Selinexor alone and in combination with JAK inhibitors suppresses pro-inflammatory cytokine secretion from primary myelofibrosis cells ex vivo

Abstract Introduction: Selinexor (SEL) is an oral selective inhibitor of the nuclear export protein exportin 1 (XPO1). In combination with ruxolitinib (RUX), SEL has shown rapid, deep, and sustained spleen and symptom responses including disease modifying potential in patients with JAK inhibitor (JAKi)–naïve myelofibrosis (MF), with associated reductions in proinflammatory cytokines (Tantravahi SK, et al. Blood. 2023;142(suppl 1):622). MF is characterized by elevated levels of pro-inflammatory cytokines, which correlate with disease severity. Many of these cytokines, including IL-6 and TNFα, are transcriptionally regulated by NF-κB and activate the JAK/STAT signaling pathway. JAKis used in MF therapy, including RUX, momelotinib (MOM), and pacritinib (PAC), have limited impact on cytokine modulation. SEL inhibits NF-κB activity by preventing nuclear export of its inhibitor IκB. We hypothesize that SEL may attenuate cytokine release in MF by targeting NF-κB signaling, and its combination with JAKi may enhance therapeutic benefit. Methods: NF-κB transcriptional activity was assessed in TNFα-stimulated UKE1 (JAK2-V617F, TP53wt) and ELF-153 (JAK2/CALR/MPL wt and TP53-I251N) cells treated with SEL (250 nM) alone or in combination with JAKi (100 nM). Ex vivo analyses were conducted using viably frozen peripheral blood mononuclear cells (PBMCs) from pretreatment samples of seven patients with JAKi-naïve MF who were being evaluated for participation in the Phase 3 SENTRY trial (NCT04562389). Cells were stimulated with the TLR8 agonist R848 (0.5 µg/mL) to induce cytokine production and incubated for 24 hours with single-agent or combination treatments. Supernatants were collected, and IL-6 and TNFα levels were quantified using immunoassays. Results: SEL significantly reduced NF-κB activity in UKE1 and ELF-153 cells regardless of JAK2 or TP53 mutation status. In contrast, JAKi alone had varied, but minimal, effect on NF-κB transcriptional activity based on their different target profiles. SEL-mediated NF-κB inhibition was maintained in combination with JAKi. We investigated how the ex vivo treatment of PBMCs obtained from patients with MF could affect the production of pro-inflammatory cytokines. The molecular characteristics of the patients included bone marrow fibrosis grade (MF-2: 2/7, MF-3: 2/7, N/A: 3/7), driver mutation (JAK2: 4/7, CALR: 2/7, N/A: 1/7), and high–molecular risk (HMR) mutations (6/7). R848 stimulation of primary MF PBMCs led to a 30.5 ± 44–fold and 2.5 ± 0.8–fold increase in IL-6 and TNFα secretion, respectively, compared with unstimulated control. Single-agent SEL (100–500 nM) inhibited IL-6 secretion (range: -8.7% to -37.9%) in a dose-dependent manner compared to DMSO control. SEL had the highest single-agent activity of the tested compounds and reduced IL-6 levels by -22.1 ± 23.4%. Comparatively, RUX (+7.9 ± 15.7%), MOM (-2.3 ± 24.4%), and PAC (+2.1 ± 15.9%) alone had negligible effects. In combination with JAKi, SEL improved or maintained the inhibition of IL-6 production with RUX (-21.4 ± 27.5%), MOM (-26.4 ± 27.3%), and PAC (-30.4 ± 20.8%). Additionally, TNFα secretion was evaluated in 3 patients and SEL alone again showed the highest inhibition (-30 ± 33.5%) compared to JAKi [RUX (-4.5 ± 30.6%), MOM (-14.2 ± 13.3%), PAC (+14 ± 26.8%)]. The combination of SEL with RUX (-26.9 ± 47.1%) and MOM (-17.3 ± 22.3%) showed sustained TNFα inhibition, while combination with PAC showed moderate inhibition (-3.2 ± 16.9%). Interestingly for a high–molecular risk TP53mut patient, neither single–agent nor combination treatment showed IL-6 or TNFα suppression, though this phenomenon warrants further evaluation in follow–up studies. Notably, cytokine suppression was limited in patients with baseline IL-6 levels near the normal range. Cells from additional patients with MF are being evaluated and will be included in the presentation. Conclusion: XPO1 inhibition is a potentially fundamental mechanism that addresses key inflammatory and pathobiological features of MF. Here we demonstrate that SEL effectively suppresses ex vivo NF-κB–regulated pro-inflammatory cytokine production in PBMCs derived from patients with MF, both as a single agent and in combination with JAKi, further validating the potential of disease modification. The combinatory activity of XPO1 inhibition and JAK/STAT inhibitors supports the clinical use of SEL plus RUX in JAKi-naïve MF, which is being evaluated in the ongoing Phase 3 SENTRY trial (NCT04562389).