Hashmi, Shahrukh

Background Graft-versus-host disease (GVHD) remains a major determinant of morbidity and mortality following allogeneic hematopoietic stem cell transplantation (allo-HSCT). Existing GVHD prediction models demonstrate modest discrimination and limited generalizability, and calibration drift across external populations is rarely characterized despite its essential role in the clinical interpretability of predicted probabilities. Objectives To develop and externally validate an explainable machine learning framework for predicting acute and chronic GVHD and associated overall survival in patients with acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL), and myelodysplastic syndromes (MDS) undergoing allo-HSCT, and to systematically characterize calibration across heterogeneous external validation cohorts to inform deployment requirements. Study Design The model was developed on three publicly available registry-derived datasets (N = 2,509) and externally validated across six independent cohorts (N = 14,788) comprising adult and pediatric allo-HSCT recipients, including a regional Middle Eastern cohort (UAE and Jordan). A standardized preprocessing pipeline harmonized heterogeneous datasets. Gradient boosting models (CatBoost) were used for binary GVHD prediction; exploratory overall survival analysis used a Cox proportional hazards model with predicted acute GVHD risk as a covariate. Discrimination (AUROC with bootstrap 95% CI), calibration (logistic recalibration intercept and slope with analytical 95% CI), and feature importance (SHapley Additive exPlanations, SHAP) were assessed in training out-of-fold and all external cohorts. Results In internal validation, AUROC was 0.63 (95% CI 0.61–0.65) for acute GVHD and 0.72 (95% CI 0.70–0.74) for chronic GVHD. External validation demonstrated AUROC ranges of 0.51–0.57 (acute) and 0.54–0.64 (chronic), with consistent performance across disease subgroups despite substantial heterogeneity in transplant practices and feature availability. In exploratory survival analysis, the acute-GVHD-informed Cox model achieved a training-cohort C-index of 0.679 (95% CI 0.658–0.697); external C-indices ranged from 0.47–0.53. Calibration analysis identified systematic external risk overestimation (negative calibration intercept in 10 of 11 evaluable external cohort-target combinations) with heterogeneous slope drift requiring cohort-specific recalibration. Key predictors included recipient age, graft source, conditioning intensity, GVHD prophylaxis, and HLA match ratio. Conclusions An explainable, externally validated GVHD prediction framework was developed using heterogeneous registry-derived datasets, with systematic characterization of calibration drift across multiple external cohorts, an analysis rarely reported in prior GVHD prediction literature. Predictive performance was modest for acute GVHD and moderate for chronic GVHD, constrained by missing immunobiological variables and incomplete HLA characterization. Per-cohort recalibration is required before clinical deployment, with prospective validation and benchmarking against established GVHD risk scores identified as priority next steps.

ABSTRACTBackgroundIn this study, we compared outcomes of intensified myeloablative conditioning regimens using large registry data from Japan (Japanese Society for Transplantation and Cellular Therapy) and the United States (Center for International Blood and Marrow Transplant Research).MethodsAdult patients who underwent their first myeloablative allogeneic hematopoietic stem cell transplantation (HSCT) for acute leukemia in remission between 2010 and 2018 using conditioning regimens of cyclophosphamide plus total‐body irradiation (CY/TBI), CY/TBI+cytarabine (AraC), or CY/TBI+etoposide (VP16) were included.ResultsThe acute myeloid leukemia (AML) cohort (N = 480, 38.8%) indicated that overall survival (OS) was poorer in CY/TBI+AraC (hazard ratio [HR] 1.46, p < 0.001) and CY/TBI+VP16 (HR 1.39, p = 0.059) compared to CY/TBI. Relapse was not suppressed, while treatment‐related mortality (TRM) was significantly higher (HR 1.78 and 1.74, p < 0.001 and 0.018, respectively). In the acute lymphoblastic leukemia (ALL) cohort (N = 3901, 61.2%), OS was comparable among these regimens. With intensified regimens, relapse was significantly suppressed in CY/TBI+VP16 (HR 0.74, p = 0.005), while TRM was higher (HR 1.21, p = 0.077). No interactions were observed regarding the country.ConclusionIn AML adding AraC and VP16 to CY/TBI had an adverse effect on OS. Conversely, in ALL, adding VP16 or AraC to CY/TBI did not affect survival, but the addition of VP16 reduced the risk of relapse.Clinical Trial RegistrationThe authors have confirmed clinical trial registration is not needed for this submission.

AbstractAcute myeloid leukemia (AML) is a challenging cancer in terms of achieving and maintaining long‐duration remissions. Many novel therapies have been added to the standard regimen (combining cytarabine and anthracycline “7 + 3”) to achieve such goals. Nilotinib is an oral multikinase inhibitor that is active against KIT tyrosine kinase, an important stem cell target. In this trial, we combined nilotinib with 7 + 3 induction (daunorubicin 60 mg/m2), high‐dose cytarabine consolidation, and subsequently, if the patient was a candidate, for 2 years' maintenance therapy in patients with AML and KIT (CD117) expression. Patients were allowed to proceed to allogeneic hematopoietic cell transplantation (HCT) if deemed necessary. Our primary goal was increased complete remission rate with this combination. Thirty‐four patients (with a median age 58.5 years) were enrolled on a single‐arm phase II bi‐institutional study; 21 (62%) patients achieved remission. The complete remission rate was 78% in evaluable patients. Thirteen of 34 (38%) patients had allogeneic HCT, all thirteen of which are still alive (100%). Common (>20%) grade 3 non‐hematological toxicities included febrile neutropenia, hypophosphatemia, elevated liver enzymes, and hypertension. Only one patient (3%) died in induction due to liver failure, which was thought secondary to daunorubicin. Our current study reveals good outcomes in patients who received HCT and may warrant a larger study to confirm our findings in that specific population.

Abstract Allogeneic hematopoietic cell transplantation (allo‐HCT) is a potentially curative treatment for acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL). While many factors influence the outcomes of allo‐HCT, the independent impact of donor–recipient ABO mismatching remains unclear. Using the Center for International Blood and Marrow Transplant Research (CIBMTR) database, we identified patients aged ≥18 years with AML or ALL who underwent allo‐HCT between 2008 and 2018. Our objectives were to analyze the outcomes of allo‐HCT based on the donor–recipient ABO status (match, minor mismatch, major mismatch, bidirectional mismatch). Among 4946 eligible patients, 2741 patients (55.4%) were ABO matched, 1030 patients (20.8%) had a minor ABO mismatch, 899 patients (18.1%) had a major ABO mismatch, and 276 patients (5.6%) had a bidirectional ABO mismatch. In multivariable analyses, compared to ABO matched allo‐HCT, the presence of a major ABO mismatch was associated with worse overall survival (HR 1.16, 95% CI 1.05–1.29; p = 0.005), inferior platelet engraftment (HR 0.83, 95% CI 0.77–0.90; p < 0.001), and higher primary graft failure (HR 1.60, 95% CI 1.12–2.30, p = 0.01). Relapse, acute graft versus host disease (GVHD) grades III‐IV and chronic GVHD were not significantly associated with ABO status. While donor age was not significantly associated with outcomes, older recipient age was associated with worse survival and non‐relapse mortality. Our study demonstrates that donor–recipient ABO status is independently associated with survival and other post‐transplantation outcomes in acute leukemia. This underscores the importance of considering the ABO status in donor selection algorithms and its impact in acute leukemia.

Abstract There is a lack of large comparative study on the outcomes of reduced intensity conditioning (RIC) in acute myeloid leukemia (AML) transplantation using fludarabine/busulfan (FB) and fludarabine/melphalan (FM) regimens. Adult AML patients from Center for International Blood and Marrow Transplant Research who received first RIC allo-transplant between 2001 and 2015 were studied. Patients were excluded if they received cord blood or identical twin transplant, total body irradiation in conditioning, or graft-versus-host disease (GVHD) prophylaxis with in vitro T-cell depletion. Primary outcome was overall survival (OS), secondary end points were leukemia-free survival (LFS), nonrelapse mortality (NRM), relapse, and GVHD. Multivariate survival model was used with adjustment for patient, leukemia, and transplant-related factors. A total of 622 patients received FM and 791 received FB RIC. Compared with FB, the FM group had fewer transplant in complete remission (CR), fewer matched sibling donors, and less usage of anti-thymocyte globulin or alemtuzumab. More patients in the FM group received marrow grafts and had transplantation before 2005. OS was significantly lower within the first 3 months posttransplant in the FM group (hazard ratio [HR] = 1.82, P &lt; .001), but was marginally superior beyond 3 months (HR = 0.87, P = .05). LFS was better with FM compared with FB (HR = 0.89, P = .05). NRM was significantly increased in the FM group during the first 3 months of posttransplant (HR = 3.85, P &lt; .001). Long-term relapse was lower with FM (HR = 0.65, P &lt; .001). Analysis restricted to patients with CR showed comparable results. In conclusion, compared with FB, the FM RIC showed a marginally superior long-term OS and LFS and a lower relapse rate. A lower OS early posttransplant within 3 months was largely the result of a higher early NRM.

Paroxysmal nocturnal hemoglobinuria (PNH) is a rare, acquired hematopoietic stem cell disorder, and eculizumab and ravulizumab are its two approved therapies. Only few case series/reports have reported the outcomes of pregnancies in patients with PNH despite the increased risk of thrombosis. Similarly, there is limited knowledge regarding the effect of the approved treatments on conception and pregnancy outcomes. Here, we report the first series of pregnancies in PNH patients from the Middle Eastern region from our tertiary care hospital. Ten pregnancies in four females after diagnosis with PNH were identified. In terms of PNH management, only eculizumab was used, as the safety of ravulizumab use in pregnancies has not yet been established. In the antepartum period, the patients had variable symptoms that ranged from mild symptoms including epistaxis, tea-colored urine and vaginal bleeding to life-threatening vessel thrombosis. Further, red blood cell and platelet transfusions were required because of bleeding and hemolysis in four pregnancies. The pregnancy outcomes varied, but based on these, the safety of eculizumab use during pregnancy remained inconclusive. The postpartum period was complicated in one case by portal vein thrombosis and was managed accordingly. In conclusion, pregnant females with PNH are at an increased risk for complications due to PNH, and thus experienced hematologists and obstetricians should be involved jointly in their care.

AbstractSecondary acute myeloid leukemia (sAML) includes AML as a complication of an antecedent hematological disorder or a therapy‐related AML. Large registry‐based data identified sAML as an independent poor‐outcome type of AML post allogeneic hematopoietic cell transplantation (allo‐HCT). In our study, we tried to define factors affecting outcomes of sAML post allo‐HCT, and identify patients with sAML who may truly benefit from allo‐HCT. We retrospectively analyzed the data of 64 patients aged (14‐61 years) with sAML who received allo‐HCT between September 2010 and February 2018 at our institute. Most of the patients were transplanted from matched related donors (MRD; 54, 84.4%). Our results showed that poor‐risk cytogenetics were identified in 31 patients (48.4%), and their presence was an indicator of poor overall survival (OS) and disease‐free survival (DFS; P‐value = .009, and .004, respectively). The cumulative incidence of chronic graft‐versus‐host disease (cGVHD) was significantly lower in sAML patients with poor‐risk cytogenetics (P‐value = .003) resulting in a high risk of death without cGVHD in this group of patients (P‐value = .02). Besides, GVHD relapse‐free survival (GRFS) analysis showed that most of our studied patients experienced either relapse or debilitating grade II‐IV cGVHD in the first 2 years post allo‐HCT. We conclude that sAML patients with poor‐risk cytogenetics have a significantly lower DFS post allo‐HCT with a high risk of death without active cGVHD.

Abstract Social determinants of health, including poverty, contribute significantly to health outcomes in the United States; however, their impact on pediatric hematopoietic cell transplantation (HCT) outcomes is poorly understood. We aimed to identify the association between neighborhood poverty and HCT outcomes for pediatric allogeneic HCT recipients in the Center for International Blood and Marrow Transplant Research database. We assembled 2 pediatric cohorts undergoing first allogeneic HCT from 2006 to 2015 at age ≤18 years, including 2053 children with malignant disease and 1696 children with nonmalignant disease. Neighborhood poverty exposure was defined a priori per the US Census definition as living in a high-poverty ZIP code (≥20% of persons below 100% federal poverty level) and used as the primary predictor in all analyses. Our primary outcome was overall survival (OS), defined as the time from HCT until death resulting from any cause. Secondary outcomes included relapse and transplantation-related mortality (TRM) in malignant disease, acute and chronic graft-versus-host disease, and infection in the first 100 days post-HCT. Among children undergoing transplantation for nonmalignant disease, neighborhood poverty was not associated with any HCT outcome. Among children undergoing transplantation for malignant disease, neighborhood poverty conferred an increased risk of TRM but was not associated with inferior OS or any other transplantation outcome. Among children with malignant disease, a key secondary finding was that children with Medicaid insurance experienced inferior OS and increased TRM compared with those with private insurance. These data suggest opportunities for future investigation of the effects of household-level poverty exposure on HCT outcomes in pediatric malignant disease to inform care delivery interventions.

Abstract Comparative outcomes of allogeneic hematopoietic cell transplantation (HCT) for BCR-ABL1− myeloproliferative neoplasms (MPNs) in blast phase (MPN-BP) vs de novo acute myeloid leukemia (AML), and AML with prior myelodysplastic syndromes (MDSs; post-MDS AML), are unknown. Using the Center for International Blood and Marrow Transplant Research (CIBMTR) database, we compared HCT outcomes in 177 MPN-BP patients with 4749 patients with de novo AML, and 1104 patients with post-MDS AML, using multivariate regression analysis in 2 separate comparisons. In a multivariate Cox model, no difference in overall survival (OS) or relapse was observed in patients with MPN-BP vs de novo AML with active leukemia at HCT. Patients with MPN-BP in remission had inferior OS in comparison with de novo AML in remission (hazard ratio [HR], 1.40 [95% confidence interval [CI], 1.12-1.76]) due to higher relapse rate (HR, 2.18 [95% CI, 1.69-2.80]). MPN-BP patients had inferior OS (HR, 1.19 [95% CI, 1.00-1.43]) and increased relapse (HR, 1.60 [95% CI, 1.31-1.96]) compared with post-MDS AML. Poor-risk cytogenetics were associated with increased relapse in both comparisons. Peripheral blood grafts were associated with decreased relapse in MPN-BP and post-MDS AML (HR, 0.70 [95% CI, 0.57-0.86]). Nonrelapse mortality (NRM) was similar between MPN-BP vs de novo AML, and MPN-BP vs post-MDS AML. Total-body irradiation–based myeloablative conditioning was associated with higher NRM in both comparisons. Survival of MPN-BP after HCT is inferior to de novo AML in remission and post-MDS AML due to increased relapse. Relapse-prevention strategies are required to optimize HCT outcomes in MPN-BP.