BackgroundPrimary myelofibrosis [PMF] is a myeloproliferative neoplasm associated with reduced overall survival (OS). Management strategies for PMF have evolved over the last two decades, including approval of ruxolitinib as the first Janus kinase 1 (JAK1)/JAK2 inhibitor for patients with intermediate or high-risk myelofibrosis. This study assessed changes in mortality before and after ruxolitinib approval, independent of ruxolitinib treatment.MethodsThis retrospective study investigated mortality trends among US veterans with PMF in 2 time periods, pre-ruxolitinib approval (01/01/2007–12/31/2010) and post-ruxolitinib approval (01/01/2015–09/30/2018). Deidentified patient-level data were extracted from US Veterans Health Administration (VHA) databases using PMF diagnosis codes; index was the first PMF diagnosis date. The analysis included adults with ≥2 PMF claims during the analysis periods who were continuously enrolled in the VHA plan 1 calendar year prior to and 6 months post-index and had ≥1 available International Prognostic Scoring System (IPSS) risk factor (available factors were age > 65, hemoglobin < 10 g/dL, and white blood cell count > 25 × 109/L; each counted as one point). Patients with ≥1 MF diagnosis for 12 months before the index period were excluded. Ruxolitinib treatment was not a requirement to be included in the post-ruxolitinib approval cohort. Mortality rates and OS were estimated using the Kaplan-Meier approach; all-cause mortality hazard ratio was estimated using univariate Cox regression.ResultsThe pre- and post-ruxolitinib approval cohorts included 193 and 974 patients, respectively, of which 80 and 197 had ≥2 IPSS risk factors. Ruxolitinib use in the post-ruxolitinib cohort was 8.5% (83/974). At end of follow-up, median (95% CI) OS was significantly shorter in the pre-ruxolitinib cohort (1.7 [1.2–2.6] years vs not reached [3.4–not reached]; P < 0.001). Overall mortality rates for the pre- versus post-ruxolitinib approval cohorts were 79.8% versus 47.3%, respectively, and overall risk of death was 53% lower in the post-ruxolitinib period (hazard ratio, 0.47; 95% CI, 0.37–0.58; P < 0.001). Mortality rates were lower among patients with < 2 vs ≥2 IPSS risk factors.ConclusionsAlthough veterans with PMF have high overall mortality rates, and results in this population might not be generalizable to the overall population, there was a significant lowering of mortality rate in the post-ruxolitinib period.