Abstract
INTRODUCTION: In the era before Janus kinase (JAK) inhibitors, cytogenetic information was used to predict survival in myelofibrosis patients. However, the prognostic value of cytogenetics in the setting of JAK inhibitor therapy remains unknown. PATIENTS AND METHODS: We performed a retrospective analysis of 180 patients with bone marrow biopsy-proven myelofibrosis from 3 US academic medical centers. We fit Cox proportional hazards models for overall survival and transformation-free survival on the bases of 3 factors: JAK inhibitor therapy as a time-dependent covariate, dichotomized cytogenetic status (favorable vs. unfavorable), and statistical interaction between the two. The median follow-up time was 37.1 months. RESULTS: Among patients treated with best available therapy, unfavorable cytogenetic status was associated with decreased survival (hazard ratio = 2.31; P = .025). At initiation of JAK inhibitor therapy, unfavorable cytogenetics was (nonsignificantly) associated with increased survival compared to favorable cytogenetics (hazard ratio = 0.292; P = .172). The ratio of hazard ratios was 0.126 (P = .034). These findings were similar after adjusting for standard clinical prognostic factors as well as when measured against transformation-free survival. CONCLUSION: The initiation of JAK inhibitor therapy appears to change the association between cytogenetics and overall survival. There was little difference in survival between treatment types in patients with favorable cytogenetics. However, the use of JAK inhibitor therapy among patients with unfavorable cytogenetics was not associated with worse survival compared to favorable cytogenetics. Our analyses suggest that initiation of JAK inhibitor therapy nullifies the negative prognostic implication of unfavorable cytogenetics established in the pre-JAK inhibitor therapy era.