Citations

Abstract Introduction: While newly diagnosed patients (pts) with acute myeloid leukemia (AML) who are ineligible for intensive chemotherapy have prolonged survival outcomes with venetoclax plus azacitidine (VEN+AZA) over AZA alone (2-year overall survival rate: 37.5% vs 16.9%), more can be done to further enhance outcomes (Pratz et al., 2024). CD123 is overexpressed on AML blasts and leukemia stem cells in AML. Pivekimab sunirine (PVEK) is a first-in-class antibody-drug conjugate comprised of a high-affinity anti-CD123 antibody, a cleavable linker, and an indolinobenzodiazepine pseudodimer payload. Previously, in pts with CD123-positive AML, PVEK combined with VEN+AZA demonstrated robust response rates with a manageable safety profile (Daver et al., 2023). Here we report a subgroup analysis from a preliminary cohort of frontline (1L)-treated pts with AML. Methods: Data are from the dose expansion phase of an ongoing, open-label, multicenter, fully enrolled, Phase 1b/2 study of PVEK+VEN+AZA in adults with 1L unfit AML deemed CD123-positive by investigator. Unfit pts were aged ≥75 years (yr), or aged <75 yr with Eastern Cooperative Oncology Group performance status (ECOG PS) score 2–3, or ≥1 defined comorbidity. Pts received PVEK (0.045 mg/kg, intravenous [IV]) on Day (D) 7 plus VEN (400 mg equivalent, oral) daily in a 28-D cycle (cohort 1 received ≥14 D of VEN; cohort 2 received 28 D of VEN) plus AZA (up to 75 mg/m2, subcutaneous or IV) on D1–7. Bone marrow assessments were performed by D12–14 (cohort 1) or D18–21 (cohort 2) to determine VEN duration in cycle 1 and at the end of every subsequent cycle to assess responses and myelosuppression. The primary endpoint was antileukemia activity (complete remission [CR] rate) and measurable residual disease (MRD) assessments. Responses were determined using the modified European LeukemiaNet 2017 criteria (plus CR + partial hematologic recovery [CRh]). MRD-negativity was defined centrally as <0.1% using multiparameter flow cytometry. Safety was assessed as a secondary endpoint. Results: Among 1L unfit pts with CD123-positive AML (n=49), the median (range) age was 77 (58–85) yr, with most pts aged ≥75 yr (n=36 [74%]). Forty-two (86%) pts reported an ECOG PS ≤1, 14% (n=7) had secondary AML, and 20% (n=10) and 51% (n=25) had intermediate or adverse cytogenetic risk, respectively; 65% (n=32) and 35% (n=17) were reported as TP53wt or TP53mut, respectively. Eight (16%) pts bridged to stem cell transplant. The median (range) number of treatment cycles and duration of treatment were 4 (1–20) and 5.5 (0.7–22.7) months (mo), respectively, for PVEK; 4 (1–20) and 5.1 (0.9–22.9) mo for VEN; and 4 (1–21) and 5.7 (0.9–23.4) mo for AZA. For cohort 1 (n=10), the median (range) duration of VEN was 4.5 (2.1–22.9) mo, and for cohort 2 (n=39), the median (range) duration of VEN was 5.7 (0.9–15.2) mo. With a median (range) follow-up of 10 (1.2–26.7) mo, 63.3% (95% CI: 48.3%–76.6%) of pts achieved CR, 79.6% (95% CI: 65.7%–89.8%) achieved CR/CR with incomplete hematologic recovery (CRi), and 73.5% (95% CI: 58.9%–85.1%) achieved CR/CRh; morphologic leukemia free state was achieved in 4% (n=2) of pts. Of pts who achieved CR or CR/CRi with a MRD-evaluable sample, 92% (23/25) and 90% (27/30) achieved flow cytometry MRD-negativity (<0.1%) respectively. The most common (≥50% of pts) any grade treatment-emergent adverse events (AEs) were neutropenia/neutrophil count decreased and thrombocytopenia/platelet count decreased (both 69%); constipation (61%); and peripheral edema (51%). Three pts (6%) experienced AEs leading to discontinuation (myocardial infarction/soft tissue infection/thrombocytopenia) and 2 pts had AEs leading to death (respiratory failure/pancytopenia). Conclusions: Unfit pts with newly diagnosed CD123-postive AML demonstrated high CR rates with PVEK+VEN+AZA. These robust response rates were observed across the select mutational profiles tested. Follow up with duration of response and survival estimates among all pts, by molecular subgroups and MRD negativity status, are being evaluated. Triplet therapy in unfit 1L pts with CD-123–positive AML was well-tolerated, and no new safety signals were observed. These data highlight the need to further evaluate this regimen in a randomized trial.

Abstract Background: Prolonged cytopenias are a well-recognized complication following CAR-T therapy, leading to increased morbidity and healthcare utilization. Autologous SCB offers a potential strategy to promote hematologic recovery. We conducted a retrospective multi-institutional analysis across 18 centers comparing outcomes of SCB versus supportive care alone in RRMM patients who experienced prolonged cytopenias >30 days after CAR-T infusion. Methods: Patients were included if they received SCB within 1 year after commercial CAR-T between 6/2021 and 3/2024. To identify matched controls, we reviewed other CAR-T recipients from the same consortium and selected those with an absolute neutrophil count (ANC) of <0.72x109/L and platelet count of <21K at day 30 post-infusion-thresholds representing the 75th percentile of cytopenia severity in the SCB cohort. These criteria ensured control patients had comparable hematologic toxicity but without SCB. Hematologic recovery in the SCB cohort was assessed using the CIBMTR engraftment criteria. For the non-SCB (nSCB) group, cell counts were analyzed cross-sectionally at day 90 post-CAR-T and compared to the SCB cohort with Kruskal Wallis tests. Progression-free survival (PFS) and overall survival (OS) were evaluated using Kaplan Meier methods. Results: Of the 590 CAR-T recipients, 91 patients (15.4%) developed prolonged cytopenias; 39 patients received SCB and 52 did not. SCB patients more frequently had ISS stage III disease (23.1% vs. 11.5%), exrtamedullary disease (59% vs. 46.2%) and penta-refractory status (43.6% vs. 34.6%). The median number of prior lines of therapy was 6 in both cohorts. Thirteen patients (33.3%) in the SCB and 11 (21.2%) in the nSCB received high intensity infusional chemotherapy (DCEP/PACE/HyperCVAD). Pre-CAR-T cytopenias tended to be less frequent in the SCB vs. nSCB cohort: ANC <1000 in 5.1% vs. 13.5%, Hgb <8 in 23.1% vs. 21.2%, and platelets <50K in 28.2% vs. 40.4%. Median plasmacytosis in the bone marrow prior to CAR-T in the SCB group was 35% and 47.5% in the nSCB group. In the SCB cohort, 2 patients harbored DNMT3A mutations, one of whom had CCUS. Ide-cel was administered in 64.8% and cilta-cel in 35.2% of all patients. The rates and grades of cytokine release syndrome did not differ between groups. Six patients (15.4%) had HLH in the SCB group vs. none in the control group. All (100%) SCB patients received growth factor support; 58% also received TPO agonists. In the nSCB group, 43 patients (82.7%) received growth factor, and 16 patients (30.8%) received TPO agonists. Out of 24 pre-SCB bone marrow biopsies available, only 1 patient had relapsed MM while the remainder had hypocellular marrows. Median CD34+ cell dose was 2.93 million/kg (range: 1.76-23.55) administered at a median of 53 days post-CAR-T (range 24-265). All but one patient (97.4%) in the SCB group achieved hematologic recovery; median time to recovery was 24 days (range 9-87). No new toxicities attributable to SCB were observed. On day 90 post CAR-T infusion, SCB patients had higher median ANC (2.12 vs. 1.59x109/L, p=0.277), Hgb (10.6 vs. 8.7g/dL, p=0.002), and platelet counts (135 vs. 35K/L (p<0.001). Minimal residual disease negativity was achieved in 82.4% (SCB) vs. 83.3% (nSCB). Infections occured in 53.8% and 48.1% of SCB and nSCB cohorts. All serious infections occurred prior to SCB. There were 12 deaths in the SCB group (7 from disease progression, 2 from infection, 1 from neurotoxicity (NT), and 1 unknown) and 27 in the nSCB group (22 from progression, 3 from infection, 1 from HLH, 1 from NT). After a median follow-up period of 12.6 months in the SCB cohort and 11.6 months in the nSCB arm, the mPFS was 11.0 months and 8.2 months, respectively. The mPFS was 10.6 months (SCB) and 5.8 months (nSCB) (p=0.216), and 11.0 months (SCB) and 10.9 months (nSCB), for ide-cel and cilta-cel treated-patients, respectively. Median OS was not reached vs.12.3 months in the SCB vs. nSCB groups, respectively. Conclusion: Overall, SCB was safe and effective in promoting rapid hematologic recovery in nearly all patients. Compared to controls, SCB recipeints had significantly higher Hgb and platelet levels by day 90 post-CAR-T infusion. SCB was associated wih a near doubling of PFS in patients who received ide-cel, though interpretation is limited by non-randomized design and small sample size. Further study is needed to optimize paient selection, timing, and integration of SCB into CAR-T care.

Abstract Background Primary plasma cell leukemia (pPCL) is a rare and aggressive variant of multiple myeloma (MM), characterized by the presence of more than 5% circulating plasma cells on a peripheral blood smear. Despite advancements in therapy, including novel agents and autologous stem cell transplantation (ASCT), pPCL remains associated with markedly poorer survival outcomes compared to newly diagnosed MM. Although disease progression is common in pPCL, individual courses vary widely upon relapse and thereafter. We aim to describe the natural history of pPCL, focusing on clinical features and survival outcomes at the time of relapse or disease progression. Methods Patients with pPCL who were followed at our institution's cancer center between 2006-2024 were included for retrospective analysis. The electronic health record was reviewed to provide detailed information on patient demographics, disease characteristics, treatment regimens, and clinical outcomes. Time to next treatment (TTNT) and overall survival (OS) analysis were performed using the Kaplan-Meier method. Statistical analyses were performed using the SAS biostatistical software JMP 17.0.1 (SAS Institute Inc). Results We analyzed a cohort of 106 patients with relapsed pPCL after first-line therapy, all of whom received induction with novel agents. To date, 64 patients (60%) have died. Across the entire cohort, median OS is 39 months (mos) (95% CI: 25-68). 59 patients (56%) were diagnosed after 2015. Baseline FISH cytogenetics were available for 100 patients; 57 (57%) had one or more high-risk cytogenetic abnormality. Nearly all patients (96%) had ≥50% clonal plasma cells on initial bone marrow biopsy. While most patients (n = 61; 58%) were planned to undergo upfront ASCT, 20 (33%) experienced disease progression during induction therapy and required a more intensive salvage regimen (often infusional chemotherapy) prior to ASCT. The median TTNT after first-line therapy (TTNT#1) for the entire cohort was 10.4 mos (95% CI: 8–14). The median TTNT after second-line therapy (TTNT#2) for the entire cohort was 5.7 mos (95% CI: 4.3–7.1). Clinical details at first relapse, prompting initiation of second-line therapy, were available for 91 patients. Of these, 26 patients (29%) had symptomatic relapses—due to CRAB features or extramedullary disease—while 65 patients (71%) experienced asymptomatic biochemical progression. Symptomatic relapses were followed by significantly shorter TTNT#2 (3.9 vs. 7.0 mos; p = 0.039) and inferior OS from the start of second-line therapy (7.5 vs. 31.2 mos; p <0.0001) compared to asymptomatic relapses. The median TTNT#2 for patients who relapsed during induction and received salvage ASCT was 6.9 mos (n=20). In contrast, those treated after relapse with non-ASCT regimens (n = 86) fared worse; those receiving novel agent-based combinations, including doublets, triplets, or quadruplets (n=64) had respective median TTNT#2 of 3.3, 6.0, and 3.3 mos. Patients treated with infusional chemotherapy (n = 9) had the poorest outcomes, with a median TTNT#2 of just 1.4 mos. The median TTNT diminished with each subsequent line of therapy: TTNT#3 (n = 73) was 5.8 mos, TTNT#4 (n = 53) was 4.0 mos, and TTNT#5 (n = 35) was 2.6 mos. Patients with pPCL harboring the t(11;14) cytogenetic abnormality who were treated with venetoclax-based regimens (n = 8) during any line of therapy had a longer corresponding median TTNT of 15.0 mos. Additionally, 23 patients received anti-BCMA CAR-T cell therapy as a post-relapse line of therapy. At last follow-up, 15 of these patients (65%) had not relapsed, and after a median 20.7 mos of follow-up, the median TTNT following CAR-T cell therapy had not yet been reached. Patients who received CAR-T after 1-3 prior lines (n=15) had a median TTNT that was not reached; those who had received 4 or more prior lines (n=8) had a median TTNT of 20 mos. ConclusionsDespite the use of novel agent-based combination regimens and ASCT as first-line therapy, most patients with pPCL eventually experience disease progression. Consistent with MM, symptomatic relapse in pPCL is associated with poorer outcomes compared to asymptomatic biochemical progression. Additionally, the duration of response shortens with each successive line of therapy. Notable exceptions include venetoclax-based regimens in patients with t(11;14), and anti-BCMA CAR-T cell therapies: each appear to confer more durable responses than historical treatments in patients with relapsed pPCL.

Abstract Introduction Pronounced disparities in pediatric B-lymphoblastic leukemia (B-ALL) outcomes by race and ethnicity persist in the contemporary era. Gaps in overall survival (OS) from initial diagnosis are wider than event-free survival, suggesting that post-relapse mechanisms may contribute. However, data on treatment and outcomes after relapse are lacking. In a large, multicenter, real-world cohort of children and young adults with high-risk 1st relapse of B-ALL, we sought to (1) define survival outcomes by race/ethnicity and (2) explore if differences in therapy utilization drive outcomes. Methods We conducted an analysis within the Retrospective Study of Contemporary Approaches to 1st Relapse in B-ALL (ReCALL-1) cohort, which includes patients <30 years who initiated 1st relapse therapy between 2018-2022. This analysis was restricted to patients treated at US centers for “high-risk” 1st relapse, defined as relapse for which hematopoietic cell transplant (HCT) is generally standard-of-care: Early relapse (medullary <36m; isolated extramedullary <18m), late relapse with persistent end of reinduction disease, Ph+ B-ALL or age ≥18 years. Patients with Trisomy 21, those who received cell therapy before relapse and those who died during reinduction were excluded. The primary exposure was race/ethnicity categorized as Hispanic, non-Hispanic Black (NHB), non-Hispanic White (NHW), and other/unknown. The primary outcome was OS from relapse. Multivariable Cox regression models were constructed to include other social drivers of health (SDOH; health insurance, preferred language) and disease characteristics. Results Of 479 patients in ReCALL-1, 311 from 31 centers met inclusion for this analysis (41% Hispanic, 7% NHB, 46% NHW, 5% NH-other/unknown). With a median follow-up of 47m, OS differed by race/ethnicity (log-rank p=0.017): Hispanic, 4y OS 59%; NHB, 37%; NHW, 68%. Compared to NHW patients, NHB patients had an increased hazard of death (HR 2.3, 95% CI 1.3-4.3, p=0.006), while Hispanic patients were more similar to NHW patients (HR 1.3, 95% CI 0.9-2.0, p=0.16). Adjustment for SDOH only marginally attenuated the risk for NHB patients (aHR 2.0, 95% CI, 1.1-3.7, p=0.031) and led to equivalent outcomes for Hispanic patients (aHR 0.9, 95% CI, 0.5-1.5, p=0.66). Adjusting for disease characteristics that differed by race/ethnicity at p<0.2 (time to relapse, Ph+, Ph-like, KMT2Ar, age ≥18, or relapsed infant ALL) did not diminish the risk for NHB patients (aHR 2.0, 95% CI, 1.1-3.7, p=0.032) while outcomes remained equivalent for Hispanic patients (aHR 1.2, 95% CI, 0.8-1.8, p=0.40). Point estimates of cumulative incidence of non-relapse mortality (NRM) were highest in NHB patients (Hispanic, 4y NRM 13%, NHB 22%, NHW 11%), but this trend was not statistically significant (Gray’s p=0.52). Only 17% of patients initiated therapy for relapse on a clinical trial with no differences observed by race/ethnicity. Most patients received definitive treatment for relapse with either CAR T cell therapy (CART) or HCT: 79% overall, 74% Hispanic, 91% NHB, and 83% NHW. CART and HCT were utilized equally (39% and 40%, respectively) and the proportion that received each therapy did not vary by race/ethnicity. Time to CART/HCT, external referral for treatment, receipt of investigational v commercial CART, and use of post-CART consolidative HCT revealed no differences by race/ethnicity. Of the 66 patients (21%) who did not receive cell therapy, 26 had planned for CART or HCT, but did not receive due to disease progression or NRM (n=24; 12 Hispanic, 10 NHW), or insurance barriers (n=2, both Hispanic). After HCT, OS was worse for NHB patients (log-rank p=0.03) with 4y OS of 42% v 85% for NHW and 75% for Hispanic. Post-CART, 4y OS was 40% for NHB v 59% for NHW and 51% for Hispanic, but this difference was not statistically significant. Conclusion In a large real-world cohort of children with high-risk 1st relapse of B-ALL, we found a two-fold increased hazard of death for NHB patients compared to NHW patients, even after adjustment for other SDOH factors and high-risk disease features. In contrast to previous studies of adult patients, children appeared to have equal access to cell therapies across races/ethnicities. We found that Hispanic patients, who comprised >40% of the cohort, had similar survival to NHW children, suggesting that availability of cellular and immunotherapies may be mitigating historically observed disparities for children with relapsed B-ALL.

Abstract Background: Recent advances in computer vision have given rise to image foundation models (FMs), large-scale neural networks pre-trained on millions of unlabeled histopathology patches, which can learn generalizable representations of tissue morphology. In many cancers, these FMs have predicted genetic mutations and prognoses directly from histology, suggesting that morphology encodes “hidden” molecular information. In lymphoma, this field is nascent but growing.General-purpose histopathology FMs have limited representation of lymphoid morphology. We developed and evaluated LymphoVision, a lymphoma-specialized image foundation model pretrained with self-supervised learning, to extract clinically meaningful features from routine hematoxylin and eosin (H&E) histology. Methods: LymphoVision was pre-trained using knowledge distillation and consistency (DINOv2) on over 37 million multi-resolution patches (5x, 10x, 20x, 40x) sampled using a density-aware strategy from 31,211 H&E whole slide images (WSIs) from 9,155 archival lymphoma and reactive lymphoproliferative cases. To ensure diverse, representative samples, our patch sampling method uses an efficient graph-based clustering algorithm and prioritizes morphologically informative regions by clustering patch embeddings and performing density sampling within each cluster. The pretrained encoder was evaluated using a clustering-constrained attention multiple instance learning (MIL) framework on three diagnostic tasks: (1) cell-of-origin (COO) classification in diffuse large B-cell lymphoma (DLBCL), distinguishing germinal center B-cell (GCB) from non-GCB subtypes based on the Hans immunohistochemistry algorithm; (2) follicular lymphoma (FL) grading, differentiating low-grade (grades 1–2) from high-grade (grade 3); and (3) multiclass lymphoma subtype classification across six categories: DLBCL, FL, marginal zone lymphoma (MZL), mantle cell lymphoma (MCL), classical Hodgkin lymphoma (CHL), and benign/reactive lymphoid conditions. WSI-level labels were derived from pathology reports, and none of the evaluation cases were included in pretraining. Models used a ViT-Giant LymphoVision backbone to generate 20x patch embeddings, followed by MIL using 50% training data and 2-fold cross-validation (25% validation, 25% test). Performance was assessed using area under the receiver operating characteristic curve (AUC), accuracy, recall (sensitivity) and F1 score. Results: For DLBCL COO classification (total n=358 cases), LymphoVision achieved a mean test AUC of 0.93 (range 0.92–0.94), accuracy of 84.5% and F1 of 0.85 across two independent test folds (n=91 per fold, 50% GCB). These results outperformed general-purpose FMs on the same task and cases, including Virchow-v2 (AUC 0.85) and UNI-v2 (AUC 0.88), despite those models being pretrained on 10–100 times more images. For FL grading (total n=660 cases), the test set in each fold included 166 cases (18% grade 3), yielding a mean test AUC of 0.88 (range 0.87–0.90), accuracy of 89.2% and F1 of 0.80. In multiclass lymphoma subtype classification (total n=1,176 cases), the test set included 296 cases per fold. LymphoVision achieved a mean test AUC of 0.98 (range 0.97-0.99), mean test accuracy of 87.1% and mean F1 of 0.84. Class-wise recall was 92.0% for DLBCL (n=112), 88.2% for FL (n=72), 77.2% for MZL (n=29), 70.8% for MCL (n=35), 87.5% for CHL (n=24), and 95.8% for benign/reactive lymphoid conditions (n=24). Conclusion: LymphoVision is a lymphoma-specialized FM trained using self-supervision that achieves high diagnostic accuracy across multiple clinically meaningful tasks using weak supervision and standard H&E images, suggesting the encoder has learned fine-grained lymphoma morphology. Based on historical benchmarks from the 1997 International Lymphoma Study Group classification project, expert hematopathologists achieved diagnostic agreement rates of 55% to 84% using H&E morphology alone for the non-Hodgkin lymphoma (NHL) entities included in our multiclass task. LymphoVision approached or exceeded these benchmarks. These findings underscore the value of our approach using disease-specialized pretraining and intelligent patch sampling for data-efficient histopathology learning. LymphoVision has the potential to augment diagnostics, enable scalable AI-powered classification in real-world workflows, and serve as a foundation for prognostic and theragnostic applications. External validation and multimodal integration are ongoing.

Abstract Introduction: Leukemogenesis is driven by nucleophosmin 1 mutations(NPM1-m) or lysine methyltransferase 2A rearrangements(KMT2A-r) in ~35–40% of acute myeloid leukemia (AML) cases. Nearly half of AML patients will develop relapsed/refractory (R/R) disease within a year, with <20% expected response rate following venetoclax/azacitidine (Ven/Aza) and progressively poorer outcomes with each subsequent line of therapy. Ziftomenib–a potent, highly selective, oral, investigational menin inhibitor–has demonstrated clinical activity as both monotherapy and in combination for adults with R/R NPM1-m or KMT2A-r AML. KOMET-007 (NCT05735184) is an ongoing, open-label, ph1a/b study of ziftomenib in combination with standard chemotherapies in adults with newly diagnosed and R/R NPM1-m or KMT2A-r AML. Here we present updated safety and clinical activity in patients with R/R AML treated with the recommended ph2 dose (RP2D) of ziftomenib 600 mg in combination with Ven/Aza across ph1a/b. Methods: Adults (≥18y) with R/R NPM1-m or KMT2A-r AML enrolled independently into separate arms; patients in dose escalation (ph1a) and expansion (ph1b) were treated with oral ziftomenib 600 mg once daily (QD; continuously from C1D8 onward) plus standard doses of Ven/Aza. Primary endpoints: AEs, complete remission (CR; ELN 2022), dose limiting toxicities (DLT; ph1a only). Key secondary endpoints: composite CR (CRc; CR with full, partial or incomplete hematologic recovery), overall response, duration of response. Results: As of June 25, 2025, 80 patients (51 NPM1-m, 29 KMT2A-r) with R/R AML were enrolled (20 ph1a; 60 ph1b) and treated with ziftomenib 600 mg QD + Ven/Aza. Median age was 63y (range 19–85), 50% were female, 74% had ECOG PS 0–1; 22 (28%) had FLT3 co-mutations and 7 (9%) had IDH co-mutations. Median number of prior therapies was 1 (range 1–6); 18% (14/80) had prior stem cell transplant; 25 (49%) NPM1-m and 19 (66%) KMT2A-r patients had prior Ven exposure. 71 patients (89%) had Gr ≥3 treatment-emergent AEs, most commonly (≥20% of patients) febrile neutropenia (31%), decreased platelet count (28%), decreased white blood cell count (26%) and decreased neutrophil count (23%). Gr ≥3 ziftomenib-related AEs occurred in 30 patients (38%), most commonly febrile neutropenia (9%) and anemia (8%). 6% of patients discontinued ziftomenib treatment due to AEs. Differentiation syndrome (DS) occurred in 1 (1%) NPM1-m patient (Gr 3), which lasted 1d and successfully resolved with protocol-specified DS mitigation. No ziftomenib-related QTc prolongation was reported with the combination, and no DLTs were observed in ph1a. 70 patients (43 NPM1-m; 27 KMT2A-r) were response-evaluable (≥1 response assessment or death). Median follow-up was 18.0 wks for NPM1-m and 16.4 wks for KMT2A-r. Overall response rates (ORRs) were 65% (28/43) for NPM1-m and 33% (9/27) for KMT2A-r. CRc rates were 49% (21/43) for NPM1-m and 22% (6/27) for KMT2A-r after median time to first CRc of 4.9 wks (range 2.7–15.6) and 5.5 wks (range 2.6–18.9), respectively; measurable residual disease (MRD)-negativity rates (local) among tested CRc responders were 50% (9/18) for NPM1-m and 60% (3/5) for KMT2A-r after median time to first MRD-negativity of 5.9 wks (range 2.9–15.6) and 8.1 wks (range 7.7–18.9), respectively. During continuous ziftomenib administration in patients who achieved CRc, median time to neutrophil recovery (≥1×109/L) was 43d overall (43d for NPM1-m; 71d for KMT2A-r), and median time to platelet recovery (≥100×109/L) was 27d overall (27d for NPM1-m; 23d for KMT2A-r). As of the data cutoff, overall median duration of CRc was not estimable; median overall survival was not estimable for NPM1-m and 21.1 wks for KMT2A-r. In Ven-naive patients, CRc rates were 71% (15/21) for NPM1-m and 33% (3/9) for KMT2A-r; CR rates were 48% (10/21) and 11% (1/9); and ORRs were 81% (17/21) and 56% (5/9), respectively. The study is ongoing with 44% (19/43) of NPM1-m and 11% (3/27) of KMT2A-r patients still on treatment, with responses continuing to evolve.Updated data to be presented.Conclusions: In the ongoing KOMET-007 study, ziftomenib RP2D of 600 mg QD + Ven/Aza was well tolerated with robust clinical activity in patients with R/R NPM1-mor KMT2A-rAML. No ziftomenib-related QTc prolongation was reported. One case of DS (NPM1-m, Gr 3) successfully resolved with protocol-specified mitigation. These data support further investigation of ziftomenib-based combinations in R/R NPM1-mand KMT2A-rAML.

Abstract Introduction: Nucleophosmin 1 mutations (NPM1-m) drive leukemogenesis in approximately 30% of acute myeloid leukemia (AML) cases. Ziftomenib is a potent, highly selective, oral, investigational menin inhibitor that has demonstrated clinical activity as both monotherapy and in combination for adults with NPM1-m or lysine methyltransferase 2A-rearranged (KMT2A-r) AML. KOMET-007 (NCT05735184) is an ongoing, open-label, dose-escalation (phase 1a) and expansion (phase 1b) study of ziftomenib in combination with standard chemotherapies in NPM1-m or KMT2A-r AML. Here we present phase 1b safety and clinical activity in patients with newly diagnosed NPM1-m AML who received the recommended phase 2 dose (RP2D) of ziftomenib 600 mg in combination with venetoclax and azacitidine (Ven/Aza). Methods: Adults (≥18 years) with newly diagnosed NPM1-m AML who were not eligible for intensive chemotherapy were enrolled in phase 1b and received ziftomenib 600 mg once daily plus standard doses of Ven/Aza. Ziftomenib was orally administered once daily continuously from Cycle 1 Day 8 onward. Phase 1b primary endpoints were adverse events (AEs) and complete remission (CR) per ELN 2022. Secondary endpoints included composite CR (CRc; defined as CR with full, partial, or incomplete hematologic recovery), overall response, and duration of response. Local measurable residual disease (MRD) was assessed by next-generation sequencing, RT-qPCR, and multiparameter flow cytometry. Results: As of June 25, 2025, 39 patients with newly diagnosed NPM1-m AML were enrolled and treated with ziftomenib 600 mg once daily plus Ven/Aza in phase 1b. Median age was 75 years (range 53–93), 54% were female, 46% had ECOG PS 0–1; 12 (31%) had FLT3 co-mutations and 9 (23%) had IDH co-mutations. Median duration of follow-up was 16.3 weeks (range 1.6–41.1). Twenty-nine patients (74%) had grade ≥3 treatment-emergent AEs, most commonly (≥20% of patients) decreased neutrophil count (31%) and decreased platelet count (23%). Grade ≥3 ziftomenib-related AEs occurred in 14 patients (36%), most commonly decreased neutrophil count (15%) and decreased platelet count (13%). No ziftomenib treatment discontinuations occurred due to AEs. Differentiation syndrome occurred in 1 (3%) patient (grade 2), which successfully resolved with protocol-specified mitigation. One patient (3%) had investigator-assessed ziftomenib-associated QTc prolongation (grade 3); however, there were concomitant significant electrolyte abnormalities and event resolved with electrolyte repletion. At time of data cutoff, 31 NPM1-m patients were response-evaluable (≥1 response assessment or death). CRc rate was 84% (26/31) after a median time to first CRc of 3.5 weeks (range 2.4–9.4), with local MRD-negativity rates among tested CRc responders of 54% (13/24) after a median time to first MRD-negativity of 8.4 weeks (range 2.9–17.4). CR rate was 58% (18/31) and overall response rate was 94% (29/31), with responses and MRD-negativity rates continuing to evolve. During continuous ziftomenib administration in patients who achieved CRc, median time to neutrophil recovery (≥1×109/L) was 36 days, and median time to platelet recovery (≥100×109/L) was 28 days. The median duration of response and median overall survival were not reached as of the data cutoff, at a median follow-up of 16.3 weeks. The study is ongoing, with 81% (25/31) of NPM1-m patients still on-study and 77% (24/31) still on ziftomenib treatment. Updated safety and clinical activity results will be presented.Conclusions: In the ongoing KOMET-007 study, ziftomenib RP2D of 600 mg once daily combined with Ven/Aza was well tolerated and demonstrated robust clinical activity in patients with newly diagnosed NPM1-m AML, including 84% CRc after a median of 3.5 weeks and 54% CRc MRD-negativity after a median of 8.4 weeks. Low rates of ziftomenib-related cytopenia and no additional myelosuppression were observed with this combination. One case each of differentiation syndrome (grade 2) and investigator-assessed QTc (grade 3) were successfully resolved. Taken together, these data support the RP2D determination and advancement of this ziftomenib-based combination in the KOMET-017 (NCT07007312) randomized phase 3 study in patients with newly diagnosed NPM1-m AML.

Abstract Introduction Nadir bone marrow biopsies (BM Bx) are routinely performed around day 14 to assess early response to AML intensive induction therapy. The presence of residual disease prompts initiation of reinduction therapy, often with another cycle of intensive chemotherapy. Whether there are characteristics of nadir BM Bx that predict for response to intensive reinduction is unknown. To evaluate this, we analyzed data from ECOG-ACRIN clinical trials in AML of intensive induction chemotherapy with nadir biopsy information available to determine predictors of outcomes after reinduction. Methods We included patients from ECOG-ACRIN E1900 (Fernandez NEJM 2009) and arm A of E2906 (Foran ASH 2015) who received induction therapy with daunorubicin and cytarabine (7+3), had residual disease (blasts ≥5% or cellularity ≥20%) at nadir BM Bx and received reinduction chemotherapy. Nadir BM Bx characteristics that were evaluated included blast percentage and cellularity as well as both absolute and relative (percentage change, e.g. 20% to 10% = -50%) differences calculated between diagnosis and nadir for each parameter. Complete remission (CR) after reinduction therapy, disease free survival (DFS) and overall survival (OS) were outcomes for all analyses. Logistic regression models were used to evaluate associations with CR, and Cox proportional hazards models were used for DFS and OS. For models evaluating changes in blast percentage or cellularity as predictors, baseline blast percentage or cellularity was included as a covariate, respectively. Results Of 1013 patients identified, 750 (74%) had a nadir biopsy, 233 (23%) had evidence of residual disease, received reinduction therapy, and were included in the primary analysis. The median age was 58 and 42% were female. Cytogenetic risk category (by ELN2017) was adverse in 31%, intermediate in 66% and favorable in 3%. The median nadir BM Bx blast was 30% (range 0 to +97%) and cellularity was 20% (0 to +100%). The median absolute difference in blast percentage between baseline and nadir BM Bx was -15% (-95% to +66%) with a median relative difference of -36% (-100% to +400%). After reinduction, 118 patients (52%) achieved a CR. Neither absolute difference in blast percentage (OR 0.99, 95%CI 0.98-1.00, p=0.20) nor relative difference (OR 1.00, 95%CI 0.99-1.00, p=0.33) predicted CR. There was similarly no association with DFS (HR = 1.00, 95%CI: 0.99-1.01, p=0.65) or OS (HR = 1.00, 95%CI 0.99-1.01, p=0.76). Blast reduction ≥50% similarly did not predict these outcomes. We also evaluated the impact of any reduction in blast percentage (responders) from baseline to nadir and again found no association with CR (OR 1.65, 95%CI 0.90-3.03, p=0.11). Similarly, there was no association with DFS (HR 0.74, 95%CI 0.51-1.09, p=0.13) or OS (HR 0.81, 95%CI 0.58-1.13, p=0.21) in responders vs nonresponders. No association with CR, DFS or OS were identified when performing the same analysis in intermediate and adverse cytogenetic subgroups. We found no association with CR when nadir cellularity was assessed as a continuous variable (OR 0.99, 95%CI 0.98-1.00, p=0.10) or by relative difference (OR 0.99, 95%CI 0.99-1.00, p=0.10). There was a trend towards association with absolute reduction in cellularity from baseline to nadir and CR (OR 0.99, 95%CI 0.98-1.00, p=0.06). No association was identified between relative difference in cellularity and DFS and OS, nor was there an association among cytogenetic risk subgroups with any outcome. Discussion In intensively treated patients with residual disease at nadir BM Bx (i.e. day 14) who received intensive reinduction, the likelihood of achieving a CR was not influenced by the magnitude of disease reduction nor the extent of disease at the time of nadir biopsy. Similarly, there were no nadir biopsy features that predicted DFS or OS after intensive reinduction therapy. Given alternative reinduction therapies such as hypomethylating agents and venetoclax are available, these results suggest that the decision to attempt a second intensive induction cycle following the unequivocal presence of residual disease should not be influenced by characteristics of the nadir BM Bx. Although patient-specific factors such as patient fitness may favor a switch to less intensive options, these results support consideration of intensive reinduction therapy in patients with residual disease at nadir biopsy, even without changes in blasts or cellularity.

Abstract Introduction: NAVD has become a standard-of-care for frontline (1L) treatment (tx) of advanced stage cHL, based on the results of the S1826 study. S1826 demonstrated a 1-year progression free survival (PFS) of 94% for NAVD compared to 86% for brentuximab vedotin-AVD, which remained durable at 2 years (Herrera et al, NEJM 2024). Adverse events (AEs) and efficacy have not been reported in patients (pts) treated with NAVD outside of a clinical trial. Hence, we sought to evaluate safety profile and efficacy outcomes of pts with cHL treated with 1L NAVD in a real-world setting (RWS). Methods: In this multicenter retrospective cohort study, we included pts ≥18 years with cHL treated with 1L standard of care NAVD between 2023 and 2025 at 20 US academic sites. The safety evaluable cohort received at least 1 cycle of NAVD and included pts who have not completed tx. The efficacy evaluable cohort had completed 6 cycles of tx including end of treatment (EOT) PET-CT and included pts who intended to complete tx but died or discontinued due to progressive disease or AEs. The primary endpoint was EOT overall response rate (ORR). Secondary endpoints included EOT complete response rate (CRR), PFS and overall survival (OS). PFS and OS were calculated from the start date of tx and analyzed using the Kaplan Meier method. Results: A total of 311 pts with cHL were treated with NAVD with 26% of pts with tx ongoing (n=83) and 73% with tx completed (n=228). Baseline demographics included 55% male pts, 60% with stage IV disease (stage IIB, n=4) and 84% with nodular sclerosis subtype. Median age was 37 years (range 18-88), 29% had bulky disease (>7.5 cm), 28% were EBV positive (n=74/268), 8% had history of an autoimmune disease, and 31% had an IPS 4-7. G-CSF was given in 64% of pts. All patients were safety evaluable. Tx was interrupted (any length) in 24%, reduced in 12%, and discontinued in 14%. The most common reason for dose reduction was neuropathy in 4%. Nivolumab was discontinued in 10% and all therapy in 3.7% due to toxicity. Any AE and any grade 3 or higher AE (>G3) were observed in 92% and 62%, respectively. Neutropenia of any grade occurred in 68% (48% >G3), infection in 25% (10% >G3), and neuropathy in 30% (2.8% >G3). Hospitalization for an AE occurred in 20% (4.5% febrile neutropenia). Among pts with complete data (n=305), any immune related AE (irAE) occurred in 78 pts (25.6%) with any >G3 irAE in 23 pts (7.5%). The most common irAEs included hypothyroidism (7.9%, all grade 1-2), hepatotoxicity (1.6% G1/G2, 3.3% >G3), inflammatory arthritis (3% G1/G2, 0.6% >G3), rash (2.6%, all grade 1-2), and colitis (1% G1/G2, 1% >G3). Other notable >G3 irAEs included adrenal insufficiency (n=2), AKI/ATN (n=1), myocarditis (n=1), diabetes (n=1) and pancreatitis (n=1). Of 84 irAE events, 47 (56%) have had complete resolution. Among all pts, 14.4% (n=45) required systemic steroids at any time (intravenous in 4.1%), 3.2% (n=10) require ongoing steroids/immunosuppression and 8.3% other therapy for an irAE (n=23 levothyroxine, n=1 hemodialysis, n=1 heart failure tx, n=1 insulin). Among all pts, 265 (85%) had an interim PET-CT after 2-4 cycles. The ORR and CRR at interim PET-CT were 97% and 78%, respectively. In the efficacy evaluable population (n=228), the EOT ORR and CRR were 96% and 90%. The median follow up was 10.2 mo (range 0.27-28.52). The 1-year PFS and OS estimates were 93.8% (95% CI: 90.8-98.4%) and 99.1% (95% CI: 97.9-100%), respectively. In multivariable analysis (MVA), only IPS score (OR 0.6, 95%CI: 0.59-0.60) was associated with lower CRR. In Cox univariable analysis, EBV+ status (HR=14.8, 95% CI: 3.2-69.2), IPS score (HR=1.5, 95% CI: 1.0-2.2), and ESR (HR=1.02, 95% CI: 1.00-1.04) were associated with worse PFS. Among the 11 pts with progression/relapse, 9 were primary refractory. 10 pts were started on second line (2L) therapy, and all pts are alive (1 pt pending 2L tx initiation). There were 2 deaths: one related to infectious complications and other unknown and both occurred after the first cycle of therapy. Conclusions: Pts treated with NAVD in a large RWS showed comparable response rates and 1-year survival outcomes to those in the pivotal S1826 trial. Most AEs were similar and grade 1-2, but close monitoring of irAEs is necessary given the potential long-term toxicity of checkpoint inhibitors and potential for ≥G3 irAEs (7.5%) in this young patient population. Our data supports the use of NAVD for 1L advanced stage cHL.

Abstract Background Accurate reconstruction of electronic health record (EHR) treatment sequences is essential for leveraging real-world data (RWD) in clinical research and trial-eligibility platforms in multiple myeloma (MM). However, inconsistencies in RWD treatment documentation and the complexity of line-of-treatment (LOT) assignments in MM hinder reproducibility and limit comparisons against published trials. To address these limitations, we developed HEAL-MM, an automated rule-based algorithm that systematically parses structured medication data to calculate LOTs and segment treatment into standardized, clinically meaningful phases. The algorithm is designed to be interpretable, adaptable to evolving therapeutic frameworks, and compatible with widely adopted EHR data standards. Here, we report its real-world validation against expert-adjudicated treatment sequences to assess concordance, identify edge-case failure points, and support RWD integration. Methods A stratified cohort of 100 MM patients (9,136 structured treatment events) was curated from 114 unique healthcare systems compiled from the multi-institutional HealthTree Foundation harmonized registry. Cohort selection was randomized to ensure coverage across diagnostic eras (pre-2010, 2015–2020, 2022–2025) and treatment exposures, including CAR T-cell therapies, bispecific antibodies, and anti-CD38-based regimens. HEAL-MM sequenced MM regimens chronologically and applied predefined logic rules to identify LOT transitions, the initiation of a new therapeutic course based on drug changes, treatment gaps, and procedural anchors such as stem cell transplant or CAR T. Within each LOT, the algorithm segments and labels therapy into phases, clinically distinct intervals (e.g., induction and maintenance) using rule-based clinical logic. Outputs were benchmarked against expert-labeled adjudications. Concordance was evaluated using concordance index (C-index), stratified by patient, LOT depth and diagnostic era. Results: HEAL-MM demonstrated high concordance with expert-labeled treatment sequences: 0.94 for LOT identification, 0.98 for treatment phase segmentation, and 0.96 for phase label assignment. Achieving complete agreement with all expert annotations in 90% of patients for LOT assignment, 94% for the number of treatment phases identified, 77% for the specific labels applied to each phase, and 75% across all three dimensions simultaneously. Among discordant patients, the absolute difference in final LOT count was 0.5 ± 0.9. Performance was highest in early lines (LOTs 1–3; C-index: 0.97), with seven patients showing discrepancies at this stage. Concordance declined in later lines (LOT ≥4; C-index: 0.87), where 50% of the discordant cases (n=4) were attributable to misclassifications in earlier LOTs, such as medication restarts following >3-month gaps, demonstrating the compounding effect of early errors on longitudinal accuracy. Temporal stratification revealed improved accuracy in post-2020 cases (C-index: 0.94) compared to 2015–2020 cohorts (0.91), reflecting the increasing adoption of structured treatment standards and EHR data quality improvements. To assess HEAL-MM performance at clinically significant junctures, we analyzed treatment transitions that initiated a new LOT or marked a change in phase, achieving C-indices of 0.97 for LOT transitions, 0.97 for treatment phase shifts, and 0.94 for phase label assignment. Conclusions: HEAL-MM provides a validated, scalable solution for standardizing MM treatment structures in RWD environments, as it accurately detects the exact treatment entry of a new LOT 97% of the time. This rule-based framework enables reproducible LOT and phase assignment across heterogeneous EHR systems, supporting real-time cohort identification across institutions and EHR platforms. The algorithm's modular design allows for continuous integration of novel therapeutic categories and evolving treatment logic. While the algorithm maintained high fidelity overall, early-stage misclassifications consistent with increased heterogeneity and documentation ambiguity occasionally propagated downstream, underscoring the compounding effect of upstream errors in longitudinal reconstruction. With its generalizable structure and compatibility with industry-standard data models, HEAL-MM lays the foundation for structured treatment reconstruction in MM across RWD infrastructure and serves as a transferable blueprint for other hematologic malignancies.