The activin receptor-like kinase-2 inhibitor zilurgisertib (INCB000928) as monotherapy or with ruxolitinib in patients with anemia due to myelofibrosis: Phase 1/2 study final results

Abstract Background: Hepcidin dysregulation contributes to anemia of chronic inflammation observed in several malignancies, and hepcidin levels are elevated in myelofibrosis (MF). It has been proposed that activin receptor-like kinase-2 (ALK2 [ie, ACVR1]) contributes to MF-associated anemia via hepcidin upregulation and that ALK2 inhibition may improve anemia in patients (pts) with MF. Here we present final results from a phase 1/2 study evaluating ALK2 inhibition as a treatment approach for management of MF-related anemia, including safety, efficacy, pharmacokinetics (PK), and pharmacodynamics (PD) of the potent and selective oral ALK2 inhibitor zilurgisertib. Methods:This open-label, phase 1/2, dose-escalation/expansion study (INCB00928-104/NCT04455841) evaluated zilurgisertib alone (treatment group A [TGA]), as an add-on to the Janus kinase (JAK)1/JAK2 inhibitor ruxolitinib (TGB), or with ruxolitinib in JAK inhibitor–naive pts (TGC). Eligible pts (≥18 years) had intermediate (Int)-1 (TGB, TGC) or Int-2/high-risk (all groups) MF and anemia, and they were transfusion-dependent (TD) or non–transfusion-dependent (NTD) with symptomatic anemia. In TGA, zilurgisertib monotherapy was evaluated at a 50-mg once-daily (qd) starting dose in escalation. In TGB, pts received a stable ruxolitinib regimen for ≥12 wk before initiating zilurgisertib at a 100-mg qd starting dose. In TGC, zilurgisertib and ruxolitinib were initiated concurrently at starting doses of 400 mg qd and 10 mg twice daily (bid), respectively. Ruxolitinib dose increases were permitted after the 28-day dose-limiting toxicity (DLT) period. Zilurgisertib dose escalation was performed independently in each group, with ≤2-fold dose increases evaluated until identification of maximum tolerated dose (MTD) or recommended dose(s) for expansion. Primary endpoints were safety and tolerability, including DLTs and MTD. Secondary endpoints included PK, PD, and anemia response (defined as transfusion independence for any rolling 12-wk period for TD pts, and hemoglobin [Hb] increase ≥1.5 g/dL above baseline sustained for any rolling 12-wk period for NTD pts). Results: 84 pts were enrolled in dose escalation to TGA (n=32), TGB (n=38), or TGC (n=14). Median (range) age was 73.5 (53‒84) years for TGA, 74.5 (54‒85) for TGB, and 70.0 (60–82) for TGC. Baseline median (range) Hb (g/dL) was 7.9 (6.5–9.7), 8.1 (5.1–9.9), and 8.0 (6.1–9.4) for TGA, TGB, and TGC, respectively. The maximum zilurgisertib dose evaluated in each treatment group was 600 mg daily (administered as 600 mg qd or 300 mg bid); MTD was not reached. No pts in TGA or TGC experienced a DLT; 3 pts in TGB had DLTs (grade 1 gastrointestinal hemorrhage and epistaxis, n=1; grade 3 alveolar hemorrhage, n=1; grade 3 diarrhea, n=1). Treatment-emergent adverse events (TEAEs) occurred without apparent dose dependence, and <50% of pts in each group experienced serious TEAEs (TGA, n=11 [34.4%]; TGB, n=17 [44.7%]; TGC, n=5 [35.7%]). Fatal TEAEs occurred in 2 pts in TGA (death, n=1; sepsis, n=1) and 1 pt in TGB (sepsis, n=1). Hepcidin suppression following zilurgisertib dosing was observed in all 3 treatment groups, indicating target engagement; longitudinal PD hepcidin suppression was also observed. Among NTD pts included in the Wk 24 and 48 anemia response analysis (TGA, n=16; TGB, n=22; TGC, n=5), response occurred in 2 pts in TGA, 1 in TGB, and none in TGC at Wk 24 and 3 pts in TGA, 4 in TGB, and 1 in TGC at Wk 48. Among anemia response–evaluable TD pts (TGA, n=9; TGB, n=4; TGC, n=4), anemia response at Wk 24 was not achieved by any pt; 2 pts achieved response by Wk 48 (1 in TGB; 1 in TGC). Median values for the largest increase (g/dL) from baseline in mean Hb values over any rolling 12-wk treatment period through Wk 24 were 0.3, 0.4, and –0.03 for TGA, TGB, and TGC, respectively. Dose expansion was not initiated in any treatment group following analysis of anemia response rates in dose escalation.Conclusions:Zilurgisertib monotherapy or in combination with ruxolitinib had a favorable tolerability profile, with few DLTs and no apparent correlation between dose and TEAE frequency. However, despite evidence of on-target hepcidin suppression, zilurgisertib provided minimal improvements for anemia. These findings highlight the multifactorial nature of anemia in MF and suggest that selective inhibition of ALK2/ACVR1 may not be sufficient to effectively ameliorate MF-related anemia.