Yeast Mitochondrial Escape 1-Like 1 (YME1L) is an inner mitochondrial membrane–anchored ATP-dependent metalloprotease. Here, we investigated its expression and functional role in acute myeloid leukemia (AML). YME1L protein was upregulated in 6 out of 7 AML cell lines and 17 of 23 primary AML patient samples compared to normal hematopoietic stem cells. Analysis of the TCGA dataset revealed that high YME1L mRNA expression correlated with inferior 5-year overall survival in AML patients. To evaluate the essentiality of YME1L in AML, we knocked down the protease in AML cell lines (OCI-AML2, NB4, TEX, and THP-1) with shRNA. YME1L knockdown decreased cell proliferation and clonogenic growth in all tested cell lines. Moreover, YME1L depletion decreased the engraftment of TEX cells into the marrow of immune deficient mice. Interestingly, YME1L knockdown did not induce cell death as measured by Annexin V/PI staining. Rather, YME1L knockdown induced AML differentiation as evidenced by upregulation of CD14, a marker of monocytic differentiation. To understand how YME1L promoted differentiation and decreased proliferation, we performed RNA sequencing of AML cells after YME1L depletion. Knockdown of YME1L upregulated genes associated with type I interferon (IFN) signaling and viral mimicry pathways. By qPCR, we confirmed upregulation of IFN-β, IFN-γ, and interferon-stimulated genes, ISG15, IFI44, IFIT2 after YME1L knockdown. We demonstrated that activation of the IFN signaling was secondary to activation of the cGAS–STING pathway as pharmacologically inhibiting STING abolished the upregulation of IFN signaling after YME1L depletion. We also showed that YME1L knockdown promoted the release of mitochondrial double-stranded DNA into the cytoplasm, a known trigger of IFN signaling. Finally, we demonstrated that inhibiting the VDAC channel with VBIT-4 blocked the upregulation of IFN signaling after YME1L knockdown, thus supporting the functional importance of mtDNA efflux into the cytoplasm as the mechanism explaining increased IFN signaling after YME1L depletion. In summary, we demonstrated that YME1L is overexpressed in a subset of AML samples and is required for AML proliferation and clonogenic growth. YME1L regulates the leakage of mtDNA into the cytoplasm, activation of the cGAS–STING–IFN axis, and AML differentiation. Thus, we have uncovered novel functions for the mitochondrial protease YME1L in AML and new mechanisms for how a mitochondrial protease regulates nuclear gene expression and progenitor function.

Background: Acute myeloid leukemia with extramedullary disease (EMD-AML) represents a distinct clinical entity associated with diagnostic and therapeutic challenges, and its prognostic significance remains uncertain. Methods: A retrospective study of 617 adults with newly diagnosed AML (2005-2018) was conducted, analyzing 246 patients with EMD-AML and 371 without EMD involvement. The clinical characteristics and treatment outcomes were analyzed. Propensity score matching (PSM) was applied to adjust for baseline confounders. Results: Patients with isolated EMD-AML and those with concurrent bone marrow involvement had comparable clinical outcomes. NPM1 mutations (48% vs. 25%, p = 0.0002) and t(8;21) translocation (23.2% vs. 3.7%, p < 0.001) were enriched in the EMD-AML cohort. After PSM, EMD-AML patients achieved a higher overall response rate compared with non-EMD-AML (88.1% vs. 72.0%, p = 0.0002) but experienced significantly higher relapse rates (35.7% vs. 15.5%, p < 0.0001). Despite the achievement of a higher response rate, EMD-AML was associated with shorter median overall survival (OS) (14.2 vs. 64.1 months, p < 0.0001) and event-free survival (EFS) (9.5 vs. 55.9 months, p < 0.0001). In a multivariable analysis, EMD-AML remained independently associated with worse OS and EFS (OS HR 1.79, p = 0.01; EFS HR 1.95, p = 0.001). Allogeneic hematopoietic stem cell transplantation did not confer a survival advantage in EMD-AML patients. Conclusions: EMD-AML, whether isolated or concurrent with bone marrow disease, represents a high-risk entity characterized by poor long-term outcomes despite strong initial response rates. Obtaining tissue biopsies for molecular profiling may help improve risk stratification, identify targetable mutations and guide individualized treatment.

Tyrosine kinase inhibitors (TKIs) added to chemotherapy have improved outcomes of adult patients with Philadelphia-positive B-cell acute lymphoblastic leukemia (Ph+ B-ALL). These improvements initially led to a larger proportion of patients realizing allogeneic stem cell transplantation (alloSCT), long considered essential for cure, but there has been a re-evaluation of alloSCT. At Princess Margaret Hospital (PM), adult patients with Ph+ B-ALL have been treated with a pediatric-inspired chemotherapy protocol with mostly imatinib. In the last two decades, we have witnessed many iterative changes in our approach. Here, we examine the outcomes of all Ph+ B-ALL patients treated at our institution from 2001 to 2019. During this time, there were two major protocol changes-omission of asparaginase in 2009 and discontinuation of routine referral for first complete remission (CR1) alloSCT from the early 2010s. Median follow-up was 41.13 months (range, 0.46-228.79). In total, 141 patients (91.56%) achieved CR1. Patient outcomes improved iteratively, with the best results seen in the final (2016-2019) cohort: no asparaginase, no routine alloSCT referral in CR1; 4-year overall survival (OS) and relapse-free survival (RFS) were 87.0% and 69.3%, respectively. The long-term OS in this patient group retained statistical significance in the multivariable analysis (p = 0.0176) when BCR::ABL1 molecular measurable residual disease (MRD) was considered.

Non-canonical Polycomb repressive complex 1 (ncPRC1) sustains leukemic stem cells and disease propagation in acute myeloid leukemia (AML), yet how its activity is preserved remains poorly understood. UBE2E3 is an E2 ubiquitin-conjugating enzyme and AML dependency of undefined function. Using proximity-dependent biotinylation and mass spectrometry, we identified UBE2E3 as a proximal interactor of multiple ncPRC1 components, including the PRC1.1 proteins BCOR, KDM2B and PCGF1, the PRC1.6 proteins MGA, L3MBTL2 and PCGF6, and the shared core proteins RING1A/B. Transcriptomic profiling after UBE2E3 depletion in AML recapitulated gene signatures of BCOR loss and revealed widespread deregulation of PRC1.1 target genes. Mechanistically, UBE2E3 depletion reduced BCOR and RING1B protein abundance while KDM2B, PCGF1 and PCGF6 were preserved. Re-expression of wild-type, but not catalytically dead, UBE2E3 restored RING1B abundance and normalized representative PRC1 target genes, demonstrating a requirement for E2 activity. Functionally, UBE2E3 depletion impaired AML cell proliferation, reduced clonogenic growth, induced differentiation-associated transcriptional programs and markedly diminished leukemic engraftment in xenograft models. These findings identify UBE2E3 as a previously unrecognized regulator of ncPRC1 function in AML. By maintaining the BCOR-RING1B axis of PRC1.1, UBE2E3 preserves leukemic stemness, clonogenicity and leukemia-initiating activity, linking ubiquitin signaling to Polycomb-dependent transcriptional control.

Extramedullary disease (EMD) in acute myeloid leukemia (AML) involves leukemic infiltration outside the bone marrow and is observed in up to 25% of patients. Despite its frequency, EMD is not currently incorporated into the European LeukemiaNet (ELN) 2022 risk stratification system. Recent studies have demonstrated molecular discordance between leukemic cells in the bone marrow and those found at extramedullary (EM) sites, raising concerns that molecular profiling based solely on bone marrow samples may not adequately guide clinical management of patients with EMD. To characterize the clinical features, treatment outcomes, and prognostic implications of EMD in a large, retrospective cohort of AML patients treated over two decades at Princess Margaret Cancer Centre. We conducted a retrospective analysis of 617 adult AML patients, categorizing them into non-EMD (n=371) and EMD groups (n=246). The EMD group was further subdivided into isolated EM involvement (n=53) and concurrent marrow and EM involvement (n=193). We compared complete remission (CR) rates, relapse rates, rates of allogeneic hematopoietic stem cell transplantation (allo-HSCT), overall survival (OS), and event-free survival (EFS). Propensity score matching (PSM) was used to adjust for differences in age, ELN risk category, induction regimen, and allo-HSCT. Multivariate analyses (MVA) were performed to identify independent prognostic factors. Among the 246 patients with EMD, 53 (21.5%) had isolated EMD, while 193 (78.5%) presented with concurrent bone marrow involvement. CR rates following induction therapy were similarly high in both groups (96% in isolated EMD vs. 89% in combined involvement; p=0.18). Relapse rates (32% vs. 34%; p=0.77) and rates of allogeneic HSCT (40% vs. 32%; p=0.32) were also comparable, indicating similar outcomes between isolated EMD and concurrent marrow involvement with EMD.Prior to PSM, comparing non-EMD from 2015-2018 (n=371) patients to EMD patients (n=246), EMD patients were significantly younger (p=0.04) and had more intermediate-risk disease (60% vs. 19%, p<0.01) and NPM1 mutations (48% vs. 25%, p<0.01). They received intensive induction therapy more frequently (89% vs. 74%, p<0.01) but underwent allo-HSCT less often (34% vs. 43%, p=0.02). OS and EFS were significantly inferior in the EMD group (median OS: 14.2 vs. 53.2 months; median EFS: 10.6 vs. 43.7 months; both p<0.0001). After PSM (168 matched pairs), EM remained associated with inferior OS (median 14.2 vs. 64.1 months; and EFS :9.5 vs. 55.9 months; both p<0.0001).MVA confirmed EMD as an independent adverse prognostic factor for OS (HR 1.87, 95% CI 1.2–2.91, p=0.005) and EFS (HR 1.93, 95% CI 1.36–2.75, p=0.0002). Among EMD patients, allo-HSCT was associated with improved EFS (HR 0.49, 95% CI 0.24–0.99, p=0.048), although OS benefit did not reach significance (HR 0.60, p=0.20).The most common EM involvement sites were the skin (39%) and central nervous system (CNS; 15%). Neither specific EM site involvement, nor mutations frequently identified in EMD patients (NPM1, TET2, ASXL1, DNMT3A) were significantly associated with differences in OS or EFS.Notably, within the ELN 2022 favorable-risk subgroup, EMD presence was associated with higher relapse rates (31% vs. 14%, p=0.059) and trends toward inferior median OS (54.7 months vs. not reached, p=0.08) and EFS (40.6 months vs. not reached, p=0.05), highlighting its prognostic significance even in traditionally lower-risk AML patients. This large, retrospective study emphasizes the negative prognostic impact of EMD in AML, independent of ELN risk stratification, even among patients classified as favorable-risk. Although allo-HSCT may offer some improvement in disease control, it does not fully overcome the poor prognosis linked to EMD. Additionally, these results underscore the importance of site-specific molecular profiling in guiding therapeutic decisions. EMD warrants inclusion in future AML risk stratification models and prospective therapeutic trials.

e18522 Background: The outcome of FLT3 mut R/R AML is dismal. Use of Giltertinib monotherapy results in modest benefits with an improvement in EFS (2.8 months) and OS (9.3 months). Triplet (Azacytidine, Venetoclax and Giltertinib) and doublet combinations (Venetoclax and Gilteritinib) have shown high rates of CR/CRi, mCRc and longer survival. However, on cross-trial comparison, gilteritinib combinations have similar survival to monotherapy (ADMIRAL trial). We presented our experience of FLT3 mut R/R AML, where we showed that gilteritinib combinations resulted in better mCRc and higher transplant feasibility. In this analysis we intend to look at genomic predictors which could predict survival in this population. Methods: We conducted aretrospective, single center study to evaluate the impact of co-mutations on the survival outcomes associated with gilteritinib based therapy for FLT3 mut R/R AML. Results: A total of 68 FLT3 mut R/R AML patients were treated with Giltertinib or its combinations between January 2017 and March 2024. Giltertinib monotherapy was used in 47 while 21 received combination. Combinations included Giltertinib+ Venetoclax (n = 8), Giltertinib+ Azacytidine+ Venetoclax(n = 11) and Gilteritinib plus azacytidine (n = 2). The median OS with monotherapy and combination were 5.8 months and 14.9 months, respectively ( p = 0.0958). The most common co-occurring mutations were DNMT3A (44%), NPM1 (41%), RUNX1 (19%), ASXL1 (15%) and IDH2 (15%). Notably, 21 patients (30.8%) harbored co-mutations in NPM1 and DNMT3A (termed triple-mutated).Triple-mutated patients showed significantly higher rates of CR/CRi (42.9% vs 10.6%; p = 0.007), mCRc (71.4% vs 34%; p = 0.008), and a lower probability of an EFS event (52.4 % vs 91.5%; p = 0.001) or death (42.9% vs 87.2%; p < 0.001) in comparison to non-triple mutated patients. The median OS in the triple-mutated patients was 45 months (95%CI: 4.9 - NA) in comparison to 5.6 months (95%CI: 3.6 – 7.5) in the non-triple mutated patients ( p = 0.0003). Therapy stratification revealed that the 24-month OS for triple-mutated patients treated with combination therapy and monotherapy were 60.0% and 54.5%, respectively. In contrast, the 24-month OS in the non-triple mutated cohort treated with combination therapy and monotherapy were 17.3% and 8.7% respectively. Multivariable analysis using the cox proportional hazards model, the factors positively predicting survival included achievement of mCRc, undergoing transplant and the presence of triple mutation. Exposure to azacytidine-venetoclax predicted for poorer survival. Conclusions: Patients with triple mutation ( FLT3 /NPM1/DNMT3A co-mutation) is a distinct subset which responds particularly well to FLT3-inhibitor. Future studies should explore whether therapy selection (monotherapy or combination therapy) can be decided based on the presence or absence of triple mutation.

Introduction Measurable residual disease (MRD) monitoring is informative for long-term prognosis in patients with acute myeloid leukemia (AML) at the time of remission assessment. While next-generation sequencing-based MRD monitoring in FLT3-ITD AML is being explored for clinical use, its adoption is slow due to technological and cost limitations. Multiparameter flow cytometry (MFC)-based MRD is widely used for MRD assessment across all AML subtypes. In our previous report, patients with lower FLT3-ITD allele frequency and negative MFC-MRD showed 12-month OS of 86% (vs 16.7%, p=0.023; ASH 2024). Concurrent mutations (e.g., NPM1, DNMT3A, WT1) in FLT3-ITD AML has diverse prognostic outcomes. DNMT3A and WT1 predict adverse outcomes, while NPM1 improves relapse-free survival (RFS). We hypothesized that MFC-MRD provides prognostic value in FLT3-ITD AML independently of these mutations and explored whether it stratifies prognosis for RFS and OS regardless of molecular subtype. Patients and method The present study evaluated the outcomes of 111 AML patients with FLT3-ITD diagnosed and treated from 2018 to 2022 at Princess Margaret Cancer Centre. We compared outcomes by MFC-MRD status at remission and molecular subtype (i.e. NPM1, DNMT3A, or WT1 mutation) using a 0.1% MRD cutoff. Primary endpoints were OS and RFS. Secondary endpoints included cumulative incidence of relapse (CIR) and non-relapse mortality (NRM). Kaplan-Meier method with log-rank test and Cox proportional hazard models were used for OS/RFS analyses, while Gray test and Fine-Grey models were used for CIR and NRM. Results A total of 111 patients were included with a mean age of 63.5 years. According to the European LeukemiaNet (ELN) 2022 classification, patients were stratified as favorable (2, 1.8%), intermediate (67, 60.4%), and adverse (42, 37.8%). NPM1 co-mutations were found in 56 patients (50.5%), DNMT3A or WT1 in 39 (35.1%), and combined NPM1 with DNMT3A or WT1 in 28 (25.2%). Ninety patients received treatment. Seventy-two patients (80%) received intensive chemotherapy with midostaurin used in combination for 60 patients (83.3%). Overall response was achieved in 79 patients (87.8%), including 69 (76.7%) with complete remission (CR) or CR with incomplete recovery (CRi). MFC-MRD data was available for 61 patients, with 44 (56.4%) negative and 17 (21.8%) positive. Fifty-four patients achieved CR1 after one induction cycle, highest in those with NPM1 co-mutations but no other founder or MDS-related mutations (92%, p=0.017). With a median follow-up of 437 days among treated patients, 50 (45%) were alive. Two-year OS and RFS rates were 55.6% and 57.1%, respectively. Among FLT3-ITD AML patients achieving CR1, MFC-MRD negativity was associated with improved RFS (2-year RFS 71.2% vs 35.2%, p=0.04), though no OS differences were observed in the overall cohort. Forty-six patients (51.1%) received an allogeneic transplant. Patients were categorized into three groups based on co-mutation profiles: FLT3-ITD and NPM1 mutation; FLT3-ITD and DNMT3A/WT1 without NPM1; and combined FLT3-ITD, NPM1, and DNMT3A/WT1. No differences in OS, RFS, CIR, or NRM were observed among these groups. When evaluating the impact of MFC-MRD status at CR1 in each subgroup according to their co-mutation profiles, we could not observe any significant differences of its impact on OS probably due to small subject numbers in each subgroup. In co-mutation subgroups, FLT3-ITD and NPM1 had a 2-year OS of 68.6% (vs 51.4%, p=0.37) and RFS of 69.2% (vs 53.2%, p=0.34). FLT3-ITD and DNMT3A/WT1 without NPM1 showed 2-year OS of 50% (vs 56%, p=0.98) and RFS of 51.9% (vs 57.8%, p=0.63). Combined FLT3-ITD, NPM1, and DNMT3A/WT1 had 2-year OS of 72.7% (vs 49.6%, p=0.63) and RFS of 48.3% (vs 60.5%, p=0.44). No differences were noted for CIR or NRM. Multivariate analysis for OS, incorporating age at diagnosis, ELN 2022 risk stratification, MFC-MRD status post-induction, and co-mutation profiles, identified MFC-MRD status as the only independent prognostic factor for OS (HR 3.07 [1.24-7.60], p=0.015). No independent factors were identified for RFS, CIR, or NRM. ConclusionThe present study suggests that MFC-based MRD assessment is feasible in AML with FLT3-ITD regardless of the co-mutations in NPM1, DNMT3A, or WT1. MFC-MRD negativity showed better RFS compared to MFC-MRD positivity. However, larger studies are needed to clarify the prognostic impact of MFC-MRD within specific co-mutational subgroups.

Patients with secondary acute myeloid leukemia (sAML) have inferior outcomes compared to de novo AML (dn-AML) when treated with intensive chemotherapy. Azacitidine plus venetoclax (Aza-Ven) is increasingly used in patients with sAML due to its improved efficacy and lower treatment-related mortality, which may also serve as a lower-risk effective approach to bridge eligible patients to allogeneic stem cell transplantation. In this study, we aimed to evaluate the outcomes of Aza-Ven in patients with therapy-related AML (t-AML) or myelodysplasia-related AML (AML-MR) at our center. We performed a retrospective study of adults with AML treated with frontline Aza-Ven at the Princess Margaret Cancer Center (Toronto, Canada) between 2017 and 2024. They were categorized into three mutually exclusive groups: t-AML, AML-MR and dn-AML. Patients who previously received cytotoxic therapy or radiotherapy were categorized as t-AML. Patients with prior myelodysplastic syndrome (MDS) or MDS/myeloproliferative neoplasm (MPN) or with MDS-related gene mutations (MRGM) or cytogenetic abnormalities (MRCA) were categorized as AML-MR. Remaining patients were categorized as dn-AML. Composite complete remission (CRc) was defined as complete remission (CR) and CR with incomplete or partial hematological recovery (CRi/CRh). Overall response rate (ORR) was defined as CRc and morphological leukemia-free state (MLFS). Kaplan–Meier analysis with log-rank tests were used to compare overall survival (OS) between subgroups. A total of 132 patients were included: 29 with t-AML, 73 with AML-MR, and 30 with dn-AML. Compared to dn-AML, patients with t-AML had similar age (median 70 vs 76 years, p=0.16), ECOG 0–1 status (74% vs 57%, p=0.26) and non-favorable ELN 2024 risk (52% vs 57%, p=0.79), but more TP53 mutations (29% vs 4%, p=0.03), complex karyotype (CK) (45% vs 13%, p<0.01) and non-favorable ELN 2022 risk (83% vs 53%, p=0.03). CRc rate at any time was 71% vs 69% in patients with t-AML and dn-AML, respectively (p=0.25), including 58% vs 59% after cycle 1 (p=0.25). ORR at any time was 75% vs 86% in patients with t-AML and dn-AML, respectively (p=0.15), including 67% vs 79% after cycle 1 (p=0.11). 60-day mortality was 17% vs 7% in t-AML and dn-AML, respectively (p=0.25). Median OS was 13.1 months in t-AML and 16.0 months in dn-AML, with 24-month OS rates of 29% and 48%, respectively (p=0.30). Within the t-AML subgroup, TP53 mutation (HR 2.67, 95% CI 0.99 – 7.15, p=0.05) and ELN 2022 non-favorable risk (HR 8.07, 95% CI 0.99–65.14, p=0.05) were marginally associated with inferior OS, while CK had a non-significant trend towards inferior OS (HR 1.79, 95% CI, 0.68-4.73, p=0.24). Prior chemotherapy or radiotherapy exposure, non-favorable ELN 2024 and presence of MRGM were not associated with OS in t-AML. Compared to dn-AML, patients with AML-MR had comparable age (median 74 vs 76 years, p=0.36), ECOG 0–1 status (74% vs 57%, p=0.14), and non-favorable ELN 2024 risk (45% vs 57%, p=0.37), but more TP53 mutations (18% vs 4%, p=0.11), CK (32% vs 13%, p=0.04) and non-favorable ELN 2022 risk (97% vs 53%, p<0.01). CRc rate at any time was 65% vs 69% in patients with AML-MR and dn-AML, respectively (p=0.82), including 45% vs 59% after cycle 1 (p=0.11). ORR at any time was 87% vs 86%, in patients with AML-MR and dn-AML, respectively (p=0.92), including 69% vs 79% after cycle 1 (p=0.15). 60-day mortality was 7% in both groups (p=1.00). Median OS was 12.4 months in AML-MR and 16.0 months in dn-AML, with 24-month OS rates of 33% and 48%, respectively (p=0.30). Within the AML-MR subgroup, signaling pathway mutations (FLT3-ITD, NRAS/KRAS) were associated with inferior OS (HR 3.32, 95% CI 1.66-6.63, p<0.01) along with ELN 2024 non-favorable-risk (HR 1.99, 95% CI 1.02–3.90, p=0.04). TP53 mutations were not associated with OS (HR 1.18, 95% CI, 0.52-2.58) along with non-favorable ELN 2022 risk, individual MRGM and CK. Patients with sAML in our cohort achieved remission rates comparable to dn-AML with frontline Aza-Ven with non-significant difference in OS. The ELN 2022 and 2024 risk classifications had different prognostic utility between t-AML and AML-MR mostly related to the inclusion of cytogenetic abnormalities in the former and weight of FLT3-ITD and NRAS/KRAS mutations in the latter. Further studies are needed to refine and adapt prognostic classification systems considering the heterogeneity of sAML in patients treated with Aza-Ven.

BACKGROUND:Central nervous system infiltration (CNS+) leads to serious complications in AML. Molecular characteristics associated with CNS disease is not well defined. MATERIAL AND METHODS:259 patients received lumbar puncture (LP) and prophylaxis intrathecal (IT) chemotherapy based on neurological symptoms and / or hyperleukocytosis RESULTS: : 31 (11.97%) patients were confirmed CNS+. The most common mutation by Next-generation sequencing (NGS) were NPM1 (48.4%), FLT3 (29.0%), TET2 (25.8%), DNMT3A (19.4%), ASXL1 (16.1%). Univariate analysis, NPM1 (OR: 2.880), FLT3-ITD (OR:3.222), and their combination (OR:3.912) revealed higher incidence in CNS+ . Multivariable analysis, confirmed independent associations with CNS+: WBC ≥100 × 10⁹/L (OR 5.61, P = .0001), WBC ≥30 (OR 3.73, P = .0039), WBC ≥40 (OR 2.93, P = .0149), LDH & NPM1 combined (OR: 2.503) and the presence of an 11q23 abnormality (OR:7.569). WBC count of ≥30 showed the largest separation with CNS+ being 71.0% and CNS- being 29.0% of the cases although it may capture lower-risk patients. In the meantime, a threshold of ≥40 maintained a strong association with CNS+ (64.5%) while improving specificity (35.5%). There is no significant difference in overall survival (OS) (47% vs. 45%) (P = .90) and cumulative incidence of relapse (CIR) (29.0% vs. 15.6%) (P = .06) between CNS positive and negative groups at 36 months CONCLUSION: : Routine screening LP and prophylactic IT chemotherapy should be considered in high-risk AML patients (WBC ≥40×109 /L, elevated LDH and NPM1 mutation, or 11q23 abnormality) to assess for CNS+ and likely preventing the occurrence of CNS relapse. A larger prospective trial is needed to confirm the results.

Azacitidine plus venetoclax (Aza-Ven) is now standard of care for newly diagnosed acute myeloid leukemia (AML) in patients (pts) ineligible for intensive chemotherapy. Whilst this is better tolerated than intensive induction therapy, myelosuppression is common and can be severe. In this study, we sought to investigate the impact of incomplete hematological recovery after Aza-Ven on long-term survival. We conducted a retrospective study of pts diagnosed with AML, who received front-line treatment with Aza-Ven at the Princess Margaret Cancer Centre between 2017 and 2024. Reponses were categorised based on the 2022 European Leukemia Network (ELN) recommendations for AML. We defined composite complete remission (CRc) rate as a combination of complete remission (CR) and CR with incomplete hematological recovery (CRi). We defined morphologic composite remission (mCR) as achievement of < 5% bone marrow blasts including CR, CRi and morphologic leukemia-free state (MLFS). The primary outcome was overall survival (OS). Our cohort consisted of 132 cases, with 82 (62%) male pts. The median age was 74 years (range, 34 – 90). Based on the International Consensus Classification (ICC), 127 (96%) pts met the diagnostic criterion for AML; five (4%) pts were classified as myelodysplastic syndrome/AML (MDS/AML). In this real-world cohort, 83 (63%) pts met VIALE-A eligibility criteria. Most pts received 28 days of venetoclax (n=106, 80%) with cycle 1. Bone marrow assessments (BMA) were performed in 117 (89%) pts after cycle 1. In evaluable pts with BMA after cycle 1, 59 (50%) achieved CRc, including 31 (26%) in CR and 28 (24%) in CRi. A total of 82 (70%) achieved mCR, including pts in CRc plus 23 (20%) pts in MLFS. Nine (8%) pts had partial response (PR) and 24 (21%) pts had no response (NR). From initiation of cycle one, 30-day and 60-day mortality were 2% and 8%, respectively. The median OS in our whole cohort was 13 months (95% CI: 11 – 16 months). To better characterise the heterogeneity within pts achieving mCR (n=82), subgroup analysis was performed. Pts who achieved MLFS were more likely to have had previous exposure to hypomethylating agents (HMA) (13% vs 0%, p<0.01) as well as mutations in ASXL1 (42.9% vs 19%, p=0.03), STAG2 (23.8% vs 3.4%, p<0.01) and TET2 (38.1% vs 15.5%, p=0.03). Pts achieving CR/CRi also more commonly had favourable risk disease as per ELN 2024 (49.2% vs 26.1%, p=0.03), compared to those who achieved MLFS. There was no significant association between achievement of MLFS, and prior history of MDS or MDS/MPN (p=0.11) or previous cytotoxic therapy (p=0.12). Mutations in FLT3-TKD (p=0.04) and STAG2 (p=0.03) were more common among pts with delayed neutrophil recovery (≥42 days from cycle 1 day 1). Conversely, mutations in FLT3-ITD (p=0.02), RUNX1 (p=0.03), and STAG2 (p=0.03) were more frequent in pts with delayed platelet recovery (≥28 days from cycle 1 day 1). There was no significant difference in measurable residual disease (MRD) positivity assessed by flow cytometry between pts who achieved CRc versus MLFS (38% vs 53%, p=0.26) and MRD status was not associated with OS (HR 0.89, 95% CI 0.42-1.89, p=0.77) among pts achieving mCR. There was no significant relationship between reduced duration of venetoclax (<25 days), MLFS, neutrophil or platelet recovery. However, due to the relatively small number of pts who received the shorter course (n=22/132), definitive conclusions cannot be drawn. Pts achieving CRc (n=59) after cycle 1 had significantly longer median OS at 22.5 months, compared to those who achieved MLFS (n=23) at 6.6 months (HR 0.45, 95% CI: 0.24 – 0.85, p=0.01). The 2-year OS rates in pts achieving CRc and MLFS were 46% and 27%, respectively. Median relapse-free survival (RFS) in pts achieving CRc and MLFS were 25.0 months and 7.9 months respectively (HR 0.62, 95% CI: 0.29 – 1.32, p=0.22). The 2-year RFS in pts achieving CRc and MLFS were 52% and 38%, respectively.Conclusion: Patients who achieve MLFS after Aza-Ven have significantly worse OS compared with those who achieve CRc. There was no significant difference in MRD positivity between pts achieving CRc or MLFS, but mutational profile may identify pts at risk of delayed or deficient hematological recovery. Future research should focus on identifying strategies to improve bone marrow failure beyond the achievement of bone marrow blast reduction.

MPAL is a rare, heterogeneous leukemia defined by myeloid and lymphoid marker co-expression. The 2022 WHO classifies subtypes as B/My, T/My, or those with BCR::ABL1 or KMT2A-r. MPAL lacks standardized treatment and has poor outcomes; its prognostic and therapeutic alignment with ALL subtypes remains unclear. Aim To characterize clinical features and treatment outcomes of MPAL patients(pts) treated with intensive chemotherapy. Methods We conducted a retrospective single-center study of adults (≥18 y) diagnosed with MPAL (2003–2023), classified per WHO 2022. Induction regimens were categorized as ALL-like (PMH-DFCI or HyperCVAD), AML-like (anthracycline + cytarabine), or FLAG-IDA; BCR::ABL1+ pts received TKIs. OS was measured from diagnosis to death or last follow-up; CIR from CR to relapse or death. Standard statistical methods were used. Each MPAL subtype was compared to its phenotypically similar ALL counterpart using 1:2 propensity score matchings, balanced for age (<60 vs. ≥60 years), gender, presence or absence of extramedullary disease (EM), and white blood cell count at diagnosis (<30 vs. ≥30 ×10⁹/L). Subgroup analyses included comparisons of B/My MPAL vs. B-ALL, T/My MPAL vs. T-ALL, and BCR::ABL1+ MPAL vs. Philadelphia chromosome-positive (Ph+) ALL Results Among 841 ALL pts treated with intensive chemotherapy, 64 cases were re-reviewed by two hematopathologists due to suspicion of MPAL. 41 pts (5%) met MPAL criteria per WHO 2022. Median age at diagnosis was 51 years (IQR 40–62); 29% were aged ≥60 years, and 59% were male. The most common MPAL subtype was B/My: 19 (46%), followed by T/My: 10 (24%), BCR::ABL1mut: 7 (17%), KMT2Ar: 3 (7%), and other rare subtypes: 2 (5%). Among 36 pts with evaluable cytogenetics, 11 (27%) had complex karyotype. EM disease was present in 10 pts (24%). Induction regimens included ALL-like protocols in 26 pts (63%), FLAG-IDA in 13 (32%), and AML-like in 2 (5%). Intrathecal chemotherapy was administered in 35 pts (85%) Thirty-three pts (80.5%) achieved CR after first induction: 21/26 (81%) with ALL-like protocols, 11/13 (85%) with FLAG-IDA, and 1/2 (50%) with AML-like regimens (p=0.503). Of the 8 pts (19.5%) who failed to achieve CR, 7 received reinduction therapy: 1 (12.5%) with an ALL-like regimen, 3 (37.5%) with AML-like, 1 (12.5%) with FLAG-IDA, and 2 (25%) with venetoclax-based regimens. Of these, 6 (85.7%) achieved CR after reinduction. Fourteen of 39 responders (36%) relapsed, with a median time to relapse of 10 (2–46) m. Immunophenotypic data at relapse were available for 8 pts (57%): 1 relapsed with the same phenotype; 2 shifted to B-ALL (both initially B/My and treated with ALL-like regimens at diagnosis); 1 to T-ALL (initially B/ T/My also treated with an ALL-like regimen); and 4 to AML (2 initially B/My, 2 T/My all treated with FLAG-IDA at diagnosis). Median follow-up was 75mo(IQR 41–88). OS was 80% (95% CI, 64–90) at 1 y and 54% (37–68) at 5 y. CIR was 10% (2–53) at 1 y and 24% (3–67) at 5 y. In the PSM analysis, MPAL had numerically inferior OS: 1 y OS was 78% (61–88) vs. 86% (82–88), and 5 y OS 54% (37–68) vs. 65% (61–68) (HR 1.52, 1–2.4; p=0.0659). CIR was significantly higher in MPAL at 1 y (24% [12–43] vs. 12% [10–15]) and 5 y (46% [29–67] vs. 33% [29–37]; HR 2.3, 1.5–3.7; p=0.0003). By subtype, B/My MPAL (n=19) had outcomes comparable to B-ALL (n=38): 1 y OS 90% (64–97) vs. 82% (65–92), 5 y OS 74% (45–88) vs. 57% (37–73) (HR 0.71, 0.3–1.7; p=0.4507). CIR was also comparable at 1 y: 27% (11–56) vs. 24% (13–42) (HR 0.61, 0.3–1.5; p=0.2645). T/My MPAL (n=10) showed a tendency toward lower OS vs. T-ALL (n=20): 1 y OS 68% (29–88) vs. 73% (46–88), 5 y OS 28% (4–59) vs. 56% (31–75) (HR 1.74, 0.6–5; p=0.2996); 1 y CIR was comparable (21% vs. 27%). BCR::ABL1+ MPAL (n=7) and Ph+ ALL (n=14) had similar survival: 1 y OS 86% vs. 93%, 5 y OS 57% vs. 71% (HR 1.43, 0.32–6.40; p=0.6408), with identical 1 y CIR (14%; HR 0.99, 0.2–4.0; p=0.9913). In PSM censoring for allo-HSCT, outcomes remained consistent except for T/My MPAL, which had significantly inferior OS vs. T-ALL (1 y OS 30% [4–62] vs. 67% [41–82]; HR 4.22, 1.3–14; p=0.0105), with no survivors beyond 5 y. Conclusion MPAL showed a trend toward worse OS and higher CIR vs. lineage-defined ALL, with significantly inferior OS in the T/My subtype post-transplant censoring. These findings underscore its high-risk nature and the need for collaborative studies to refine subtype-specific, risk-adapted treatment strategies for MPAL.

IL-23 receptor (IL-23R) is canonically a T cell surface cytokine receptor known for its role in inflammatory diseases. In this study, we demonstrated that IL-23R localizes intracellularly in AML where it regulates mitotic spindle formation. Differential gene expressions were analyzed in AML samples compared to normal controls. Of the upregulated processes, the mitotic spindle ontology was most enriched. Analyzing all protein coding genes for their correlation to the mitotic spindle ontology surprisingly returned IL-23R as a top hit. We confirmed IL-23R expression in 7 of 7 tested AML cell lines and 15 of 20 primary AML patient samples. Protein expression was >2 fold higher in AML compared to normal hematopoietic (n=5) and CD34+ sorted cells (n=3).Canonical IL-23R cell surface expression was confirmed in T cells, but minimal amounts were detected on AML cells. Rather, in AML cells, primary AML samples, and AML stem cells, IL-23R was detected intracellularly as determined by 4 different modalities using 4 different epitope targeting antibodies. IL-23R’s heterodimeric subunit, IL-12Rβ1, was also detected intracellularly in AML cells. We performed BioID mass spectrometry to identify novel interactors with IL-23R. Pathway analysis of interactors returned mitotic spindle formation as a top pathway. We confirmed interactions between IL-23R and mitotic spindle proteins (NuMA, TMEM201, TACC1, BAG6) via proximity ligation assays in bulk and leukemic stem cell (LSC) sorted AML cells. Furthermore, in bulk AML and LSC sorted fractions, we demonstrated that IL-23R colocalized with the mitotic spindle and centrosomes via confocal microscopy. IL-23R interacted with the mitotic spindle via its (S/T)x(I/L)P motif (amino acids 588-591). IL-12Rβ1 was also found at the mitotic spindle with exogenous IL-23 in culture promoting intracellular translocation of IL-23R. Depletion of IL-23R led to dysregulation of the mitotic spindle with spindle defects such as multipolarity and lagging chromosomes. Additionally, in IL-23R depleted mitotic AML cells, NuMA has decreased localization to mitotic spindle poles thus attenuating its ability to regulate mitotic spindle assembly. Knockdown of IL-23R reduced proliferation and clonogenic growth, reduced marrow engraftment, and increased differentiation of AML cells, thus demonstrating an effect on the leukemia initiating cells. In normal hematopoietic cells, however, knockdown of IL-23R did not impair engraftment. Consistent with this, constitutive homozygote IL-23R knockout mice had no difference in complete blood counts and stem cell number/function compared to wild type mice. In summary, intracellular IL-23R regulates mitotic spindle formation to maintain AML and stem cell growth and viability. We demonstrated a novel function and intracellular localization for IL-23R and its potential as a therapeutic target in AML. Nathan Duong, Dilshad H. Khan, Geethu E. Thomas, Yue Feng, Rose Hurren, Jong Bok Lee, Jonathan St-Germain, Lily Drimmer, Yongran Yan, Lan Xin Zhang, Dakai Ling, Mary L. Ma, Neil MacLean, Marcela Gronda, Vincent Rondeau, Brandon D. Brown, Laura Matellán, Courtney L. Jones, Hong Chang, Andrea Arruda, Stephanie Xie, Laurence Pelletier, Mark D. Minden, Li Zhang, Steven M. Kornblau, Brian Raught, Kevin Jacobs, Max G. Jacobs, Daniel Goede, Vito Spadavecchio, Aaron D. Schimmer. The IL-23 receptor localizes intracellularly in AML and stem cells where it regulates mitotic spindle formation to maintain cell viability [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 5411.

Clonal haematopoiesis arises when a haematopoietic stem cell (HSC) acquires a mutation that confers a competitive advantage over wild-type HSCs, resulting in its clonal expansion. Individuals with clonal haematopoiesis are at increased risk of developing haematologic neoplasms and other age-related inflammatory illnesses1-4. Suppressing the expansion of mutant HSCs may prevent these outcomes; however, such interventions have not yet been identified. The most common clonal haematopoiesis driver mutations are in the DNMT3A gene, with arginine 882 (R882) being a mutation hotspot1-3,5-7. Here we show that mouse haematopoietic stem and progenitor cells (HSPCs) carrying the Dnmt3aR878H/+ mutation, equivalent to human DNMT3AR882H/+, have increased mitochondrial respiration compared with wild-type cells and are dependent on this metabolic reprogramming for their competitive advantage. Treatment with metformin, an anti-diabetic drug that inhibits mitochondrial respiration8, reduced the competitive advantage of Dnmt3aR878H/+ HSCs. Through a multi-omics approach, we found that metformin acts by enhancing methylation potential in Dnmt3aR878H/+ HSPCs and reversing the aberrant DNA CpG methylation and histone H3 K27 trimethylation profiles in these cells. Metformin also reduced the competitive advantage of human DNMT3AR882H HSPCs generated by prime editing. Our findings provide preclinical rationale for investigating metformin as a preventive intervention against DNMT3A R882 mutation-driven clonal haematopoiesis in humans.

There are fundamental differences in the genomic drivers of childhood, adolescent and young adult (AYA) acute myeloid leukemia (AML) compared to AML that develops in older adults. Despite this, little is known about the downstream pathways that regulate leukemogenesis in young AML. Although genetic alterations in the nuclear pore complex are more common in young patients (e.g., NUP98 rearrangements, DEK-NUP214 fusions), nucleocytoplasmic transport has not been exclusively studied in childhood/AYA AML. We identified enrichment of nucleocytoplasmic transport pathways in young (<40 years; mean age 27 in BeatAML & 29 in TGCA) versus older AML patients (>40 years, mean age 64 in BeatAML & 61 in TCGA) (BeatAML FDR=0.01, TCGA FDR=3.46e-3). Young patients with nucleoporin (NUP) rearrangements had high expression of nucleocytoplasmic transport genes, but this pathway was also upregulated in young NUP98r/DEK-NUP214 wildtype patients. Of the nucleocytoplasmic export genes, CSE1L was the top differentially expressed exportin in young AML. Immunoblotting confirmed increased CSE1L protein expression in childhood/AYA (n=10, mean age 23) compared to older (n=10, mean age 77) AML patient samples. Increased CSE1L mRNA expression predicted for decreased remission rates, event free survival and overall survival, exclusively in young, but not older patients with AML (TARGET, Beat AML). To further understand the dependency of childhood/AYA AML on CSE1L, we knocked down CSE1L with shRNA in cell lines derived from young AML patients (NB4, MV4-11, THP-1) and TEX cells, derived from transducing cord blood with a FUS-ERG fusion. CSE1L depletion reduced AML growth, viability and clonogenicity. Loss of AML viability and clonogenicity was confirmed using CSE1L CRISPR knockout. Knockdown of CSE1L in primary young AML cells reduced primary and secondary leukemic engraftment into mouse marrow. In contrast, depletion of CSE1L did not impair engraftment of normal cord blood into mouse marrow. To understand the function of CSE1L, we interrogated a BioID database and discovered that the CSE1L interactome was enriched for ribosome biogenesis processes (PXD007976). We validated the interaction with BioID hits, RPL29 and RPL32 (large ribosome subunit (60S) proteins), using proximity ligation assay and Co-IP. Co-IP further confirmed that CSE1L interacts exclusively with 60S, but not 40S ribosomal subunits. Knockdownof CSE1L in TEX cells led to nuclear accumulation of RPL29 and a 3.3 fold reduction in cytoplasmic RPL29. Using a GFP reporter assay that incorporates into the 60S subunit, we induced depletion of CSE1L and confirmed decreased cytoplasmic RPL29-GFP. Polysome profiling further validated a role for CSE1L in 60S export, as CSE1L depletion decreased cytoplasmic 60S but not 40S formation. Downstream of 60S export, a reduction in protein synthesis was observed in CSE1L depleted AML cells, relative to control shRNA cells. Given increased CSE1L expression in young patients with AML and its role in ribosome biogenesis, next we analyzed the top 20 differentially upregulated gene sets in young versus older AML patient samples and discovered that ribosome biogenesis pathways were the most abundantly expressed genes in AML blasts of young patients. GO processes highly enriched in young AML cells included rRNA processing, pre-ribosome, large ribosome subunit and ribosome biogenesis. Functionally, we demonstrated increased rRNA transcription rates in young (n = 10, mean age = 16.8 years) relative to older primary AML samples (n = 10, mean age = 69.7 years ). Next, we treated these samples with the RNA polymerase I inhibitor, BMH-21, and identified a differential sensitivity to BMH-21 in young AML cells. In summary, nucleocytoplasmic transport and ribosome biogenesis are positively correlated, differentially upregulated pathways in childhood/AYA AML. Depleting CSE1L, leads to impaired large ribosome subunit export, decreased protein synthesis and leukemic cell death. Likewise, AML cells of young patients have increased rates of ribosome biogenesis and are more sensitive to RNA polymerase I inhibitors than AML cells from older patients. Our findings confirm that childhood and older adult AML are fundamentally different diseases, driven by different biologic processes and reinforce the need to develop pediatric-specific treatment approaches to manage this disease in young patients.

Context: FLT3 mutated (FLT3m) AML with co-occurring mutations in DNMT3a and NPM1 (triple mutated) have been identified to be a subset with poor prognosis. However, the evidence for this comes from trials predating the routine incorporation of FLT3 inhibitors in induction regimens. Triple mutated (TM+) patients in the relapsed-refractory setting are particularly sensitive to FLT3 inhibitor therapy. To explore whether the outcomes of TM+ patients were improved by the addition of midostaurin to intensive chemotherapy, we conducted a real-world, single-center analysis of FLT3m patients. Design: Retrospective single center study Patients and methods: Patients ≥ 18 years of age diagnosed with FLT3m AML who were treated with intensive chemotherapy at the Princess Margaret Cancer Centre between January 1st 2015 and January 31, 2023, with available NGS data were included. RFS and OS was evaluated using Kaplan-Meier method. IPTW-adjusted Kaplan Meier analysis was performed to compare survival outcome between non-midostaurin (M-) and midostaurin treated (M+) population, stratified for the TM status(TM+ vs TM-). Propensity scores were estimated using clinically relevant covariates (age, FLT3m type, de-novo versus secondary AML, baseline MRC cytogenetic group, timing of transplant (CR1 versus CR2)), and survival curves were weighted accordingly. Results: We identified 242 patients with FLT3m AML. Ninety-two (38%) did not receive a FLT3 inhibitor with their induction regime (M-) and 150 (62.0%) received midostaurin (M+) with the intensive chemotherapy. Amongst the 92 M- patients, median age was 61 years [21 – 81], 75(81.5%) had denovo AML, 17(18.5%) had secondary AML, 66(71.7%) had intermediate risk cytogenetics and 6(6.5%) had adverse risk cytogenetics. FLT3m were ITD in 63(68.5%) and TKD in 11(12.0%) patients. A CR/CR equivalent was achieved in 84 patients (91.3%). After a median follow-up of 91[95%CI: 78 – 99] months, there were 29(34.5%) relapses and 49(53.3%) deaths. A total of 50(54.3%) patients underwent HSCT. 3-year RFS and OS (uncensored) was 65%[53 - 75] and 52%[42 – 62], respectively. Mutations in addition to FLT3m, included NPM1 in 51(55.4%), DNMT3a in 35(38.0%), IDH2 in 18(19.6%), TET2 in 15(16.3%), WT1 in 11(12.0%) and IDH1 in 8(8.7%). There were 28(30.4%) patients with TM+ AML patients. On comparing TM+ with TM-, the 3-year RFS was 57%[35 – 75] and 68%[54 – 79](p 0.327), respectively and the 3-year OS (uncensored) was 39%[22 – 57] and 58%[45 – 69] (p 0.146), respectively. On censoring for transplant, the OS was significantly lower in the TM+ subgroup (3 -year OS: 13% vs 58%; p 0.048). Amongst the 150 M+ patients, median age was 58[18 – 80] years , 138(92.0%) had denovo AML, 12(8.0%) had secondary AML, 113(75.3%) had intermediate risk cytogenetics and 12(8.0%) had adverse risk cytogenetics. FLT3m were ITD in 117(78.0%) and TKD in 32(21.3%). A CR/CR equivalent was achieved in 145(96.7 %) patients. After a median follow-up of 54[50 - 64] months, there were 61(42.1%) relapses and 66(44.0%) deaths. A total of 93(62.0%) patients underwent HSCT. 3-year RFS and OS was 56%[47 – 64] and 59%[50 – 67], respectively. Mutations in addition to FLT3m, included NPM1 in 96(64.0%), DNMT3a in 54 (36.0%), TET2 in 22(14.7%), IDH2 in 21(14.0%), WT1 in 11(7.3%) and IDH1 in 10(6.7%) patients. There were 44(29.3%) TM+ AML patients. On comparing the patients with TM+ with TM-, the 3-year RFS was 55%[38 – 69] and 56%[45 – 66] (p 0.777), respectively and the 3-year OS (uncensored) was 67%[51 – 79] and 55%[45 – 64] (p 0.328), respectively. On censoring for transplant, the OS was similar between the TM+ and TM- subgroups (3-year OS: 38% vs 35%; p 0.662). Using IPTW-adjusted Kaplan-Meier analysis, the 3-year OS of TM+ patients were 65%[51 – 83] versus 36%[21 – 60] in M+ and M- patients (HR: 0.44 [0.22 – 0.87]; p = 0.019), and of TM- patients were 53%[43.9 – 64.8] versus 54% [43 – 70] in M+ and M- patients, respectively (HR: 0.89 [0.55 – 1.45]; p 0.65). Conclusions:In FLT3m AML, use of midostaurin during intensive treatment appears to attenuate the adverse prognostic impact of the TM+ patients. OS was inferior in the TM+ compared to TM- patients when midostaurin was not administered – particularly evident after censoring for transplant. IPTW-adjusted analysis also showed significant improvement in OS with the use of midostaurin in TM+ patients. These findings suggest that the efficacy of FLT3 inhibitors may be modulated by the co-mutations.

Background/Objectives: The 2022 European LeukemiaNet (ELN 2022) is a widely used genotypic risk classification tool for the treatment and prognostication of acute myeloid leukemia (AML) patients. Our study evaluates its effectiveness in categorizing adverse-risk AML patients on standard therapy based on their overall survival (OS). Methods: We conducted a retrospective study involving 256 AML patients. Results: Those in the ELN 2022 adverse-risk group had the shortest OS (p < 0.0001) and were predominantly characterized by myelodysplasia-related (MR) mutations, complex karyotype (CK), monosomal karyotype (MK), and TP53 mutation (TP53 Mut). Subclassification and analysis of this adverse-risk group based on the TP53 Mut status revealed a significantly shorter OS compared to the adverse TP53 wild-type (TP53 WT) counterparts (p = 0.0036). We propose refining the ELN 2022 adverse-risk group into two categories, adverse TP53 Mut and adverse TP53 WT groups, to represent adverse- and ultra-adverse-risk groups, respectively. We used an external validation dataset to confirm our findings. Conclusions: This refinement allows for a more accurate classification of these adverse-risk patients based on their clinical outcomes.

We conducted a multi-center, open-label, randomized phase II study to assess the efficacy of Nivolumab as maintenance therapy for patients with AML in first complete remission (CR) or CR with incomplete hematologic recovery (CRi) who were not candidates for SCT. Patients were stratified and randomized to Observation (Obs) or Nivolumab (Nivo, 3mg/kg IV every 2 weeks for 46 doses). The primary endpoint was progression-free survival (PFS) defined as time to disease relapse or death due to any reason. Secondary endpoints included overall survival (OS), and evaluation of adverse events following Nivolumab administration. Eighty patients were enrolled with median duration of follow-up of 24 months (33 months among survivors). PFS was 13.2 months in the Nivolumab arm (95% CI: 8.5-21.8) and 10.9 months in the Observation arm (5.4-14.9 months). Overall PFS curves were not statistically significantly different ((Nivo/Obs)= 0.92; 95% CI: 0.54, 1.56; one-sided p = 0.38). The median OS was 53.9 months in the Nivolumab arm and 30.9 months in the Observation arm. Cox regression model HR (Nivo/Obs)= 0.78; 95% CI: 0.40, 1.51; p=0.23 (one-sided). There were more adverse events (AEs) of any type (regardless of attribution) on the Nivolumab arm; 27 (71%) patients on the Nivolumab arm had a grade 3 or higher AE compared to 5 patients (12%) on the Observation arm (p<0.001). Nivolumab maintenance after AML chemotherapy failed to improve the PFS and OS in this randomized Phase II study. There were increased AEs and SAEs with nivolumab, but these AEs and SAEs were expected and manageable. ClinicalTrials.gov ID NCT02275533