Indirect treatment comparison (ITC) of momelotinib (MMB) vs fedratinib (FED) safety in patients (pts) with myelofibrosis (MF).

Abstract

7065 Background: JAK inhibitors (JAKi) are a mainstay of MF treatment, with proven ability to improve symptoms and spleen volume for most pts; although safety profiles vary, some approved JAKi have associated hematologic toxicity. The JAK1/JAK2/ACVR1 inhibitor MMB has shown consistent benefit in improving anemia, symptoms, and spleen volume in pts with MF. Clinical trials have directly compared MMB with ruxolitinib, with MMB showing a more favorable hematologic toxicity profile. However, there are limited data comparing MMB vs other JAKi such as FED. We performed an ITC of safety outcomes between MMB and FED in pts with MF. Methods: A matching-adjusted indirect comparison (MAIC) was undertaken to compare outcomes from phase 2 and/or 3 MMB and FED trials. Adverse events (AEs) occurring in > 10% of a treatment arm over 24 weeks were evaluated; primary outcomes were grade 3/4 anemia and thrombocytopenia. Separate analyses were performed in JAKi-naive (SIMPLIFY-1 [S1] vs JAKARTA) and JAKi-experienced (pooled SIMPLIFY-2 [S2], MOMENTUM vs JAKARTA2) pts. Evaluable outcomes were limited by published aggregate data, so not all were feasible to compare. MAIC involved reweighting MMB pt data by matching baseline characteristics prioritized by expert clinical input (DIPSS/IPSS, Total Symptom Score, platelet counts, spleen volume/length, and/or hemoglobin levels) between trial arms, and determining odds ratios (ORs), risk ratios, and risk differences (RDs) between outcomes in resulting balanced populations. As nonrandomized groups were compared, results may be biased due to unmeasured/residual confounding, despite adjusting for prognostic factors/effect modifiers. Results: Effective MMB sample sizes after adjustment were 151.1 JAKi-naive and 79.4 JAKi-experienced pts. Risk of any or grade 3/4 anemia, diarrhea, nausea, and serious AEs leading to dose reductions were statistically less likely with MMB vs FED in both populations; any or grade 3/4 thrombocytopenia was also less likely with MMB vs FED, significantly so in JAKi-naive pts. Odds/risk of other outcomes assessed varied in significance (key OR/RD results presented in the Table). No identified outcomes were statistically less likely with FED. Conclusions: MMB showed a favorable safety profile vs FED in both JAKi-naive and experienced pts, including a significantly lower risk of key hematologic AEs, diarrhea, and nausea over 24 weeks. Clinical trial information: NCT04173494 ; NCT01969838 ; NCT02101268 ; NCT01523171 ; NCT01437787 . [Table: see text]