Selinexor plus ruxolitinib in JAK inhibitor–naïve myelofibrosis: Phase 3 SENTRY trial.

Abstract

LBA6500 Background: Myelofibrosis (MF) is a debilitating myeloproliferative neoplasm marked by splenomegaly, constitutional symptoms, and reduced life expectancy. JAK inhibitors (JAKi), such as ruxolitinib (R), are standard frontline therapy; however, only ~1/3 of R-treated patients (pts) achieve spleen volume reduction ≥35% (SVR35). Despite observed symptom improvements, durable modification of underlying disease biology, including reductions in variant allele frequency (VAF), is limited. Importantly, gains in overall survival remain modest, highlighting a critical unmet need. Selinexor (S), an inhibitor of XPO1-mediated nuclear export, has biologic activity in MF and synergy with R in MPN models. SENTRY evaluated S+R in pts with JAKi–naïve MF. Methods: Pts with JAKi-naïve MF were randomized 2:1 to S 60 mg weekly plus R (per label) or placebo plus R, stratified by DIPSS risk, spleen volume, and baseline platelet count. Eligibility included spleen volume ≥450 cm³, active symptoms, DIPSS Int-1 or higher and platelets ≥100×10⁹/L. Co-primary endpoints were SVR35 and absolute mean change in TSS (AbsTSS; excluding fatigue) at Week 24. SVR35 used a stratified Cochran–Mantel–Haenszel test; TSS used a mixed-effects model for repeated measures. Hierarchical testing (one-sided α=0.025) evaluated SVR35 then AbsTSS. Secondary endpoints included safety and overall survival (OS). Changes in VAF were exploratory. Results: 353 pts were randomized (S+R n=235; R n=118). Baseline characteristics were balanced. At Week 24 SVR35 was achieved in 49.8% of pts in S+R vs 28.0% in R (difference, 21.8%; OR 2.58; 95% CI 1.60 to 4.17; P < .0001). Responses occurred early and were sustained, with SVR35 rates of 49.4% vs 20.3% at Week 12 and 46.9% vs 23.0% at Week 36. SVR35 was achieved at any time in 67.7% vs 44.9%. Mean percent change in spleen volume at Week 24 was −40.0% vs −26.7%. Mean (95% CI) AbsTSS change at Week 24 was −9.9 (−11.2 to −8.6) vs −10.9 (−12.6 to −9.1); adjusted mean difference 0.97 (95% CI -1.07 to 3.02; P = .825). As of Feb 20, 2026, 224 (95.3%) pts in S+R and 106 (89.8%) in R were alive. With median follow-up of 11.6 and 12.6 months, OS favored S+R (HR 0.43; 95% CI 0.19 to 1.00; nominal P = .022) with early separation of Kaplan–Meier curves emerging around Month 9. VAF reduction ≥20% at Week 24 occurred in 32.0% vs 23.9% and correlated with SVR35 response. TEAEs occurred in 99.1% in S+R and 97.4% in R (grade ≥3 in 70.1% vs 50.0%); however, TEAEs leading to treatment discontinuation were low (14.5% and 8.6%). TEAEs leading to death occurred in 0.9% vs 2.6%. Confirmed leukemic transformation was 1.7% in each arm. Conclusions: S+R significantly improved spleen response vs R alone with earlier, deeper, and sustained response rates, comparable symptom improvement from baseline, and a manageable safety profile. The improved spleen response, early OS signal, and VAF reductions observed in SENTRY position S+R as a novel combination approach for frontline treatment of MF. Clinical trial information: NCT04562389 .