(SLC-035) Effect of Anemia on Mortality in Janus Kinase Inhibitors Users with Myelofibrosis: A Nationwide Cohort Study

Background: Myelofibrosis (MF) is a rare myeloproliferative neoplasm characterized by fibrosis of bone marrow, developing anemia as its complication. While Janus kinase inhibitors (JAKi) are standard treatment choice for MF, they can also trigger anemia. However, the effect of anemia on all-cause death in routine clinical practice is not fully understood.

Objectives: To assess the association between anemia and the risk of all-cause death among JAKi users with MF.

Methods: We conducted a nationwide, population-based cohort study using the National Health Insurance Service database from 2011 to 2023. New users of JAKi with MF aged ≥20 years, without history of hematopoietic stem cell transplantation (HSCT) were included in the study cohort. The index date was defined as the JAKi start date and their anemic status was assessed 180 days prior to the index date, defined as MF-related anemia. Patients were followed up from the index date until the earliest of all-cause death, HSCT, or study end date. To assess possible effect modifications, subgroup analyses were performed by age, sex, and type of MF (primary and secondary). As an exploratory analysis to evaluate the effect of JAKi-related anemia, defined as incident anemia within 180 days of JAKi initiation, patients without anemia 180 days prior to JAKi initiation were included. The index date of this analysis was defined as 180 days after the JAKi initiation. To ascertain 180 days for anemic status assessment, patients who died or received HSCT within 180 days of JAKi initiation were excluded. Poisson regressions were applied to estimate crude incidence rates (IR) per 1000 person-years. Cox proportional hazard models adjusted with demographic and clinical characteristics were used to estimate the adjusted hazard ratios (aHR) of all-cause death.

Results: Among a total of 2121 JAKi new users, 1217 patients (52.8% older than 70 years; 57.4% male) were anemic and 904 patients (40.0% older than 70 years; 52.1% male) were non-anemic at baseline. During mean follow-up of 2.58 years, IRs per 1000 person-years of all-cause death were 5.42 (95% CI 5.35 – 5.50) and 4.62 (4.48 – 4.75) in anemic and non-anemic patients, with aHR of 1.86 (1.58 – 2.20). In subgroup analyses, we did not find any significant effect modifications. In exploratory analyses for JAKi-related anemia, IRs per 1000 person-years were 5.38 (5.29 – 5.48) and 5.32 (5.04 – 5.60) in incident-anemic and non-anemic patients, with aHR of 1.03 (0.38 – 1.45).

Conclusions: In this nationwide cohort study, anemia at JAKi initiation was an independent risk factor for all-cause death, whereas anemia after JAKi initiation did not associate with increased risk of all-cause death.