(C-175) Myelofibrosis Patients Treated with Momelotinib vs. Best Available Therapy: Evaluating the Use of External Comparator Patients from the Optum Electronic Health Records Database.

Background: Myelofibrosis (MF) is a rare blood cancer most often observed in older adults. Momelotinib (MMB), a Janus kinase inhibitor (JAKi) for the treatment of intermediate and high-risk MF, may improve anemia responses and alleviate symptoms compared to ruxolitinib (RUX) or other best available therapy (BAT). However, the overall survival (OS) benefits of MMB compared to RUX/BAT are unknown, as phase 3 clinical trials of MMB have used crossover designs that make it difficult to evaluate long-term OS. An external control arm study could extend trial findings by comparing OS among MMB trial patients to external comparator patients sourced from real-world data.

Objectives: To evaluate the comparability in baseline patient characteristics and data availability between published MMB trial cohorts and external comparator cohorts of patients receiving RUX/BAT in routine care.

Methods: Using the Optum Electronic Health Record (EHR) database, we identified 2 MF cohorts: (1) patients treated with RUX intended as external comparators for JAKi naïve/limited trial patients receiving MMB, and (2) patients treated with BAT intended as external comparators for JAKi-experienced patients receiving MMB. Comparators were identified by applying eligibility criteria from relevant clinical trials to Optum EHR data.

Results: Application of trial eligibility criteria to EHR patients resulted in 404 and 293 patients in the JAKi naïve/limited and JAKi-experienced external comparator cohorts, respectively. External comparators had similar distributions of key lab values at baseline as MMB clinical trial patients, based on published data. Comparing the JAKi-naïve/limited external comparator patients receiving RUX within Optum EHR data to JAKi-naïve patients receiving MMB in the SIMPLIFY-1 trial, mean (SD) hemoglobin (g/dL) was 9.5 (1.3) versus 10.6 (2.1) and platelet count (x 109 cells/L) was 264.3 (192.8) versus 301.1 (207.0), respectively. Comparing JAKi-experienced external comparator patients receiving BAT within Optum EHR data to JAKi-experienced patients receiving MMB in the SIMPLIFY-2 and MOMENTUM trials, mean (SD) hemoglobin (g/dL) was 7.9 (1.2) in the external comparators versus 9.4 (1.9) and 8.1 (1.1) in the trials, respectively. Mean (SD) platelet count was 175.4 (194.3) in the external comparators versus 170.8 (148.0) and 151.7 (130.9) in the trials. Missingness of key labs in the Optum cohorts was limited to 5 or fewer patients ( < 2%).

Conclusions: By applying trial eligibility criteria to Optum EHR data, external comparator cohorts were comparable to MF patients from published clinical trials of MMB on key baseline lab values. This demonstrates the feasibility of using Optum EHR data as a source of real-world data in future external comparator studies.