Selinexor for myelofibrosis or a drug in search of a disease?

Abstract

Selinexor for the Treatment of Myelofibrosis. SVR35- spleen volume reduction ≥ 35%, TSS- total symptom score.

Selinexor is an oral selective inhibitor of nuclear export (SINE) that targets exportin-1 (XPO1), a key regulator of nuclear-cytoplasmic transport [1]. Inhibition of XPO1 results in nuclear entrapment of tumor suppressor proteins and growth regulators, resulting in their reactivation and induction of apoptosis in malignant cells [1]. Such a mechanism of action suggests the potential for non-specific therapeutic activity that extends across a broad range of hematologic malignancies. The drug is currently approved for the treatment of multiple myeloma and was previously approved for relapsed/refractory diffuse large B-cell lymphoma, although this indication has since been withdrawn [2, 3]. Preclinical studies have also suggested a therapeutic role in myelofibrosis (MF) by identifying XPO1 as a therapeutic vulnerability in JAK2V617F-mutant cells by selectively suppressing clonal cells while sparing normal progenitors and demonstrating activity in murine models [4, 5].

What are the current unmet needs in MF?

MF is associated with considerable symptom burden, impaired quality of life, and shortened survival. In a Mayo Clinic series of 1,282 patients with primary myelofibrosis (PMF), splenomegaly, transfusion-dependent anemia, and constitutional symptoms were documented at the time of referral in 72%, 33%, and 29% of patients, respectively [6–8]. In addition, leukemic transformation occurred in 9% of patients after a median follow-up of 3.2 years, and median survival was 4.4 years [8]. These observations underline the need for therapies that not only alleviate symptoms but also alter the natural history of the disease. Between 2011 and 2023, four JAK2 inhibitors (JAKi) received approval by the Food and Drug Administration (FDA) for the treatment of MF. However, none of these drugs has consistently demonstrated disease modification or a clear overall survival benefit [6]. To that effect, a Mayo Clinic study of 183 patients with MF treated on clinical trials found no survival differences among various JAKi, with allogeneic hematopoietic stem cell transplantation (AHSCT) remaining the only intervention associated with improved survival [9].

On the other hand, ruxolitinib, the first approved JAKi, demonstrated meaningful improvements in splenomegaly and constitutional symptoms [10, 11]. Toxicities included exacerbation of anemia and thrombocytopenia. Fedratinib, the second JAKi approved for use in MF, shares a similar toxicity profile [12]. More recent approvals include pacritinib for use in patients with severe thrombocytopenia (platelet count < 50 ×10⁹/l) [13], and momelotinib for use in the setting of associated anemia, through its ACVR1/ALK2 inhibitory activity [14, 15]. Taken together, it is reasonable to list the lack of disease-modifying activity and suboptimal control of anemia as the current unmet needs in MF. Accordingly, drug development in MF has recently expanded beyond JAK–STAT signaling to target alternative pathways, including XPO1, bromodomain, extraterminal domain (BET) proteins, BCL-xL, and the BMP6/SMAD pathway [16]. Preliminary reports from ongoing clinical trials were most encouraging with the latter approach, with several anemia-directed therapies, including DISC-0974 (selcodebart; an anti-hemojuvelin antibody), elritercept, and luspatercept (TGF-β ligand traps) [17–20].

Where does selinexor fit in addressing currently unmet needs in MF?

In this issue of Blood Cancer Journal, Deininger et al, report the results of an investigator-initiated phase 2 study evaluating selinexor 40 mg (n = 5), 60 mg (n = 6) or 80 mg (n = 6) once-weekly in 17 patients with JAKi–refractory or intolerant MF, including 88% with JAKi-refractory disease and 59% with high-risk molecular mutations [21, 22], Among 11 evaluable patients at week 24, spleen response, defined as ≥35% spleen volume reduction (SVR35) was achieved in 27%, while 45% achieved SVR ≥ 25% [21, 22]. Notably, none of the patients initiated on selinexor 40 mg weekly achieved SVR35 [22]. Among 8 patients with non-transfusion-dependent anemia, three (38%) achieved major and one (13%) minor response [22]. Treatment was associated with frequent nausea, fatigue, anorexia, diarrhea, and a 100% discontinuation rate [22].

Subsequently, a multicenter phase 1 study evaluated selinexor in combination with ruxolitinib in 24 JAKi–naïve patients with MF [23]. In the recommended phase 3 dose cohort (selinexor 60 mg weekly), SVR35 and symptom response rates at week 24 were 79% and 58%, respectively [23]. While limited by its single-arm design, the encouraging efficacy signals observed with combination therapy provided the rationale for the randomized phase 3 SENTRY trial, in which 353 JAKi–naïve patients with MF were randomized to receive selinexor plus ruxolitinib (n = 235) or placebo plus ruxolitinib (n = 118) [24]. Baseline characteristics reflected a relatively favorable-risk patient population, with 56% of patients classified as Dynamic International Prognostic Scoring System (DIPSS) intermediate-1 risk. Approximately one-third of the study patients harbored high molecular risk mutations (ASXL1, EZH2, IDH1/2, SRSF2, or U2AF1), 6% displayed transfusion-dependent anemia, and roughly a quarter had a hemoglobin level <10 g/dL [24]. Notably, patients in both treatment arms received median ruxolitinib doses below the optimal dose of 20–25 mg twice daily (23 and 29.9 mg/day). The study met its primary endpoint, with SVR35 at week 24 achieved in 49.8% of patients receiving selinexor plus ruxolitinib compared with 28% receiving placebo plus ruxolitinib. However, the median treatment duration was only 11 months, limiting accurate assessment of durability of spleen response [24]. For context, SVR35 rates with single-agent ruxolitinib were 41.9% at week 24 in COMFORT-I (ruxolitinib vs placebo) [10], and 28% at week 48 in COMFORT-II studies (ruxolitinib vs best available therapy) [11].

Importantly, the superior spleen responses achieved with selinexor in combination with ruxolitinib were accompanied by greater toxicity, potentially explaining the lack of additional symptom benefit. Grade ≥3 adverse events occurred in 70% of patients receiving selinexor plus ruxolitinib compared with 50% in the placebo-ruxolitinib arm [24]. Nausea was reported in more than half of treated patients (57% vs 17%) despite dual antiemetic prophylaxis [24]. Furthermore, thrombocytopenia (59% vs 43%) and neutropenia (27% vs 9%) were notably more frequent with selinexor [24]. The co-primary symptom endpoint was not met, with similar improvements in total symptom score (TSS) in both arms (−9.9 vs −10.9), and no anemia benefit was observed [24]. Although an early overall survival signal was reported, the analysis was based on a limited number of events with ~12 months of follow-up and was not risk-adjusted [24]. Furthermore, separation of the survival curves appeared to diminish beyond 21 months [24], underscoring the need for longer follow-up before drawing conclusions regarding any survival advantage. Taken together, selinexor plus ruxolitinib improved spleen responses but failed to improve symptoms, while increasing toxicity and providing no convincing evidence of disease modification.

Where do we go from here?

The above observations underline a central conundrum in clinical trials in MF, which primarily focus on spleen volume reduction and symptom improvement, endpoints that have driven regulatory approval of JAKi but do not fully capture outcomes most relevant to patients, while also overlooking the financial implications of new therapies [25]. Results of the SENTRY study are no different from those reported in the MANIFEST-2 [26] (pelabresib plus ruxolitinib) and TRANSFORM-1 [27] phase 3 trials (navitoclax plus ruxolitinib) in JAKi naïve patients with MF; SVR35 at week 24 was significantly higher with combination treatment than placebo-ruxolitinib (65.9% vs 35.2%, and 63.2% vs 31.5%, respectively); however, symptom responses were not significantly improved (52.3% vs 46.3%, and 39.2% vs 41.7%, respectively) and neither combination improved anemia.

Of importance to all three above-mentioned ruxolitinib add-on clinical trials, it should be recognized that effective control of splenomegaly in routine clinical practice can often be achieved through optimal ruxolitinib dosing (20 to 25 mg twice daily) or by switching to an alternative JAKi. In other words, it is currently uncertain whether an ruxolitinib add-on treatment strategy would perform better than a control arm with higher doses of JAKi. Regardless, whether or not a higher spleen response rate translates into improved survival remains uncertain, and the potential value in improved quality of life is undermined by increased toxicity. On the other hand, given the association between marked splenomegaly and adverse post-transplant outcomes, effective pre-transplant spleen reduction remains an important therapeutic objective, and this might be one area where the above-discussed combination drug strategy might come in handy. To that end, a formal clinical trial designed to determine value in the pre-transplant setting might provide practically more relevant information.

Author contributions

NG wrote the paper. AT reviewed and edited the paper.

Competing interests

NG has served on the Advisory Board for Agios and DISC Medicine.