Nuvisertib, an oral investigational selective PIM1 kinase inhibitor, showed clinical responses strongly correlating with cytokine modulation in patients with relapsed/refractory myelofibrosis in the ongoing global phase I/II study

Abstract Background: PIM1 is overexpressed in hematologic malignancies, including myelofibrosis (MF), contributing to disease progression by modulating cytokine-driven pathways such as PI3K/AKT and JAK/STAT. Elevated pro-inflammatory cytokines are a hallmark of MF and are closely linked to symptom burden and poor prognosis. Preclinically, PIM1 knockout was shown to prevent MF progression without affecting the platelet (PLT) counts, whereas pan-PIM knockout caused thrombocytopenia (TCP). Nuvisertib (TP-3654), an oral, investigational, highly selective PIM1 kinase inhibitor, alone or in combination with JAK inhibitor ruxolitinib has demonstrated reduction in spleen size, bone marrow (BM) fibrosis, and expression of cytokine response genes in JAK2V617F and MPLW515L MF mouse models. Methods: The ongoing global phase 1/2 study evaluates the safety and efficacy of nuvisertib monotherapy in patients (pts) with MF (NCT04176198, Arm 1). Study population includes primary or secondary MF, previously treated with or ineligible for JAK inhibitor, DIPSS intermediate or high-risk MF, PLT ≥25 x 109/L, splenomegaly (≥450 cm3 by imaging), and ≥2 measurable symptoms per MFSAF v4. The study aims to identify the RP2D of nuvisertib monotherapy and assess the safety, clinical activity (spleen volume reduction [SVR], total symptom score [TSS] improvement), and PK and PD markers (cytokine, BM fibrosis etc.). Results: As of 29 May 2025, total 77 pts enrolled in 5 dose levels of nuvisertib from 480 mg QD to 720 mg BID. At baseline, median age 71 years (49, 85); spleen volume 1988 cm3 (270, 7718); TSS 23 (4, 62); hemoglobin (Hgb) 9.7 g/dL (5.6, 17.2; 52% pts were <10 g/dL; 39% pts required transfusion); PLT 96 x 109/L (24, 816; 51% pts were <100 x 109/L). 75% pts were DIPSS Int-2 or high risk; 41% pts had high molecular risk mutation; and 30% pts received ≥2 prior JAK inhibitors. Median nuvisertib treatment duration was 22 weeks (2, 197), and 16 (21%) pts on active treatment. No DLT occurred. Treatment-related adverse events (TRAEs) occurring in ≥20% of pts were primarily grade 1/2 diarrhea, nausea, and vomiting. Grade ≥3 TRAE occurring in ≥3 pts included TCP (n=8, 7 of 8 pts had baseline TCP). Mean Hgb and PLT remained stable throughout the 24-week treatment. In pts treated with 720 mg BID dose for ≥12 weeks, ≥25% SVR was observed in 4 of 20 pts (20% SVR25 response) and ≥50% reduction in TSS in 9 of 20 pts (44% TSS50 response) at any time. Absolute improvement in symptoms was seen across all 7 parameters. A strong correlation (p<0.001) between cytokine modulation (e.g. ↓ENRAGE, ↓MIP1β, ↓PAI-1, ↓IL-1Ra, and ↑adiponectin) and SVR25, TSS50, and individual symptom improvement were observed. In pts with baseline Hgb <10 g/dL, 6 of 26 (23%) pts showed Hgb response [mean ≥1.0 g/dL Hgb increase for ≥12 weeks without transfusion, including 3 pts with ≥1.5 g/dL Hgb increase]. Hgb responses were also observed in pts with baseline Hgb >10 g/dL. In pts with baseline PLT <100 × 109/L, 8 of 30 (26.7%) pts showed PLT response [≥30×109/L increase maintained ≥4 weeks], and PLT recovered to ≥100 × 109/L in all responders. Modulation of circulating biomarkers was observed in PLT responder pts including increased TN-C, a protein reported to be involved with PLT recovery, and decreased VCAM-1, a marker involved in PLT endothelial cell adhesion and chronic inflammation. 13 of 34 (38.2%) evaluable pts (assessments at baseline and every 24 weeks) showed ≥1 grade reduction in BM fibrosis which correlated with cytokine reduction (e.g. ↓MIP1β, ↓TNFR1); and 11 of 13 pts also showed at least one of SVR25, TSS50, Hgb or PLT responses (5 pts showed dual Hgb and PLT responses). The 1-year overall survival rate following nuvisertib treatment in this heavily pretreated pts with relapsed/refractory (R/R) MF was 81% which also correlated with cytokine modulation (e.g. ↓MIP1β, ↓TNFR1, and ↑FVII). Conclusions: Nuvisertib monotherapy appeared well tolerated with no DLTs. Preliminary data in pts with R/R MF showed that nuvisertib treatment leads to significant modulation of cytokine profiles, demonstrating a strong correlation with clinical responses, including SVR25 and TSS50 responses, and improvements in Hgb, PLT and BM fibrosis, suggesting that selective PIM1 inhibition may offer disease-modification with limited hematologic toxicity. Emerging data supports ongoing clinical development of nuvisertib in combination with ruxolitinib and momelotinib (NCT04176198, Arms 2 and 3, respectively).