Ruxolitinib has been a cornerstone of treatment for myelofibrosis (MF) since its approval, offering meaningful improvements in splenomegaly and constitutional symptoms. However, not all cases responded, and many patients developed resistant disease. Selinexor, an oral selective inhibitor of exportin 1 (XPO1), has shown activity in MF and was hypothesized to complement ruxolitinib’s mechanism of action by targeting the malignant clone through a separate pathway. The phase III SENTRY trial (Bose et al, 2026) was designed to evaluate whether the addition of selinexor to ruxolitinib could improve outcomes beyond what ruxolitinib alone achieves in JAK inhibitor-naive MF.
SENTRY was a double-blind, randomized, phase III trial in which patients with JAK inhibitor-naive MF were randomly assigned 2:1 to receive selinexor plus ruxolitinib or placebo plus ruxolitinib. The coprimary endpoints were spleen volume reduction of ≥35% (SVR35) and the absolute mean change in Total Symptom Score (AbsTSS, excluding fatigue) from baseline to week 24. The enrolled population was representative of the broader MF landscape: median age was 67 years, 57.5% of patients were male, and disease subtypes included primary MF (50.7%), post-essential thrombocythemia MF (25.8%), and post-polycythemia vera MF (23.2%). By Dynamic International Prognostic Scoring System risk category, 56.1% of patients were intermediate-1, 35.1% were intermediate-2, and 8.8% were high risk.
At week 24, SVR35 was achieved in 49.8% of patients in the selinexor plus ruxolitinib group compared with 28.0% in the placebo plus ruxolitinib group—a between-group difference of 21.8 percentage points (odds ratio [OR] 2.58; 95% CI, 1.60 to 4.17; P <.0001). This superiority was apparent as early as week 12 (49.4% vs 20.3%) and maintained through the final study assessment at week 36. The coprimary endpoint of AbsTSS at week 24 was not met; both arms demonstrated comparable symptom improvement from baseline (–9.9 vs –10.9), with no significant between-group difference. Exploratory analyses showed reductions in driver mutation variant allele frequency (VAF) and peripheral blasts in the selinexor-containing arm, consistent with XPO1 inhibition exerting a direct cytotoxic effect on the malignant clone. Patients with ≥20% VAF reduction at week 24 were significantly more likely to demonstrate SVR35 (OR 3.22; 95% CI, 1.81 to 5.72). Early overall survival (OS) data showed a hazard ratio of 0.43 (95% CI, 0.19 to 1.00; nominal P = .022) at a median follow-up of approximately 12 months, though this finding remains immature. Grade ≥3 adverse events occurred in 70.1% vs 50.0% of patients, most commonly cytopenias; nausea was more frequent with selinexor but predominantly low-grade and early.
These results suggest that the combination of selinexor plus ruxolitinib delivers a meaningful and rapid improvement in spleen response in the frontline MF setting, with an emerging but still preliminary signal of potential OS benefit.
High Level
The dissociation between the significant SVR35 improvement and the equivalent AbsTSS outcome raises important questions about endpoint design in MF combination trials, specifically whether ruxolitinib’s robust symptomatic effect creates a floor effect that limits the sensitivity of symptom-based coprimary endpoints when a JAK inhibitor anchors the backbone. The exploratory VAF and peripheral blast reduction data suggest that XPO1 inhibition plays a role in driving meaningful clonal suppression, warranting prospective validation of molecular response as a pharmacodynamic biomarker and potential surrogate endpoint. Close monitoring of the early OS hazard ratio of 0.43 is needed as follow-up matures; if this signal holds, and deeper and earlier spleen responses driven by clonal cytotoxicity translate into a survival benefit, it may fundamentally reframe frontline treatment goals in MF.
Ground Level
The SENTRY trial may provide community oncologists with a meaningful new option for patients with JAK inhibitor-naive MF and significant splenomegaly, as nearly half of patients on the combination demonstrated SVR35 vs just over a quarter on ruxolitinib alone, with benefit visible as early as 12 weeks. Importantly, however, symptom improvement was equivalent between arms, so the principal advantage of the combination is spleen control, not additional symptom relief. If this combination becomes integrated into NCCN Guidelines or receives US regulatory approval, physicians considering this regimen might enact proactive management of cytopenias and nausea, given the higher rate of grade ≥3 adverse events with the combination (70.1% vs 50.0%), Patient selection is likely to be multifactorial, incorporating disease burden, fitness, and tolerability alongside spleen response goals, although expert guidance will be necessary to provide specific cutoff details on how to approach potential integration of this combination into the therapeutic repertoire for first-line treatment of MF.