Durable efficacy and long-term safety with pelabresib plus ruxolitinib in JAK Inhibitor–Naive myelofibrosis: 96-week Results from the Phase III MANIFEST-2 study

Abstract Janus kinase inhibitor (JAKi) monotherapy, the standard of care for myelofibrosis (MF), improves splenomegaly and symptom burden but provides limited depth and durability of response. There is a significant unmet need for new combination strategies that adequately address the underlying disease biology and further improve clinical outcomes in MF. Pelabresib (PELA) is an oral, small molecule inhibitor of bromodomain and extraterminal domain (BET) proteins and can act in a complementary manner with the JAK1/2 inhibitor ruxolitinib (RUX) by targeting inflammatory pathways that are regulated by BET-mediated gene expression. PELA+RUX is being investigated in MANIFEST-2 (NCT04603495) in patients (pts) with JAKi-naive MF. The study met its primary endpoint (Rampal RK. Nat Med. 2025). Here, updated 96-wk results from MANIFEST-2 are reported. MANIFEST-2 is a double-blind, randomized, Phase III study in pts with JAKi-naive MF. Pts were randomized 1:1 to PELA or placebo (PBO) once daily with RUX twice daily. Efficacy and safety endpoints were assessed at Wk 96, including ≥35% reduction in spleen volume (SVR35) response, total symptom score (TSS), hemoglobin (Hgb) response, safety, and survival outcomes. As of March 2, 2025, pts had been followed for at least 96 wks, with 56.5% (121/214) of pts in the PELA+RUX arm and 59.3% (128/216) in the PBO+RUX arm completing 96 wks of assigned treatment. With a median follow-up of 115.9 wks at the cutoff date, approximately half of pts were still on assigned treatment (48.6% in the PELA+RUX arm vs 48.1% in the PBO+RUX arm). Most common reasons for discontinuation were adverse events (AE; 21.0% vs 13.4%), physician decision (7.5% vs 14.8%), withdrawal of consent (10.3% vs 6.0%), and protocol-defined disease progression (5.1% vs 6.9%, respectively). At Wk 96, among evaluable pts on treatment, SVR35 response was observed in 91.5% (97/106) in the PELA+RUX arm vs 57.5% (65/113) in the PBO+RUX arm (difference, 34.0%), representing 45.3% (97/214) vs 30.1% (65/216) of the intent-to-treat population in each arm (difference, 15.2%), respectively, and showing sustained benefit in pts remaining on treatment. At Wk 96, absolute change in TSS from baseline was −15.07 in the PELA+RUX arm vs −12.48 in the PBO+RUX arm (difference, −2.59; 95% CI, −5.23 to 0.05); ≥50% reduction in TSS from baseline (TSS50) was achieved by 36.9% (79/214) and 28.2% (61/216) of pts, respectively, demonstrating durable responses in symptom improvement. At Wk 96, dual SVR35 and TSS50 responses were reported in 31.8% (68/214) vs 15.7% (34/216) of pts, respectively. Hgb response was achieved by 17.8% (38/214) vs 11.6% (25/216) of pts in the PELA+RUX vs PBO+RUX arms. By 96 wks from the start of treatment, red blood cell transfusions were required in 33.1% (44/133) vs 39.9% (57/143) of evaluable pts, respectively. Ongoing Wk 96 analyses of exploratory biomarker data, including the dynamics of variant allele frequencies of MF driver genes, will be presented. Treatment-emergent AEs were mainly low grade and remained similar between arms over time (PELA+RUX n=212 vs PBO+RUX n=214: any grade, 99.5% vs 98.1%; Grade ≥3, 67.5% vs 70.1%). Externally and independently adjudicated leukemic transformations were reported in 5.1% (11/216) of pts on PELA+RUX and in 3.7% (8/214) of pts on PBO+RUX by the 96-wk data cutoff. A total of 28 deaths and 22 progression-free survival (PFS) events were reported in the PELA+RUX arm vs 32 deaths and 34 PFS events in the PBO+RUX arm (overall survival OS HR, 0.986; 95% CI, 0.590–1.647; PFS HR, 0.746; 95% CI, 0.432–1.291; not powered for OS/PFS). These results represent the longest follow-up to date of a randomized combination trial in MF. After at least 96 wks of follow-up, PELA+RUX continued to show deep and durable improvements in splenic response, TSS, SVR35/TSS50 dual response, and anemia vs PBO+RUX, with stable or improving response rates among pts remaining on long-term treatment. PELA+RUX had a similar safety profile to PBO+RUX. The early imbalance in leukemic transformation cases decreased over time, and the observed frequency was in line with what is historically seen in MF. Longer-term follow-up showed numerically fewer deaths and fewer progression events in the PELA+RUX arm vs the PBO+RUX arm. Overall findings suggest that PELA+RUX provides clinically meaningful benefits vs RUX monotherapy, with potential for disease modification and improving survival in pts with MF.