Claire Harrison

Background: Abnormal JAK and TGF-β superfamily signaling in myelofibrosis (MF) lead to ineffective hematopoiesis (IH), splenomegaly, constitutional symptoms and cytopenia. Anemia is prevalent and is associated with reduced quality of life and survival. JAK inhibitors improve spleen size and symptoms but are associated with dose-limiting cytopenias, exacerbating those caused by IH. Elritercept is an investigational modified activin receptor type IIA ligand trap designed to inhibit activin A and select TGF-β superfamily ligands (activin B, GDFs 8 & 11) to restore IH. Updated results from the ongoing Phase 2 RESTORE trial (NCT05037760) evaluating elritercept in MF are presented, including initial exploration of changes in iron homeostasis markers and specific symptoms. Methods: Data are presented as of 03Apr24, evaluating elritercept as monotherapy (Arm A) and in combination with ruxolitinib (Arm B) in participants with MF and anemia (transfusion-dependent [TD; ≥6 RBC U/12 wks and ≥1 transfusion within 4 wks before elritercept] or non-TD and hemoglobin [Hgb] <10g/dL). Endpoints include safety, PK/PD, and effects on anemia, spleen size, and symptoms (MPN-SAF-TSS). Baseline and safety data are for participants who received ≥1 dose of elritercept. Anemia, spleen, and total symptom score (TSS) data are for the respective evaluable populations. The recommended Part 2 dose (RP2D) of 3.75 - 5mg/kg SC Q4W was selected, consistent with a parallel Phase 2 study in myelodysplastic neoplasms (MDS), based on the totality of data including safety and PK/PD. Elritercept exposure at the RP2D in MDS and at doses ≥3 mg/kg in RESTORE was similar. Results: Overall, 23 and 31 participants were enrolled in Arm A and Arm B, respectively, with median treatment duration of 30 wks (Part 1 Dose Escalation: 41 wks, Part 2 Dose Expansion: 10 wks). At baseline, 31% of all participants were TD and most had splenomegaly (volume ≥450 cm3) and symptoms (TSS ≥10). Ferritin and hepcidin levels were elevated, particularly in TD participants, and 59% of all participants had thrombocytopenia (<150x109/L). Treatment-emergent adverse events observed in ≥15% of participants) were thrombocytopenia (19%) and diarrhea (17%). One participant had an asymptomatic Hgb increase (10.2 to 12.1 g/dL) requiring dose reduction, per protocol, that was deemed a dose-limiting toxicity. A maximum mean increase in Hgb ≥1.0g/dL over 12 consecutive wks during the first 24 wks was observed in 3/8 evaluable non-TD participants in Arm A and 7/12 non-TD participants in Arm B; among participants treated at 3mg/kg or higher in Arm B, 6/10 had a mean increase in Hgb ≥1.0g/dL. Participants with a maximum mean Hgb increase ≥1.0g/dL generally had decreases in hepcidin and ferritin. Platelets were generally maintained or increased, including in participants with baseline thrombocytopenia. Reduction in 12-wk transfusion burden was observed in 3/5 and 6/10 TD participants in Arms A and B respectively over the first 24 wks. In an expanded population with ≥3 RBC U/12 wks at baseline (TD3), 2/12 participants in Arm A achieved ≥ 12 wks transfusion independence (TI), both at doses < 3 mg/kg. In Arm B, 10/21 TD3 participants had a ≥50% reduction in RBC U/12wks including 6 participants with TI. Among Arm B TD3 participants treated at doses ≥ 3 mg/kg, 8/11 had ≥50% reduction in RBC U/12 wks and 5/11 achieved TI. Overall, 9/17 evaluable participants showed reduction in spleen size at Wk 24; 3 participants had ≥35% reduction (1 in Arm A, 2 in Arm B). Improvements in TSS were observed in 13/20 evaluable participants at Wk 24 and 3 participants (2 in Arm A, 1 in Arm B) had ≥50% reduction (improvement). Among participants with ≥50% reduction in TSS, dominant items were itching, abdominal discomfort, pain under the left rib, and satiety in Arm A and fatigue, night sweats, and abdominal discomfort in Arm B. Summary: Elritercept was generally well tolerated with potential to treat multiple aspects of MF. Observed improvements in Hgb and transfusion burden with maintenance or increases of platelets demonstrate potential to address IH and treat MF and ruxolitnib-associated cytopenias. Decreases in ferritin and hepcidin support potential improvements in iron homeostasis and/or inflammation. Data also support potential for elritercept to reduce spleen size and improve TSS as monotherapy and in combination with ruxolitinib. Updated results with additional patients and longer-term data will be presented.

• The addition of navitoclax to ongoing ruxolitinib demonstrates durable responses and evidence of potential disease modification in R/R MF • Thrombocytopenia is the most common adverse event, and is manageable and reversible with dose reduction as necessary ABSTRACT Navitoclax, an oral BCL-X L /BCL-2 inhibitor, induces apoptosis of malignant cells in myelofibrosis (MF). Here, we present results of pooled Cohort 1 of the phase II REFINE trial (NCT03222609), which evaluated navitoclax plus ruxolitinib (NAV+RUX) in patients with relapsed/refractory MF with suboptimal response to ruxolitinib (on stable dose of ≥10 mg twice daily for ≥12 weeks [Cohort 1a] or ≥24 weeks [Cohort 1b]). Patients in Cohort 1a received add-on navitoclax 50 mg/day, with escalation to ≤300 mg if platelet count was ≥75 x 10 9 /L. Patients in Cohort 1b received navitoclax 100/200 mg/day if platelet count was ≤150 or >150 x 10 9 /L, respectively. Primary endpoint was spleen volume reduction ≥35% (SVR 35 ) at Week (W) 24. Secondary endpoints included ≥50% reduction in total symptoms score (TSS 50 ) at W24, changes in bone marrow fibrosis (BMF) grade, anemia response, and safety. In total, 125 patients received ≥1 dose of NAV+RUX. With median follow-up of 21 months, SVR 35 rate was 23% at W24 and 39% at any time on study (median duration: 11 months). TSS 50 rate was 24% at W24 and 46% at any time on study. BMF improved by ≥1 grade, any time on study, in 39% of patients. Anemia responses were achieved in 23% of patients. Median overall and progression-free survival were 52.3 and 22.1 months, respectively. No new safety signals were observed. The most common adverse event was thrombocytopenia (86%) without clinically significant bleeding. NAV+RUX had tolerable safety, and resulted in early improvement in parameters of disease modification in this difficult-to-treat population.

6502 Background: Pelabresib (PELA) is an oral, small-molecule, investigational BET inhibitor that aims to decrease expression of genes involved in MF. MANIFEST-2 (NCT04603495), a global, randomized, double-blind, Phase 3 study, investigated the efficacy and safety of PELA + ruxolitinib (PELA+RUX) vs placebo + RUX (PBO+RUX) in JAKi treatment-naïve patients (pts) with MF. Methods: Eligible pts had DIPSS score ≥ INT-1, platelet count ≥100 × 10 9 /L, spleen volume ≥450 cm 3 , ≥2 symptoms with an average score ≥3 or total symptom score (TSS) ≥10 (MFSAF v4.0), peripheral blast count <5%, and ECOG PS ≤2. Pts were randomized 1:1. PELA or PBO was administered (QD for 14 consecutive days of 21) with RUX (BID for 21 days [1 cycle]). Primary endpoint was ≥35% spleen volume reduction from baseline (BL) (SVR35) at Week (Wk) 24. Secondary endpoints included absolute change in TSS and ≥50% reduction in TSS from BL (TSS50) at Wk 24, and safety. Other endpoints included hemoglobin (Hb) response (≥1.5 g/dL mean increase from BL without transfusions in the prior 12 wks), RBC transfusion number and bone marrow fibrosis (BMF). Results: As of Aug 31, 2023, 430 pts were randomized. At Wk 24, 65.9% (141/214) vs 35.2% (76/216) (p<0.001) of pts had an SVR35 response in the PELA+RUX vs PBO+RUX arms, respectively. SVR35 responders at any time were 80.4% (172/214) vs 50.0% (108/216); 80% (137/172) vs 63% (68/108) of responders reached SVR35 at Wk 12 scan; 83.7% (144/172) vs 79.6% (86/108) maintained response at cutoff. Mean change in absolute TSS was -15.99 (SE 1.028) vs -14.05 (SE 0.986) (p=0.0545), and TSS50 was 52.3% (112/214) vs 46.3% (100/216) (p=0.216) at Wk 24. There was a 2-fold difference in pts with both SVR35 and TSS50 with PELA+RUX (40.2% [86/214]) vs PBO+RUX (18.5% [40/216]). Hb response occurred in 10.7% (23/214) vs 6.0% (13/216) of pts, with differences in mean Hb levels maintained at 48 wks. In pts with anemia (Hb BL <10 g/dL), Hb response occurred in 16.4% (11/67) vs 14.1% (10/71). A total of 30.8% (66/214) vs 39.8% (86/216) of required RBC transfusion during the first 24 wks. BMF improvement ≥1 grade occurred in 38.5% (40/104) vs 24.2% (24/99) of pts (odds ratio 2.09; p=0.019). Of 426 pts evaluated for safety, the most common treatment-emergent AEs (≥20%) in the PELA+RUX vs PBO+RUX arms were anemia (43.9% vs 55.6% [Grade ≥3, 23.1% vs 36.4%]), thrombocytopenia (32.1% vs 23.4% [9% vs 5.6%]), platelet count decreased (20.8% vs 15.9% [4.2% vs 0.9%]), and diarrhea (23.1% vs 18.7% [0.5% vs 1.4%]). Updated results will be presented at the congress. Conclusions: PELA+RUX significantly and durably reduced splenomegaly, with a trend toward reduced TSS, and improved anemia and BMF at Wk 24 compared with PBO+RUX in JAKi treatment-naïve pts with MF, addressing key hallmarks of MF. Resultssupport a potential paradigm shift to combination therapy for MF. CH and JM contributed equally. Clinical trial information: NCT04603495 .

Background Myelofibrosis (MF) is characterized by splenomegaly, MF-associated symptoms, cytopenias (eg, anemia), and impairment of the bone marrow (BM) microenvironment (including fibrosis). Pelabresib (CPI-0610; PELA) is an investigational, oral, small molecule drug that inhibits BET proteins and subsequent BET-mediated gene expression involved in MF pathogenesis. The Phase 3 MANIFEST-2 study (NCT04603495) met its primary endpoint, showing a statistically significant higher proportion of patients (pts) with ≥35% reduction in spleen volume (SVR35) from baseline (BL) at Week (Wk) 24 with PELA+ruxolitinib (RUX) vs placebo (PBO)+RUX (p<0.001) in Janus kinase inhibitor-naïve pts with MF. PELA+RUX also showed a trend toward improved total symptom score (TSS) from BL at Wk 24, as well as improvements in multiple measures of anemia (eg, hemoglobin [Hb] response) and in the BM microenvironment, vs PBO+RUX (Rampal R, et al. Presented at ASH 2023 [Oral 628]). Aim To present follow-up efficacy and safety outcomes at Wk 48 from the Phase 3 MANIFEST-2 study. Methods Eligible pts had a DIPSS ≥intermediate-1 risk, platelet count ≥100 × 109/L, spleen volume ≥450 cm3, ≥2 symptoms with an average score ≥3 or TSS ≥10 by MF Symptom Assessment Form v4.0, peripheral blast count <5%, and ECOG PS ≤2. Pts were randomized 1:1. PELA or PBO was administered once daily for 14 consecutive days of 21-day cycles in combination with RUX, which was administered twice daily for 21-day cycles. Primary endpoint was SVR35 at Wk 24. Key secondary endpoints were absolute change in TSS and ≥50% reduction in TSS from BL (TSS50) at Wk 24. Other prespecified endpoints included SVR35 at Wk 48, absolute change in TSS and TSS50 at Wk 48, Hb response (≥1.5 g/dL mean increase from BL without transfusions in the prior 12 wks), BM fibrosis (BMF), and safety. Mutation profiles were assessed by next-generation sequencing. Informed consent was obtained from all pts. Results As of March 29, 2024, all pts assessed had been followed for ≥48 wks; 58.9% (126/214) and 62.0% (134/216) of pts continued on double-blind treatment in the PELA+RUX and PBO+RUX arms, respectively. Preliminary data are presented here; analyses are ongoing and data points are subject to change. At Wk 48, 56.5% (121/214) vs 37.5% (81/216) of pts had SVR35 responses in the PELA+RUX vs PBO+RUX arms, showing sustained benefit beyond Wk 24. SVR35 responders at any time were 82.2% (176/214) vs 58.3% (126/216) in the PELA+RUX vs PBO+RUX arm; loss of SVR35 response was observed in 13.1% (23/176) vs 19.8% (25/126) of pts in the PELA+RUX vs PBO+RUX arm. At Wk 48, least squares mean (standard error) absolute change in TSS was −16.24 (1.133) vs −14.11 (1.085) in the PELA+RUX vs PBO+RUX arms, with a greater difference in absolute TSS observed between treatment arms at Wk 24 and Wk 48 for pts with higher symptom burden at BL. TSS50 response at Wk 48 was 45.3% (97/214) vs 39.4% (85/216) with PELA+RUX vs PBO+RUX. Dual SVR35 and TSS50 response was observed in 36.0% (77/214) vs 19.0% (41/216) of pts in the PELA+RUX vs PBO+RUX arms at Wk 48. Hb response was observed in 12.6% (27/214; 95% confidence interval [CI], 8.17-17.07) vs 6.9% (15/216; 95% CI, 3.55-10.33) of pts in the PELA+RUX vs PBO+RUX arms, with differences between arms in mean Hb levels maintained at Wk 48; in pts with anemia (Hb BL <10 g/dL), Hb response was observed in 17.9% (12/67; 95% CI, 8.73-27.09) vs 14.1% (10/71; 95% CI, 5.99-22.18) of pts. The trend of higher rate of BMF improvement of ≥1 grade in the PELA+RUX arm vs the PBO+RUX arm continued to be observed at Wk 48. Of 426 pts evaluated for safety, ≥1 treatment-emergent adverse event (TEAE) was reported in 97.6% vs 96.7% of pts in the PELA+RUX vs PBO+RUX arms; Grade ≥3 events were reported in 56.6% vs 62.1% of pts. Rates of the most common TEAEs (≥10%) at Wk 48 in the PELA+RUX vs PBO+RUX arms were similar to rates at Wk 24. Updated efficacy and safety results, including data on BMF, mutational profile, and leukemic transformation, will be presented. Conclusion At Wk 48, PELA+RUX continued to show improvements in spleen volume, TSS, multiple measures of anemia, and the BM microenvironment vs PBO+RUX, impacting the four hallmarks of MF. These data suggest that PELA+RUX could lead to more profound and sustained responses in pts with MF vs PBO+RUX.