Abstract
Introduction: Myelofibrosis (MF) is a Philadelphia chromosome-negative myeloproliferative neoplasm with an estimated prevalence of 4-6/100,000 persons in the USA. Patients with MF experience aberrant hematopoiesis, bone marrow fibrosis, splenomegaly, and cytopenias, including thrombocytopenia, as well as diminished survival ranging from months to years, depending on risk status. Symptoms progressively worsen, underscoring the need for effective treatments across heterogeneous risk statuses and patient dispositions. Among intermediate- to high-risk patients, Janus kinase 2 (JAK2) inhibitors are the primary treatment for MF. Fedratinib was approved by the US Food and Drug Administration (FDA) based on the results of the JAKARTA trials (NCT01437787; NCT01523171), and a new drug application for pacritinib has been submitted to the FDA for the treatment of MF in patients with severe thrombocytopenia. A head-to-head trial of fedratinib and pacritinib for the treatment of MF has not been conducted. Moreover, there is a paucity of evidence evaluating these agents in patients with MF and thrombocytopenia (platelets < 100 × 10 9/L). Therefore, indirect treatment comparisons (ITCs) are required to assess the comparative efficacy of fedratinib and pacritinib in this population.
Methods: A systematic literature review (SLR) was conducted to identify all clinical evidence in patients with MF and thrombocytopenia. Based on the SLR results, the JAKARTA, JAKARTA-2, and PERSIST-2 (NCT01773187) trials were identified as the studies to form the basis of the ITC. A pooled analysis data set was developed based on individual patient-level data from the fedratinib 400-mg arms of the JAKARTA and JAKARTA-2 trials including patients with platelets < 100 × 10 9/L. Published summary data from the pacritinib 200-mg arm of the PERSIST-2 trial served as the comparator. Simulated treatment comparisons (STCs) were used to compare spleen volume reduction (SVR) ≥ 35% while adjusting for mutually reported baseline patient characteristics such as age, sex, Dynamic International Prognostic Scoring System, Eastern Cooperative Oncology Group (ECOG) performance status (PS), JAK2 V617F mutation status, prior ruxolitinib exposure, and laboratory tests. Final adjustment models were selected based on fit statistics and the literature review. Indirect relative risk (RR) was estimated with 95% confidence intervals (CIs) using unanchored naive ITC (unadjusted) and STC (adjusted) methodologies.
The PERSIST-2 trial included patients with baseline platelets < 50 × 10 9/L, but did not report median baseline counts. Therefore, a sensitivity analysis was conducted to evaluate the impact of differential median baseline platelet counts. In the sensitivity analysis, an outcome model was generated in the full pooled JAKARTA trial population irrespective of platelet count, and platelet count was forced into the multivariable adjustment model. Outcomes were simulated at 3 median baseline platelet counts (25, 50, and 75 × 10 9/L) and compared using similar methodology to the main analysis.
Results: The main analysis suggests that fedratinib is associated with a greater proportion of SVR ≥ 35% than pacritinib (Table, A). According to the naive ITC, fedratinib was numerically favored over pacritinib in terms of SVR ≥ 35% (RR, 1.67 [95% CI, 0.94-2.97]). After adjustment for ECOG PS, JAK2 V6127F mutation status, and prior ruxolitinib exposure, fedratinib was statistically favored relative to pacritinib for SVR ≥ 35% (RR, 1.76 [95% CI, 1.00-3.10]).
Results of the platelet-count-adjusted sensitivity analysis showed that a significant difference was observed in favor of fedratinib with respect to SVR ≥ 35% (Table, B). The resulting indirect RRs were similar regardless of simulated median baseline platelet count, which suggest that differential baseline platelet count has a minimal impact on the ITC results.
Conclusion: This analysis used a population-adjusted ITC to assess the comparative efficacy of pacritinib and fedratinib in patients with MF and thrombocytopenia. Following population adjustment, fedratinib was associated with a greater proportion of patients achieving SVR ≥ 35% than pacritinib. In lieu of a head-to-head clinical trial, further real-world evidence studies should be conducted to assess the effectiveness of these treatments in the clinical setting.
Daniele: BMS/Celgene: Consultancy; Purple Squirrel Economics: Current Employment. Abraham: Bristol Myers Squibb: Current Employment. Kee: Bristol Myers Squibb: Current Employment. Tremblay: Cytel: Consultancy; Purple Squirrel Economics: Current Employment. Rose: Bristol Myers Squibb: Current Employment, Current equity holder in publicly-traded company, Current holder of individual stocks in a privately-held company, Current holder of stock options in a privately-held company. McBride: BMS: Current Employment.
