PCN19 Beyond a Binary Endpoint: Longitudinal and Individual Symptom Analyses from the Simplify-1 Study Demonstrate Clinically Comparable Symptomatic Benefit of Momelotinib to Ruxolitinib in JAK Inhibitor Naive Myelofibrosis Patients

Abstract

Clinical trials investigating JAK1/JAK2 inhibitors for myelofibrosis subjects have measured symptom improvement as a minimum 50% reduction in total symptom score (TSS) at the end of a 24-week treatment period. However, this landmark response analysis is based on post-baseline score changes occurring only between Week 21 (W21) through W24 and does not account for variation in baseline TSS. To better understand the clinical relevance of TSS improvement we applied individual item analyses and Mixed-effect Models for Repeated Measures (MMRM) to SIMPLIFY-1, a phase 3 study which randomized 432 intermediate and high risk JAK inhibitor-naive MF patients 1:1 to momelotinib or ruxolitinib. Analyses were conducted in the intention-to-treat (ITT) population and in symptomatic subjects (baseline TSS ≥ 10). The distributions of TSS items were examined at baseline, and shift in scores at W24 (health state shifts) were determined. Odds ratios from generalized estimating equations for improvement on momelotinib relative to ruxolitinib were calculated for each item of the TSS using multiple predictive imputations for missing data. The MMRM compared mean change in TSS from baseline through W24. The meaningful change threshold (MCT) was determined using Patient Global Impression of Change. While the prespecified non-inferiority endpoint for TSS response rate was not met for momelotinib (28%) vs ruxolitinib (42%), item-level health state shifts and responder analyses showed similar improvements for momelotinib and ruxolitinib. No significant differences on any items were seen; odds ratios for each between-group comparison ranged from 0.74 to 1.20. MMRM mean TSS change score at W24 was 6.35 (momelotinib) vs 7.87 (ruxolitinib) in the ITT and 8.80 (momelotinib) vs 10.46 (ruxolitinib) in the symptomatic subset. TSS improvements were near the within-subject MCT of 8 points in the ITT and exceeded the MCT in the symptomatic subset. Item analyses and MMRM showed clinically important and comparable improvements in TSS.