Efficacy and Safety of Fedratinib in Patients with Myelofibrosis and Low Baseline Platelet Counts in the Phase 3 Randomized FREEDOM2 Trial

Introduction Fedratinib (FEDR), a Janus kinase inhibitor (JAKi), demonstrated spleen volume reduction (SVR) and symptom reduction in the phase 3, randomized, open-label FREEDOM2 trial (NCT03952039) vs best available therapy (BAT) in patients (pts) with myelofibrosis (MF) previously treated with ruxolitinib (Harrison CN, et al. Lancet Haematology. In press). Low platelet count (PLT; <100 x 109/L) is associated with inferior survival in MF, and thrombocytopenia can result in JAKi dose reductions/interruptions. Pooled analysis of the JAKARTA (NCT01437787) and JAKARTA2 (NCT01523171) trials indicated efficacy of FEDR in pts with low baseline (BL) PLT (Harrison CN, et al. Br J Haematol 2022) and exploratory FREEDOM2 subgroup analyses showed a trend towards enhanced SVR in pts with low BL PLT. This FREEDOM2 subanalysis further examined the impact of FEDR on platelet levels, efficacy, and safety in pts with low (50-<100 x 109/L) vs high (≥100 x 109/L) BL PLT. Methods Eligible pts aged ≥18 years with intermediate-2 or high-risk MF relapsed/refractory or intolerant to ruxolitinib were randomized 2:1 to receive FEDR 400 mg once daily or BAT (78% ruxolitinib); pts with BL PLT <50 x 109/L were excluded. Pts were treated until intolerance or lack of efficacy and followed up every 3 months until death, withdrawal of consent, loss to follow-up, or study closure. The primary endpoint was ≥35% SVR (SVR35) at the end of cycle 6 (EOC6). Hematology assessments were at screening, days (D) 1 and 15 of cycles (C) 1-3, D1 of each subsequent cycle up to end of treatment, and at 30-day follow-up. Results In the ITT population, 55 pts (FEDR, n=34; BAT, n=21) had low BL PLT and 124 pts (FEDR, n=85; BAT, n=39) had high BL PLT. BL characteristics were generally similar across analysis groups. Median (interquartile range [IQR]) BL PLT (x 109/L) for pts in the FEDR vs BAT arms was 76.5 (64.0-91.0) vs 72.0 (62.0-86.0) in the low PLT group and 170.0 (124.0-317.0) vs 156.0 (128.0-208.0) in the high PLT group. Pts receiving FEDR generally had greater PLT increase from baseline than those receiving BAT. Increases in median (IQR) PLT levels (x 109/L) were observed as early as C1D15 in pts with low BL PLT (FEDR: 92.5 [82.0-138.0], 43% increase; BAT: 80.0 [63.0-100.0], 11% increase); among pts with high BL PLT, median (IQR) PLT levels (x 109/L) at C1D15 were 223.0 (154.0-336.0) in the FEDR arm (19% increase) and 167.5 (129.0-243.0) in the BAT arm (4.5% decrease). There was no significant correlation between change from BL in spleen volume and PLT in the FEDR (Pearson r=0.02; P=0.87) or BAT (Pearson r=-0.22; P=0.29) arms. At data cutoff (May 10, 2023), pts receiving FEDR had significantly higher rates of SVR35 at EOC6 vs BAT with low (47% vs 0%, P=0.0001) and high (35% vs 10%, P=0.0043) PLT. Among evaluable pts, rates of total symptom score reduction ≥50% at EOC6 were also higher among pts treated with FEDR vs BAT in pts with low (38% vs 20%, P=0.2280) and high (36% vs 16%, P=0.0304) PLT. During the first 6 cycles, rates of grade 3/4 hematologic treatment-emergent adverse events (TEAEs) were highest among pts with low BL PLT treated with FEDR compared with other groups (low PLT: FEDR 76% vs BAT 57%; high PLT: FEDR 61% vs BAT 33%). The most common grade 3/4 TEAEs were thrombocytopenia (low PLT: FEDR 29% vs BAT 10%; high PLT: FEDR 11% vs BAT 0%) and anemia (low PLT: FEDR 18% vs BAT 24%; high PLT: FEDR 27% vs BAT 13%). Thrombocytopenia was the most common cause of TEAE-related dose reduction (low PLT: FEDR 18% vs BAT 5%; high PLT: FEDR 2% vs BAT 3%). Treatment discontinuation rates resulting from TEAEs were higher in pts treated with FEDR vs BAT (low PLT: FEDR 12% vs BAT 5%; high PLT: FEDR 9% vs BAT 5%); only 1 pt discontinued fedratinib due to thrombocytopenia (pt with low BL PLT). Conclusions In FREEDOM2, pts treated with FEDR showed early increases in PLT vs BAT, with a greater magnitude of increase in pts with low vs high BL PLT. Increased PLT did not correlate with spleen size change, indicating a potentially beneficial effect of FEDR on thrombopoiesis. SVR35 rates were numerically higher in pts with low vs high BL PLT treated with FEDR. Safety was consistent with previous trials. Together, these data suggest a platelet sparing effect of second-line FEDR vs BAT, and support FEDR as a promising second-line treatment option for pts with MF with low or high BL PLT.