Retrospective Real-World Analysis of Survival Outcomes in Patients with Myelofibrosis and New or Worsening Anemia Treated with Ruxolitinib in the United States

Introduction: Ruxolitinib provides symptom, spleen, and overall survival (OS) benefits in patients with myelofibrosis but does not address and may worsen anemia, a key feature of myelofibrosis that negatively impacts OS. In a previous analysis of the phase 3b expanded access study of ruxolitinib (JUMP; Gupta V, et al. ASH 2023. Poster 5174), OS in patients who were not anemic at baseline was significantly longer in those who did not develop new or worsening anemia on treatment vs those who did (hazard ratio [HR], 0.56; P=.01). In baseline anemic patients, no new or worsening anemia on treatment also trended towards longer OS (median, not reached vs 58.3 months; HR, 0.80; P=.24). Although new or worsening anemia on treatment did not impact spleen or symptom responses, it may be associated with worse survival outcomes independent of baseline anemia status. The present analysis explores this observation in the real-world setting by assessing the association between anemia status at treatment initiation, new or worsening anemia on treatment, and OS in a US-based population with myelofibrosis treated with ruxolitinib.

Methods: This was a retrospective analysis of adult patients in the US Flatiron Health electronic health record–derived deidentified database with a myelofibrosis diagnosis and ruxolitinib use between January 1, 2013, and March 1, 2023. OS was measured from the date of ruxolitinib initiation (baseline) until end of data availability or death; patients were required to have ≥3 months of follow-up after treatment initiation. Subgroups were defined by anemia status at baseline (yes/no; based on a hemoglobin level of <10 g/dL or ICD-9/10 diagnostic codes for anemia at index) and new or worsening anemia during the 12 weeks post index (yes/no; based on [1] a hemoglobin level of <10 g/dL with a hemoglobin decrease of ≥1.5 g/dL from baseline, or [2] new transfusion requirement [≥2 units transfused within 8 weeks] at weeks 4, 8, or 12). OS across the subgroups was compared using unadjusted Cox regression analysis.

Results: A total of 286 patients met the criteria for analysis; at baseline, the mean age was 72.0 years, 57% were male, and mean hemoglobin level was 10.3 g/dL. A total of 147 patients were anemic at baseline (71 [48%] with new or worsening anemia in the first 12 weeks of treatment), and 139 were nonanemic at baseline (61 [44%] with new or worsening anemia in the first 12 weeks of treatment).

Compared with OS in baseline nonanemic patients with no new or worsening anemia on treatment, OS was significantly shorter in baseline anemic patients with no worsening anemia on treatment (HR, 1.87; 95% CI, 1.13-3.08; P<.05), baseline nonanemic patients with new or worsening anemia on treatment (HR, 2.04; 95% CI, 1.21-3.45; P<.01), and baseline anemic patients with worsening anemia on treatment (HR, 2.53; 95% CI, 1.53-4.19; P<.001).

In the overall pooled population, new or worsening anemia on treatment was significantly associated with shorter OS regardless of baseline anemia status (HR, 1.68; 95% CI, 1.18-2.40; P<.01). This was most pronounced in patients without anemia at baseline, among whom OS was significantly shorter in those who developed new or worsening anemia on treatment (HR vs patients without new or worsening anemia on treatment, 2.17; 95% CI, 1.27-3.70; P<.01). In patients who were already anemic at baseline, worsening anemia on treatment was also numerically associated with shorter OS (HR vs patients without new or worsening anemia on treatment, 1.30; 95% CI, 0.81-2.08; P=.28).

Conclusion: This retrospective analysis highlights that ruxolitinib-treated patients with myelofibrosis and anemia have poorer survival outcomes compared with nonanemic patients, regardless of whether anemia was diagnosed before or during treatment. These results provide real-world evidence that new or worsening anemia on treatment with ruxolitinib, in both patients who were or were not anemic at treatment initiation, is associated with an increased risk of death, demonstrating the need for alternative therapeutic options that can address anemia without compromising survival.