction cycles, and 10% received consolidation (representing 22% of patients with CR/CRi) with CPX-351. Following induction, CR/CRi rate was 47% including 64% of patients with available information achieving measurable residual disease (MRD) negativity (<10−3) as measured by flow cytometry. After a median follow-up of 9.3 months, median overall survival (OS) was 21 months and 1-year OS rate 64%. In multivariate analysis, complex karyotype predicted lower response (p = 0.0001), while pretreatment with hypomethylating agents (p = 0.02) and adverse European LeukemiaNet 2017 genetic risk (p < 0.0001) were associated with lower OS. Allogeneic hematopoietic cell transplantation (allo-HCT) was performed in 116 patients (62%) resulting in promising outcome (median survival not reached, 1-year OS 73%), especially in MRD-negative patients (p = 0.048). With 69% of patients developing grade III/IV non-hematologic toxicity following induction and a day 30-mortality of 8% the safety profile was consistent with previous findings. These real-world data confirm CPX-351 as efficient treatment for these high-risk AML patients facilitating allo-HCT in many patients with promising outcome after transplantation.

The inclusion of mutation status improved risk stratification for newly diagnosed patients with chronic myelomonocytic leukemia (CMML). Stem cell transplantation is a potentially curative treatment option, and patient selection is critical because of relevant transplant-related morbidity and mortality. We aimed to evaluate the impact of mutation status together with clinical presentations on posttransplant outcome. Our study included 240 patients with a median follow-up of 5.5 years. A significant association with worse survival was identified for the presence of mutations in ASXL1 and/or NRAS. In multivariable analysis, ASXL1- and/or NRAS-mutated genotype (hazard ratio [HR], 1.63), marrow blasts &gt;2% (HR, 1.70), and increasing comorbidity index (continuous HR, 1.16) were independently associated with worse survival. A prognostic score (CMML transplant score) was developed, and the following points were assigned: 4 points for an ASXL1- and/or NRAS-mutated genotype or blasts &gt;2% and 1 point each for an increase of 1 in the comorbidity index. The CMML transplant score (range, 0-20) was predictive of survival and nonrelapse mortality (P &lt; .001 for both). Up to 5 risk groups were identified, showing 5-year survival of 81% for a score of 0 to 1, 49% for a score of 2 to 4, 43% for a score of 5 to 7, 31% for a score of 8 to 10, and 19% for a score &gt;10. The score retained performance after validation (concordance index, 0.68) and good accuracy after calibration. Predictions were superior compared with existing scores designed for the nontransplant setting, which resulted in significant risk reclassification. This CMML transplant score, which incorporated mutation and clinical information, was prognostic in patients specifically undergoing transplantation and may facilitate personalized counseling.