Real-world treatment duration of ruxolitinib and use of transfusion among 2268 patients with myelofibrosis: An analysis of the Medicare fee-for-service claims database

Abstract Background: Ruxolitinib (RUX) is approved by the US Food and Drug Administration for the treatment of adults with intermediate- or high-risk myelofibrosis (MF). Median RUX treatment duration in the COMFORT-I/II trials was approximately 3 years and extended up to 5 years for some patients (pts). This study examined real-world duration of RUX treatment for MF, including subgroups with anemia and those requiring transfusions, based on a retrospective long-term follow-up in a large cohort of pts. Methods:This retrospective analysis used Medicare Fee-For-Service Parts A/B claims data and Part D Prescription Drug Event data to identify adults with ≥2 claims for MF from January 2016 to December 2023 and ≥1 claim for RUX from January 2017 to December 2021 (index date was the first RUX claim on or after the first MF claim). Two anemia subgroups were evaluated based on either ≥1 anemia diagnosis (Dx) claim or ≥1 transfusion claim within 3 mo before index. Eligible pts were required to have continuous enrollment in Medicare Parts A/B/D for ≥12 mo before (baseline period) and ≥24 mo after index (follow-up period). Pts with acute myeloid leukemia or myelodysplastic syndrome before index and those with hematopoietic cell transplant during the study period were excluded. Duration of RUX treatment was analyzed using the Kaplan-Meier method; clinical characteristics, MF-related treatments before and after RUX, and RUX doses were assessed descriptively. Results:Among 2268 eligible pts, median (IQR) age was 75 (70–80) years, 1210 (53%) were female, and 2019 (89%) were White. Nearly all pts (96%; n=2175) were aged ≥65 years and therefore intermediate to high risk. The most common comorbidities were renal disease (n=667 29%), peripheral vascular disease (n=557 25%), chronic pulmonary disease (n=552 24%), diabetes without complication (n=523 23%), and congestive heart disease (n=448 20%). Pts in the subgroups (anemia Dx, n=1129; transfusion, n=387) had intermediate-2 or high-risk MF and other baseline characteristics similar to the overall population. Median (IQR) follow-up duration was 46 (33–63) mo. Median (IQR) time from MF Dx to RUX initiation was 4.5 (1–12) mo, with 1577 (70%) pts starting RUX within 12 mo of Dx. Most pts were treatment naive (n=1564 69%); 659 ( hydroxyurea (HU) prior to RUX. During the baseline period, supportive medication for anemia was received by 277 (12%) pts overall (anemia Dx, n=197 17%; transfusion, n=102 26%), most commonly erythropoiesis-stimulating agents (n=251 11%) or danazol (n=43 2%). In the 90 days prior to RUX initiation, 384 (17%) had ≥1 blood transfusion. The mean (SD) daily RUX dose was 23 (11) mg at initiation and 21 (11) mg at the last recorded dose. Overall, 2061 (91%) pts had a dose change during follow-up (mean SD daily dose change, −4.2 14.7 mg), with a median (IQR) time to first dose change of 4 (2–14) mo; 1121 (49%) pts had a dose reduction and 940 (41%) had a dose increase. Among the 384 pts who required transfusion prior to index, 70% (n=267) received transfusions during months 0–3 and 52% (n=201) during months 10–12 post-index. Overall, 807 (36%) pts remained on RUX at the end of follow-up. Median (95% CI) RUX treatment duration was 37 (36–38) mo (subgroups: anemia Dx, 34 32–36 mo; transfusion, 30 28–33 mo). After RUX discontinuation, 25% received another MF treatment; the most common treatments were HU (n=256 11%), fedratinib (n=186 8%), and pacritinib (n=115 5%).Conclusions: In this real-world study with long-term follow-up of approximately 4 years, median duration of RUX treatment was 3.1 years. On average, pts started RUX early after MF diagnosis (median, 4.5 mo), and approximately one-third remained on RUX treatment at end of study period, regardless of age or baseline anemia status. Most pts did not require transfusions at baseline. Among those who did require transfusions prior to RUX initiation, transfusion rates declined following RUX treatment, with nearly half becoming transfusion-free by the end of 1-year follow-up. These findings are consistent with data from the COMFORT trials and suggest that pts with MF can be treated long-term with RUX, including those with anemia and/or transfusion requirement, when managed with appropriate dose-optimizing strategies.