ABSTRACT
Background
Janus kinase inhibitors (JAKis), the current standard of care for myelofibrosis (MF), provide clinical benefit, but responses are frequently incomplete, non‐durable, and associated with cytopenias, underscoring the need for therapies with novel mechanisms of action. Selinexor, an oral selective inhibitor of Exportin 1 (XPO1), modulates nuclear‐cytoplasmic transport, inflammatory signaling, and cancer cell proliferation.
Methods
XPORT‐MF‐035, a global, Phase 2, randomized, open‐label study comparing selinexor monotherapy with physician's choice (PC) in MF previously treated with JAKi, randomized patients 1:1 to selinexor or PC, with crossover permitted upon disease progression or inadequate spleen response. Primary endpoint was spleen volume reduction ≥ 35% (SVR35). Additional endpoints included symptom improvement, hematologic outcomes, plasma cytokine analyses, and safety.
Results
Twenty‐four patients were enrolled and treated. In the efficacy‐evaluable population, SVR35 at any time was achieved by 29% (2/7) and 13% (1/8) of selinexor‐ and PC‐treated patients, respectively; 8% (1/12) in each arm achieved SVR35 at Week 24 in the intent‐to‐treat population. Among evaluable patients, 40% (2/5) achieved SVR35 following selinexor crossover. Symptom improvement was observed with selinexor but not PC. Selinexor‐treated patients experienced fewer Grade ≥ 3 anemia events with lower transfusion burden and reductions in multiple proinflammatory cytokines. Selinexor was generally well tolerated; common treatment‐emergent adverse events (TEAEs) included anemia, asthenia, and decreased weight. One TEAE leading to death occurred in the selinexor and two in PC; none were considered treatment‐related.
Conclusions
Despite limited sample size, XPORT‐MF‐035 provides descriptive data on the safety, tolerability, and biological activity of selinexor monotherapy in previously treated MF, supporting further evaluation in this population.
Trial Registration
ClinicalTrials.gov identifier: NCT04562870
Keywords: JAKi, myelofibrosis, Phase 2, randomized, selinexor
1. Introduction
Myelofibrosis (MF) is a debilitating, chronic myeloproliferative neoplasm characterized by clonal stem cell–derived myeloproliferation, bone marrow fibrosis, splenomegaly, anemia, and constitutional symptoms, resulting in substantial morbidity and reduced survival [1, 2, 3, 4, 5]. MF is driven by a proinflammatory state commonly associated with JAK2, CALR, or MPL mutations, leading to dysregulated JAK–STAT signaling and aberrant cytokine production that contribute to disease progression and clinical manifestations [3, 6, 7, 8, 9].
Janus kinase inhibitors (JAKis) are the current standard of care for MF and provide clinical benefit through reductions in splenomegaly and symptom burden [3, 10, 11]. However, responses are frequently limited and not durable, with more than 50% of patients discontinuing therapy within 3 years due to loss of efficacy, cytopenias, or other toxicities [12, 13, 14]. Moreover, JAKi therapy has not consistently demonstrated disease‐modifying effects such as reversal of bone marrow fibrosis or elimination of the malignant clone [1, 15, 16]. Patients who relapse or become resistant often switch to an alternative JAKi but may continue to experience suboptimal spleen responses, anemia, or transfusion dependence, reflecting overlapping mechanisms of action [3]. These limitations underscore the need for therapies with novel mechanisms of action that target additional pathways involved in MF pathogenesis and malignant clone survival [17, 18].
Selinexor is an investigational, oral selective inhibitor of Exportin 1 (XPO1), a nuclear export protein overexpressed in multiple malignancies. XPO1 inhibition leads to nuclear retention of tumor suppressor proteins, induction of apoptosis, and modulation of NF‐κB‐mediated inflammatory signaling [19]. Nonclinical models have identified XPO1‐mediated nuclear export as a critical dependency in MF, demonstrating selective activity of selinexor against MF cells, suppression of proinflammatory signaling, and induction of apoptosis in hematopoietic progenitors from patients with ruxolitinib‐refractory disease [20, 21, 22]. Early clinical studies suggest activity of selinexor both as monotherapy and in combination with JAKis [22, 23].
Here, we report results from the Phase 2 XPORT‐MF‐035 study evaluating selinexor monotherapy compared with the physician's choice (PC) in patients with MF previously treated with JAKi.
2. Materials and Methods
2.1. Study Design and Treatment
XPORT‐MF‐035 is a Phase 2, global, multicenter, randomized, open‐label study evaluating the safety and efficacy of selinexor monotherapy versus PC in patients with previously treated MF (NCT04562870).
Patients in the selinexor arm were administered 80 mg selinexor once weekly (QW) for two cycles, followed by 60 mg QW in subsequent cycles. To manage nausea, all patients in the selinexor arm received two antiemetics for nausea prophylaxis for the first two cycles, followed by a reduction to one antiemetic from Cycle 3 onwards. PC consisted of standard, non‐investigational therapies selected by the investigator, including ruxolitinib, fedratinib, chemotherapy (e.g., hydroxyurea), anagrelide, corticosteroids, hematopoietic growth factors, androgens, interferon, combinations thereof, or best supportive care. Investigational therapies were not permitted.
2.2. Patients
Eligible patients were ≥ 18 years of age with a diagnosis of primary MF or post‐essential thrombocythemia or post‐polycythemia vera MF according to the 2016 World Health Organization classification of MPN [24], refractory, relapsed, or intolerant to prior JAKi exposure for ≥ 6 months, with measurable splenomegaly during screening, and an Eastern Cooperative Oncology Group performance status ≤ 2. Key exclusion criteria included prior exposure to an XPO1 inhibitor, > 5% peripheral blood blasts or > 10% bone marrow blasts, recent MF‐directed therapy within 21 days prior to Cycle 1 Day 1 (C1D1), clinically significant gastrointestinal disorders affecting absorption, use of strong CYP3A inhibitors or inducers within protocol‐defined washout periods, or major surgery within 28 days of C1D1. Full eligibility criteria and key definitions are provided in Table S1.
2.3. Randomization and Crossover
Patients were randomized 1:1 to selinexor or PC, stratified by prior fedratinib exposure (yes vs. no) and baseline spleen size (< 10 cm vs. ≥ 10 cm below the left costal margin). Patients assigned to PC were permitted to crossover to selinexor upon development of progressive disease, defined as > 25% increase in spleen volume from nadir, or failure to achieve spleen volume reduction ≥ 35% (SVR35) by Week 36. Patients assigned to PC received a single investigator‐selected regimen; changes or additions to MF‐directed therapy were not permitted during treatment, with the exception of crossover to selinexor per protocol.
2.4. Endpoints and Assessments
The primary endpoint was SVR35 assessed by magnetic resonance imaging or computed tomography and evaluated by an independent radiology committee. Key secondary endpoints included spleen volume reduction of ≥ 25% (SVR25), total symptom score reduction ≥ 50% (TSS50) using the Myelofibrosis Symptom Assessment Form v4.0 (MFSAF v4.0), and absolute mean change of TSS (AbsTSS) from baseline to Week 24. Safety and tolerability were assessed as secondary endpoints. Variant allele frequency (VAF) in MF driver genes (JAK2, CALR, and MPL) and other high‐molecular‐risk genes (ASXL1, EZH2, IDH1/IDH2, SRSF2, and U2AF1) were evaluated locally. Exploratory endpoints included red blood cell transfusion, hemoglobin changes, and plasma cytokine and chemokine levels. Plasma cytokine and chemokine concentrations were centrally evaluated using a Luminex 200 system human cytokine/chemokine array 71‐plex panel (Eve Technologies, Calgary, Canada). TGF‐β and hepcidin were evaluated with the Millipore TGF‐β 3‐Plex Discovery Assay (MilliporeSigma, Burlington, MA, USA) and the R&D Human Hepcidin ELISA assay (R&D Systems, Minneapolis, MN, USA), respectively. Hemoglobin improvements were defined as ≥ 1.5 g/dL increase from baseline without transfusion in the preceding 12 weeks. Transfusion dependence was defined as ≥ 6 units of packed red blood cells in the prior 12 weeks, for a hemoglobin level of < 8.5 g/dL, in the absence of bleeding or treatment‐induced anemia. Adverse events (AEs) were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE), Version 5.0.
2.5. Statistical Analysis
The original study was designed with a planned sample size of approximately 112 patients, estimated to provide at least 80% power at a one‐sided alpha level of 0.025 (including a Haybittle–Peto stopping boundary with 0.001α), assuming true response rates of 40% for selinexor and 15% for PC. Enrollment was halted early due to a sponsor decision to prioritize resources for the SENTRY Phase 3 study evaluating selinexor or placebo in combination with ruxolitinib in JAK inhibitor–naive patients with MF; therefore, the results presented are descriptive in nature. Efficacy analyses were performed in the intent‐to‐treat (ITT) and efficacy evaluable populations after all patients received at least 24 weeks of treatment or discontinued treatment. The efficacy evaluable population included all patients who received selinexor at any time and had at least one spleen volume assessment. The ITT population included all patients randomized to study treatment, exclusive of crossover. The TSS evaluable population consisted of patients with > 0 TSS at baseline and any non‐missing TSS evaluations post‐baseline. The safety population included all patients who were randomized and received at least one dose of the study drug. Efficacy analyses were descriptive in nature. SVR35 was analyzed using a Cochran–Mantel–Haenszel (CMH) test stratified by randomization stratification factors, with nominal estimates of odds ratios and 95% confidence intervals reported. Spleen volume response at any time was assessed by investigators. For crossover patients, SVR was calculated based on new spleen volume baseline at the time of crossover. Fatigue was excluded from TSS analysis, and patients with missing symptom data at Week 24 were considered non‐responders.
3. Results
3.1. Patients
Between December 1 2021 and January 4 2023, a total of 24 patients were enrolled, with 12 randomized to each arm. Baseline demographic and disease characteristics were generally comparable between treatment arms (Table 1). Median age was 70 years (range, 43–79), and 46% of patients were 65–74 years of age. Overall, 42% of patients were female. The majority of patients carried a JAK2 V617F mutation (71%), with CALR mutations present in 13% and 17% classified as triple‐negative. Locally assessed bone marrow fibrosis grade was MF‐3 in 50% of patients, MF‐2 in 38%, and MF‐1 in 13%.
TABLE 1.
Patient baseline characteristics.
|
Physician's choice (N = 12) |
Selinexor (N = 12) |
Overall (N = 24) |
|
|---|---|---|---|
| Age (years), median (range) | 71 (60, 79) | 70 (43, 78) | 70 (43, 79) |
| Age category, n (%) | |||
| 18–50 | 0 | 1 (8) | 1 (4) |
| 51–64 | 3 (25) | 4 (33) | 7 (29) |
| 65–74 | 6 (50) | 5 (42) | 11 (46) |
| ≥ 75 | 3 (25) | 2 (17) | 5 (21) |
| Female, n (%) | 7 (58) | 3 (25) | 10 (42) |
| Region | |||
| United States | 0 (0) | 3 (25) | 3 (12) |
| European Union | 9 (75) | 6 (50) | 15 (63) |
| China | 3 (25) | 3 (25) | 6 (25) |
| Transfusion status, n (%) | |||
| Transfusion‐dependent a | 9 (75) | 10 (83) | 19 (79) |
| Transfusion‐independent | 3 (25) | 2 (17) | 5 (21) |
| Myelofibrosis grade, b n (%) | |||
| MF‐1 | 0 (0) | 1 (8) | 1 (4) |
| MF‐2 | 3 (25) | 3 (25) | 6 (25) |
| MF‐3 | 7 (58) | 3 (25) | 10 (42) |
| Missing | 2 (17) | 5 (42) | 7 (29) |
| MF type, n (%) | |||
| Primary MF | 5 (42) | 6 (50) | 11 (46) |
| Post‐ET MF | 3 (25) | 3 (25) | 6 (25) |
| Post‐PV MF | 4 (33) | 3 (25) | 7 (29) |
| DIPSS risk, n (%) | |||
| Int‐1 | 2 (17) | 6 (50) | 8 (33) |
| Int‐2 | 8 (67) | 3 (25) | 11 (46) |
| High | 2 (17) | 3 (25) | 5 (21) |
| Driver mutation, c n (%) | |||
| JAK2 | 8 (67) | 9 (75) | 17 (71) |
| CALR | 2 (17) | 1 (8) | 3 (13) |
| Triple negative | 2 (17) | 2 (17) | 4 (17) |
| Prescence of high‐risk mutations, c n (%) | |||
| Mutated | 0 | 1 (8) | 1 (4) |
| Non‐mutated | 12 (100) | 11 (92) | 23 (96) |
| Number of prior lines of treatment (min, max) | 2 (1, 3) | 2 (1, 4) | 2 (1, 4) |
| Time from last prior treatment to randomization, weeks, median (min, max) | 6 (3, 49) | 5 (3, 52) | 6 (3, 52) |
| Baseline spleen volume (cm3), mean SD) | 2373 ± 1410 | 1870 ± 910 | 2132 ± 1198 |
| TSS, mean (SD) d | 17.1 ± 15.3 | 18.6 ± 14.4 | NA |
| Hemoglobin (g/dL), mean (SD) | 9.4 ± 1.9 | 10.3 ± 2.2 | NA |
| Hemoglobin (g/dL), n (%) e | |||
| < 8.0 | 3 (25.0) | 2 (16.7) | 5 (20.8) |
| 8.0–< 10.0 | 5 (41.7) | 4 (33.3) | 9 (37.5) |
| ≥ 10.0 | 4 (33.3) | 6 (50.0) | 10 (41.7) |
| Platelets (109/L), n (%) | |||
| < 100 | 0 | 2 (16.7) | 2 (8.3) |
| 100–200 | 4 (33.3) | 5 (41.7) | 9 (37.5) |
| > 200 | 7 (58.3) | 5 (41.7) | 12 (50.0) |
| Missing | 1 (8.3) | 0 | 1 (4.2) |
Abbreviations: CALR, calreticulin; DIPSS, Dynamic International Prognostic Scoring System; ET, essential thrombocytopenia; JAK2, Janus kinase 2; JAKi, Janus kinase inhibitor; MF, myelofibrosis; NA, not applicable; PC, physician's choice; PV, polycythemia vera; RUX, ruxolitinib; SD, standard deviation; SEL, selinexor; TSS, total symptom score.
Transfusion dependence defined per IWG‐MRT/ELN 2013 criteria: ≥ 6 units of packed red blood cells (PRBC), in 12 weeks prior to study enrollment, for a hemoglobin level of < 8.5 g/dL, in the absence of bleeding or treatment‐induced anemia. In addition, the most recent transfusion episode must have occurred in the 28 days prior to study enrollment.
Myelofibrosis grading was locally assessed.
MF driver genes (JAK2, CALR, and MPL) and other high–molecular risk genes (ASXL1, EZH2, IDH1/IDH2, SRSF2, and U2AF1) were evaluated locally.
Based on the Myelofibrosis Symptom Assessment Form version 4.0 (excluding fatigue) in patients with nonzero baseline score (SEL: n = 8; PC: n = 11).
Baseline hemoglobin calculated based on the rolling mean.
Most patients had intermediate‐2 or high‐risk disease by DIPSS (83% in the PC arm; 50% in the selinexor arm). Baseline spleen volume was numerically lower in the selinexor arm compared with the PC arm; mean (standard deviation [SD]) 1870 (910) cm3 versus 2373 (1410) cm3. Mean (SD) baseline hemoglobin was 10.3 (2.2) g/L in selinexor and 9.4 (1.9) g/L in PC; 79% of patients were transfusion dependent at study entry. Patients in both arms had a mean of two prior lines of therapy (Table S2); 75% of patients received ruxolitinib as the most recent prior therapy, and 96% received ruxolitinib as part of at least one line of prior therapy, with a median of 51.5 days (range 22–2412) from the last dose of ruxolitinib prior to randomization.
At data cutoff (October 31 2024), median (range) treatment duration was 14.7 (0.1–129.6) weeks in the selinexor arm and 17.3 (1.6–113.6) weeks in the PC arm. Median (range) time on study was 27.4 (4.3–142.7) weeks in the selinexor arm and 65.1 (1.7–113.6) weeks in the PC arm. Four selinexor‐treated patients and one PC‐treated patient remained on treatment for ≥ 1 year; three patients (two in the selinexor arm and one in PC) remained on treatment at data cutoff (Figure 1). Five selinexor‐treated patients and two PC‐treated patients discontinued study participation due to death. Of these, two deaths in the selinexor arm occurred during study treatment. In the PC arm, 75% of patients received ruxolitinib as part of PC (Table S2).
FIGURE 1.
Patient disposition. Data cutoff date: October 31, 2024.
Six patients (50%) crossed over from PC to selinexor due to disease progression or inadequate spleen response. The median time from last PC treatment to first treatment with selinexor in crossover was three weeks. Four of five patients evaluable for efficacy following crossover had received ruxolitinib as PC. One crossover patient was not evaluable for efficacy due to withdrawal from the study after two doses of selinexor prior to the next spleen assessment.
3.2. Efficacy
In the ITT population, SVR35 was achieved at Week 24 by 8% of patients in each arm, while SVR25 was observed in 25% of selinexor‐treated patients and 17% of PC‐treated patients (Table 2). SVR35 at any time was achieved in 17% of selinexor‐treated patients and 8% of PC patients, with corresponding SVR25 rates of 33% and 25%. Among efficacy‐evaluable patients, SVR35 was observed at Week 24 in 14% of selinexor‐treated patients and 13% of PC patients, and at any time in 29% of selinexor‐treated patients and 13%, respectively (Figure 2). SVR25 was observed at Week 24 in 43% and 25% of evaluable selinexor‐treated and PC patients, respectively, and at any time in 57% and 38%.
TABLE 2.
Efficacy outcomes.
|
Physician's choice (n = 12) |
Selinexor (n = 12) |
All selinexor treated (n = 18) |
|
|---|---|---|---|
| Spleen volume reduction | |||
| Intent‐to‐treat | |||
| SVR25 at Week 24, n/N (%) | 2/12 (17) | 3/12 (25) | 3/18 (17) |
| SVR35 at Week 24, n/N (%) | 1/12 (8) | 1/12 (8) | 1/18 (6) |
| Efficacy evaluable | |||
| SVR25 at any time, n/N (%) | 3/8 (38) | 4/7 (57) | 8/12 (67) |
| SVR35 at any time, n/N (%) | 1/8 (13) | 2/7 (29) | 4/12 (33) |
| Total symptom score response | |||
| TSS50 at Week 24, n/N (%) | 0/9 (0) | 2/7 (29) | 2/11 (18) |
| Baseline TSS, mean (SD) | 17.3 (17.1) | 18.8 (15.6) | 17.6 (15.5) |
| AbsTSS at Week 24, mean (SD) | −0.0 (5.4) | −5.9 (14.0) | −3.7 (15.6) |
Note: AbsTSS, absolute mean change from baseline to Week 24 with last observation carried forward imputation for missing data at Week 24.
Abbreviations: SD, standard deviation; SVR, spleen volume reduction; SVR25, spleen volume reduction ≥ 25%; SVR35, spleen volume reduction ≥ 35%.TSS, total symptom score; TSS50, total symptom score reduction of ≥ 50%.
FIGURE 2.
Efficacy outcomes. Spleen volume reduction and symptom improvement with selinexor monotherapy compared with physician's choice (efficacy evaluable). Arrows denote crossover patients. Spleen volume reduction anytime, as assessed by investigator in the efficacy evaluable population. Baseline spleen volume for the crossover patients is the volume at the time of crossover. For crossover patients, all spleen volume changes were calculated based on new spleen volume baseline at the time of crossover. Of these 5/6 patients received 80 mg and 1/6 received 60 mg (prior peginterferon) as initial selinexor dose. One patient withdrew consent shortly after crossover to selinexor arm and is not included in the analysis. Physician's choice treatment prior to crossover: ruxolitinib (PC4X, PC6X, and PC7X), peginterferon (PC1X), danazol, prednisone, and folic acid (PC3X), and ruxolitinib, thalidomide, and prednisone (PC2X).
Among evaluable crossover patients, two of five (40%) achieved SVR35, and four of five (80%) achieved SVR25 following initiation of selinexor monotherapy (Figure 2).
Improvement in symptom burden was observed in selinexor‐treated patients, but not in PC‐treated patients, reflected in both TSS50 and AbsTSS (Table 2). At Week 24, TSS50 was achieved in 29% (2/7) of selinexor‐treated patients and in none of the PC‐treated patients (0/9). Mean absolute change in TSS at Week 24 was −5.9 (SD 14.0) with selinexor and −0.0 (SD 5.4) with PC.
3.3. Hematologic Outcomes and Transfusion Burden
The change in mean hemoglobin levels at Week 24 relative to matched baseline was +4.03% in the selinexor arm and −3.05% in the PC arm (Figure 3). Hemoglobin improvement was observed in 25% of selinexor‐treated patients and 8% of PC‐treated patients.
FIGURE 3.
Standard mean hemoglobin levels during treatment. PC, Physician's choice; SEL, selinexor.
Consistent with these findings, selinexor‐treated patients required fewer red blood cell transfusions than those receiving PC (Figure 4). The mean number of transfused units per patient‐month was 0.54 in the selinexor arm and 1.14 in the PC arm. One selinexor‐treated patient who received only a single dose of selinexor and subsequently required nine transfusion units was excluded from transfusion analyses.
FIGURE 4.
Time on study and frequency of transfusions. Available mutation and variant allele frequency data are shown on the left of each row. Patients highlighted were transfusion dependent at baseline. PC, physician's choice; RBC, red blood cell; SEL, selinexor.
3.4. Cytokine and Chemokine Analyses
Paired pre‐ and post‐treatment plasma samples (screening and C2D1) were available for five patients in each arm. All paired samples in the PC arm were from patients receiving ruxolitinib. Reductions in several proinflammatory cytokines and chemokines relevant to MF were observed following treatment with selinexor, but not PC (Figure 5). Median changes in the selinexor and PC arms were −29% and +19% for interleukin‐6 (IL‐6), −37% and +25% for interleukin‐8 (IL‐8), −27% and +31% for monocyte chemoattractant protein‐1 (MCP‐1), and +25% and −4% for tumor necrosis factor alpha (TNFα), respectively. Notably, increases in T‐cell‐associated cytokines (IL‐2, IL‐1β, and IL‐17A) were also observed. Hepcidin, an iron‐regulating hormone primarily produced in the liver and typically elevated in anemia and pro‐inflammatory states, showed median changes of −30% in the selinexor arm compared with +6% in the PC arm.
FIGURE 5.
Myelofibrosis‐relevant cytokine and chemokine levels in patients. Heatmap shows the median % change from baseline to Week 4. The top dendrogram shows unsupervised hierarchical clustering of cytokines and is colored orange to highlight clusters of cytokines and hepcidin reduced by SEL but not PC, and green to highlight a cluster of T‐cell activating cytokines. PC, physician's choice; SEL, selinexor.
3.5. Safety and Tolerability
All 24 enrolled patients were evaluable for safety. Treatment‐emergent adverse events (TEAEs) occurred in 92% of patients, with Grade ≥ 3 TEAEs reported in 54%. The most common (≥ 10%) TEAEs with selinexor and PC were anemia (25% and 50%), asthenia (42% and 25%), and decreased weight (50% and 17%) (Table 3). After crossover to selinexor (n = 6), anemia (n = 3), diarrhea (n = 2), nausea (n = 3), thrombocytopenia (n = 2), and decreased weight (n = 4) were reported. Grade ≥ 3 anemia was reported in 17% of selinexor‐treated patients and 50% of PC‐treated patients. No TEAEs led to treatment discontinuation in either arm. TEAEs leading to dose reduction and dose interruption were reported in 33% and 50% of selinexor‐treated patients and 25% and 25% of PC‐treated patients, respectively. One death in the selinexor arm (unexplained death) and two deaths in the PC arm (cardiac failure and sudden death) were reported as a Grade 5 TEAE. The remaining deaths occurred outside the TEAE reporting period. All deaths were adjudicated by the investigators as unrelated to study treatment based on clinical circumstances.
TABLE 3.
Treatment‐emergent adverse events (TEAEs).
| Treatment‐emergent adverse events, n (%) |
Physician's choice (n = 12) |
Selinexor a (n = 12) |
Overall (N = 24) |
|---|---|---|---|
| Any grade, > 10% | |||
| Anemia | 6 (50) | 3 (25) | 9 (38) |
| Asthenia | 3 (25) | 5 (42) | 8 (33) |
| Weight decreased | 2 (17) | 6 (50) | 8 (33) |
| Upper respiratory tract infection | 4 (33) | 3 (25) | 7 (29) |
| Cough | 1 (8) | 4 (33) | 5 (21) |
| Thrombocytopenia | 1 (8) | 4 (33) | 5 (21) |
| Nausea | 0 (0) | 4 (33) | 4 (17) |
| Pyrexia | 1 (8) | 3 (25) | 4 (17) |
| COVID‐19 | 1 (8) | 2 (17) | 3 (13) |
| Constipation | 0 (0) | 3 (25) | 3 (13) |
| Hypokalemia | 2 (17) | 1 (8) | 3 (13) |
| Hypotension | 0 (0) | 3 (25) | 3 (13) |
| Patients with 1+ serious TEAE | 5 (42) | 5 (42) | 10 (42) |
| Patients with 1+ Grade 3+ event | 9 (75) | 7 (58) | 16 (67) |
| Grade 3+ data | |||
| Anemia | 6 (50) | 2 (17) | 8 (33) |
| Asthenia | 1 (8) | 0 (0) | 1 (4) |
| Nausea | 2 (17) | 0 (0) | 2 (8) |
| Hypotension | 1 (8) | 0 (0) | 1 (4) |
| Cardiac failure | 0 (0) | 2 (16) | 2 (8) |
| Decreased appetite | 2 (17) | 0 (0) | 2 (8) |
| Dehydration | 1 (8) | 0 (0) | 1 (4) |
| Leukopenia | 1 (8) | 1 (8) | 2 (8) |
| Neutropenia | 1 (8) | 0 (0) | 1 (4) |
| Pneumonia | 0 (0) | 1 (8) | 1 (4) |
| Vomiting | 1 (8) | 0 (0) | 1 (4) |
| Arthritis | 0 (0) | 1 (8) | 1 (4) |
| Cerebral ischaemia | 1 (8) | 0 (0) | 1 (4) |
| Confusional state | 1 (8) | 0 (0) | 1 (4) |
| Death | 1 (8) | 0 (0) | 1 (4) |
| Febrile neutropenia | 1 (8) | 0 (0) | 1 (4) |
| Gastroenteritis | 1 (8) | 0 (0) | 1 (4) |
| Hepatic fibrosis | 1 (8) | 0 (0) | 1 (4) |
| Hyperkalaemia | 0 (0) | 1 (8) | 1 (4) |
| Myelodysplastic syndrome | 1 (8) | 0 (0) | 1 (4) |
| Pneumonia pneumococcal | 1 (8) | 0 (0) | 1 (4) |
| Pneumonia viral | 0 (0) | 1 (8) | 1 (4) |
| Sepsis | 1 (8) | 0 (0) | 1 (4) |
| Sudden death | 0 (0) | 1 (8) | 1 (4 |
Abbreviation: TEAE, treatment‐emergent adverse event.
For patients who crossed over, treatment‐emergent adverse events that occurred after the crossover are not included. After crossover to selinexor (n = 6), anemia was reported in three patients, diarrhea in two patients, nausea in three patients, thrombocytopenia in two patients, and decreased weight in four patients.
4. Discussion
XPORT‐MF‐035 evaluated selinexor monotherapy compared with PC in patients with MF previously treated with JAKi and demonstrated encouraging signals of clinical activity. Selinexor was generally well tolerated, with AEs consistent with prior reports. SVR and symptom improvement were observed in a subset of selinexor‐treated patients, including after crossover among those refractory or intolerant to PC. These findings align with preclinical and translational data supporting XPO1 inhibition by selinexor as a novel therapeutic strategy to address limitations of JAKis [21, 22].
While SVR35 rates at Week 24 were similar across the two arms in the ITT, SVR25 was numerically higher for selinexor arm. Furthermore, in the efficacy‐evaluable population, more than half of selinexor‐treated patients achieved SVR25 at any time, and one‐third achieved SVR35. Responses were also observed after crossover from PC, most commonly ruxolitinib, suggesting that XPO1 inhibition may engage disease pathways distinct from JAK–STAT signaling, thereby circumventing JAKi resistance [23].
Anemia and transfusion dependence remain major unmet needs in MF and are frequently exacerbated by JAKi therapy [25]. Anemia is a negative prognostic marker for progression and overall survival in MF and commonly leads to dose modifications, interruptions, or discontinuation of frontline JAKi therapy due to increased requirements for RBC transfusions for management of hemoglobin levels [25]. In this study, selinexor‐treated patients, including those who crossed over, experienced fewer Grade ≥ 3 anemia events and a lower transfusion burden than those receiving PC. While baseline hemoglobin differences and treatment heterogeneity limit interpretation, hemoglobin trends and transfusion rates indicate that these differences are unlikely to fully account for the observed differences and suggest that selinexor may be administered without uniformly worsening anemia in patients with MF. Comparisons with emerging anemia‐directed therapies were beyond the scope of this study and will require larger trials.
Certain subsets of elevated cytokines in MF can be correlated with a pro‐inflammatory tumor microenvironment and increased constitutional symptom burden [26, 27, 28, 29]. MF samples can have continued aberrant cytokine production following ruxolitinib treatment [27], highlighting the need for therapeutic strategies that induce inhibition of inflammatory signaling. Exploratory biomarker analyses in the current study demonstrated reductions in several proinflammatory cytokines and chemokines following selinexor treatment but not with PC, consistent with preclinical evidence that XPO1 inhibition modulates inflammatory signaling through mechanisms distinct from JAK–STAT signaling. However, baseline differences and prior treatment exposure may have influenced these findings, which should be confirmed in larger prospective studies. Similarly, hepcidin reduction was observed to be deeper for selinexor treatment versus PC, but low sample sizes hinder the ability to draw a conclusion regarding treatment correlation with hepcidin and anemia changes.
Selinexor was generally well‐tolerated with no new safety signals identified. Clinically relevant AEs remain important considerations, particularly in settings where quality‐of‐life preservation is a key goal and where alternative treatments may offer different tolerability profiles. Most AEs were manageable with dose modifications or supportive care. Further optimization of treatment regimens and supportive care will be critical in future studies to assess relative risk‐benefit profiles.
4.1. Study Limitations and Future Directions
The study was limited by early termination and small sample size, precluding definitive conclusions regarding comparative efficacy or safety. Given the small sample size, baseline imbalances may have had a disproportionate impact on observed outcomes. Of note, the lower baseline hemoglobin and other increased high‐risk features, such as baseline DIPSS and spleen size in the PC group, may have influenced study outcomes, such as transfusion needs. In addition, the heterogeneity of therapies included under PC and the descriptive nature of the analyses limit direct comparisons between treatment groups. Yet, even with these limitations, there is still evidence of clinical activity for single‐agent selinexor in patients who failed PC and underwent crossover. Study enrollment was halted early, resulting in a limited sample size that was not powered to detect statistically significant differences between treatments, and all efficacy analyses were descriptive. Accordingly, the findings should be interpreted cautiously and viewed as hypothesis‐generating.
Despite these limitations, the results support the biological and clinical rationale for further evaluation of selinexor in MF. The potential to target pathways beyond those impacted by JAKis suggests a role for XPO1 inhibition both as monotherapy and in combination with JAKis. Ongoing and planned studies, including the Phase 2 SENTRY‐2 trial of selinexor monotherapy (NCT05980806) and the Phase 3 SENTRY trial evaluating selinexor in combination with ruxolitinib in JAKi‐naïve MF (NCT04562389), will further clarify selinexor's therapeutic potential.
5. Conclusion
In this small cohort, selinexor monotherapy demonstrated preliminary evidence of clinical activity in patients with MF previously treated with JAKi, including those refractory or intolerant to ruxolitinib, with indications of SVR, symptom improvement, hemoglobin stabilization, reduced transfusion burden, and modulation of proinflammatory cytokines. These data demonstrate that XPO1‐mediated nuclear export may play a critical role in MF pathophysiology, supporting further evaluation of selinexor in MF as monotherapy or in combination with JAKis.
Author Contributions
S.G., H.Y., R.A., S.C., H.L., F.B., P.G., M.B., P.S., G.M., P.T., J.Z., T.M., Y.C., A.E., and A.L. contributed to the conceptualization of the study and investigation, wrote the original draft, review, and editing. T.M., A.E., and Y.C. contributed to the formal analysis and interpretation of data. All authors provided a critical review of the manuscript and approved the final submitted version.
Funding
This study was supported by research funding from Karyopharm Therapeutics, Inc.
Ethics Statement
The study protocol was approved by the institutional review board or ethics committee at each participating site and conducted in accordance with the ethical principles that originate from the Declaration of Helsinki [30] and that are consistent with International Council for Harmonisation Good Clinical Practice (GCP) and regulatory requirements as applicable.
Consent
All patients provided written informed consent prior to participation.
Conflicts of Interest
Habte Yimer reports speakers bureau for AbbVie, Genmab, Pfizer, Amgen, Janssen, BeiGene, G1 Therapeutics, and Karyopharm. Rosa Ayala reports speakers bureau for Astellas, Bristol Myers Squibb, Novartis; consultancy for Incyte, Novartis; and holds equity for Altum Sequencing. Françoise Boyer reports an advisory board role for Novartis. Paola Guglielmelli reports speakers bureau for Novartis, AOP, BMS, and GSK; advisory board for Novartis, BMS, GSK, and Incyte. Gabor Mikala reports speaker fees from AbbVie, Amgen, BMS, Johnson‐Johnson, Novartis, Pfizer, Sanofi, Takeda, and research support from AbbVie. Tomer Mark is an employee of Karyopharm Therapeutics Inc. Yi Chai is an employee of Karyopharm Therapeutics Inc. Andrea Ellero is an independent consultant contracted by Karyopharm Therapeutics Inc. Alessandro Lucchesi reports consulting or advisory role for Amgen, Grifols, MorphoSys, Novartis, Protagonist Therapeutics, Sanofi; Speakers bureau for Amgen, Bristol Myers Squibb, Grifols, Incyte, Novartis, Pfizer, Sanofi, Sobi. The other authors declare no conflicts of interest.
Supporting information
Supporting File 1: jha270301‐sup‐0001‐SuppMat.docx
Acknowledgments
JetPub Scientific Communications LLC assisted the authors in the preparation of this manuscript. We thank the patients who participated in this trial, their families and caregivers, the investigators, the study staff at each of the clinical sites, and members of the following departments at Karyopharm Therapeutics: Clinical Operations, Clinical Development, Translational Medicine, and Biostatistics.
Data Availability Statement
Karyopharm Therapeutics is committed to providing qualified scientific and medical researchers with access to anonymized data and protocols from the company's clinical trials for the purpose of conducting legitimate research. Karyopharm is also obligated to protect the rights and privacy of trial participants and, as such, has a procedure in place for evaluating and fulfilling requests for sharing company clinical trial data with qualified external scientific researchers, as informed by industry guidelines. Data requestors can submit a proposal to medicalinformation@karyopharm.com.
References
- 1. Pemmaraju N., Verstovsek S., Mesa R., et al., “Defining Disease Modification in Myelofibrosis in the Era of Targeted Therapy,” Cancer 128 (2022): 2420–2432, 10.1002/cncr.34205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. O'Sullivan J. M. and Harrison C. N., “JAK‐STAT Signaling in the Therapeutic Landscape of Myeloproliferative Neoplasms,” Molecular and Cellular Endocrinology 451 (2017): 71–79. [DOI] [PubMed] [Google Scholar]
- 3. Tefferi A., “Primary Myelofibrosis: 2023 Update on Diagnosis, Risk‐Stratification, and Management,” American Journal of Hematology 98 (2023): 801–821, 10.1002/ajh.26857. [DOI] [PubMed] [Google Scholar]
- 4. Dunbar A. J., Kim D., Lu M., et al., “CXCL8/CXCR2 Signaling Mediates Bone Marrow Fibrosis and Is a Therapeutic Target in Myelofibrosis,” Blood 141 (2023): 2508–2519, 10.1182/BLOOD.2022015418. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Fisher D. A. C., Fowles J. S., Zhou A., and Oh S. T., “Inflammatory Pathophysiology as a Contributor to Myeloproliferative Neoplasms,” Frontiers in Immunology 12 (2021): 683401, 10.3389/fimmu.2021.683401. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Tefferi A., “Myelofibrosis With Myeloid Metaplasia,” New England Journal of Medicine 342 (2000): 1255–1265, 10.1056/NEJM200004273421706. [DOI] [PubMed] [Google Scholar]
- 7. Hernández‐Boluda J.‐C., Correa J.‐G., Alvarez‐Larrán A., et al., “Clinical Characteristics, Prognosis and Treatment of Myelofibrosis Patients With Severe Thrombocytopenia,” British Journal of Haematology 181 (2018): 397–400, 10.1111/bjh.14601. [DOI] [PubMed] [Google Scholar]
- 8. Kc D., Falchi L., and Verstovsek S., “The Underappreciated Risk of Thrombosis and Bleeding in Patients With Myelofibrosis: A Review,” Annals of Hematology 96 (2017): 1595–1604, 10.1007/s00277-017-3099-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Polverelli N., Breccia M., Benevolo G., et al., “Risk Factors for Infections in Myelofibrosis: Role of Disease Status and Treatment. A Multicenter Study of 507 Patients,” American Journal of Hematology 92 (2017): 37–41, 10.1002/ajh.24572. [DOI] [PubMed] [Google Scholar]
- 10. Stuckey R., Díaz A. S., and MT G.‐C., “Myelofibrosis: Treatment Options After Ruxolitinib Failure,” Current Oncology 32 (2025): 339, 10.3390/CURRONCOL32060339. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Harrison C. N., Kiladjian J.‐J., Koschmieder S., and Passamonti F., “Myelofibrosis: Current Unmet Needs, Emerging Treatments, and Future Perspectives,” Cancer 130 (2024): 2091–2097, 10.1002/cncr.35244. [DOI] [PubMed] [Google Scholar]
- 12. Pardanani A. and Tefferi A., “How I Treat Myelofibrosis After Failure of JAK Inhibitors,” Blood 132 (2018): 492–500, 10.1182/blood-2018-02-785923. [DOI] [PubMed] [Google Scholar]
- 13. Scherber R. M. and Mesa R. A., “Management of Challenging Myelofibrosis After JAK Inhibitor Failure and/or Progression,” Blood Reviews 42 (2020): 100716, 10.1016/j.blre.2020.100716. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Palandri F., Breccia M., Bonifacio M., et al., “Life After Ruxolitinib: Reasons for Discontinuation, Impact of Disease Phase, and Outcomes in 218 Patients With Myelofibrosis,” Cancer 126 (2020): 1243–1252, 10.1002/cncr.32664. [DOI] [PubMed] [Google Scholar]
- 15. Maffioli M., “Making a Case for Disease‐Modifying Agents in Myelofibrosis,” Lancet Haematology 9 (2022): e391–e392, 10.1016/S2352-3026(22)00140-5. [DOI] [PubMed] [Google Scholar]
- 16. Vachhani P., Verstovsek S., and Bose P., “Disease Modification in Myelofibrosis: An Elusive Goal?,” Journal of Clinical Oncology 40 (2022): 1147–1154, 10.1200/JCO.21.02246. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Pemmaraju N., Mead A. J., Somervaille T. C. P., et al., “Transform‐1: A Randomized, Double‐Blind, Placebo‐Controlled, Multicenter, International Phase 3 Study of Navitoclax in Combination With Ruxolitinib Versus Ruxolitinib Plus Placebo in Patients With Untreated Myelofibrosis,” Blood 142 (2023): 620–620, 10.1182/blood-2023-173509. [DOI] [Google Scholar]
- 18. Rampal R. K., Grosicki S., Chraniuk D., et al., “Pelabresib Plus Ruxolitinib for JAK Inhibitor‐Naive Myelofibrosis: A Randomized Phase 3 Trial,” Nature Medicine 31 (2025): 1531–1538, 10.1038/s41591-025-03572-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Azizian N. G. and Li Y., “XPO1‐Dependent Nuclear Export as a Target for Cancer Therapy,” Journal of Hematology & Oncology 13 (2020): 1–9, 10.1186/s13045-020-00903-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Yan D., Pomicter A. D., Tantravahi S., et al., “Nuclear–Cytoplasmic Transport Is a Therapeutic Target in Myelofibrosis,” Clinical Cancer Research 25 (2019): 2323–2335, 10.1158/1078-0432.CCR-18-0959. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Lu M., Mia M. D. B., Xia L., Ellero A., Walker C. J., and Hoffman R., “Selinexor Depletes Ruxolitinib Refractory Myelofibrosis Hematopoietic Stem Cells by Inducing Apoptosis and Blunting the Pro‐Inflammatory Milieu,” Blood 144 (2024): 1377–1377, 10.1182/blood-2024-205114. [DOI] [Google Scholar]
- 22. Lu M., Xia L., and Hoffman R., “Use of Combination Therapies Including the XPO1 Inhibitor Selinexor in Is a Potential Effective Therapeutic Strategy to Treat Myelofibrosis Patients,” Blood 142 (2023): 1792–1792, 10.1182/blood-2023-186674. [DOI] [Google Scholar]
- 23. Maloof M. E. E., Wu S., Ellero A., Walker C. J., Kye S., and Taverna P., “Activity of Selinexor as a Single Agent and Synergistic Activity With Approved/Investigational Myelofibrosis Therapies In Vitro,” Blood 142 (2023): 6353, 10.1182/blood-2023-180134. [DOI] [Google Scholar]
- 24. Michiels J. J., Tevet M., Trifa A., et al., “2016 WHO Clinical Molecular and Pathological Criteria for Classification and Staging of Myeloproliferative Neoplasms (MPN) Caused by MPN Driver Mutations in the JAK2, MPL and CALR Genes in the Context of New 2016 WHO Classification: Prognostic and Therapeutic Implications,” Maedica 11 (2016): 5. [PMC free article] [PubMed] [Google Scholar]
- 25. Passamonti F., Harrison C. N., Mesa R. A., Kiladjian J.‐J., Vannucchi A. M., and Verstovsek S., “Anemia in Myelofibrosis: Current and Emerging Treatment Options,” Critical Reviews in Oncology/Hematology 180 (2022): 103862, 10.1016/j.critrevonc.2022.103862. [DOI] [PubMed] [Google Scholar]
- 26. Pozzi G., Carubbi C., Cerreto G. M., et al., “Functionally Relevant Cytokine/Receptor Axes in Myelofibrosis,” Biomedicines 11 (2023): 2462, 10.3390/biomedicines11092462. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Fisher D. A. C., Miner C. A., Engle E. K., et al., “Cytokine Production in Myelofibrosis Exhibits Differential Responsiveness to JAK‐STAT, MAP Kinase, And NFκB Signaling,” Leukemia 33 (2019): 1978, 10.1038/S41375-019-0379-Y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Masselli E., Pozzi G., Gobbi G., et al., “Cytokine Profiling in Myeloproliferative Neoplasms: Overview on Phenotype Correlation, Outcome Prediction, and Role of Genetic Variants,” Cells 9 (2020): 2136, 10.3390/cells9092136. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Tremblay D. and Mesa R., “Addressing Symptom Burden in Myeloproliferative Neoplasms,” Best Practice & Research Clinical Haematology 35 (2022): 101372, 10.1016/j.beha.2022.101372. [DOI] [PubMed] [Google Scholar]
- 30. World Medical Association, “World Medical Association Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects,” JAMA 310 (2013): 2191–2194. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting File 1: jha270301‐sup‐0001‐SuppMat.docx
Data Availability Statement
Karyopharm Therapeutics is committed to providing qualified scientific and medical researchers with access to anonymized data and protocols from the company's clinical trials for the purpose of conducting legitimate research. Karyopharm is also obligated to protect the rights and privacy of trial participants and, as such, has a procedure in place for evaluating and fulfilling requests for sharing company clinical trial data with qualified external scientific researchers, as informed by industry guidelines. Data requestors can submit a proposal to medicalinformation@karyopharm.com.