Real-World Outcomes with Fedratinib in Myelofibrosis Patients Post-Ruxolitinib Discontinuation

Background

Dysregulation of the Janus kinase 2 (JAK2) hematopoiesis-signaling pathway in myelofibrosis (MF) disrupts bone marrow production of blood cells, leading to anemia, fatigue, and splenomegaly. Ruxolitinib (RUX), a dual JAK1/JAK2 inhibitor, was the first drug approved in the USA for intermediate/high-risk MF. However, many patients only have a partial response or develop cytopenia, ultimately discontinuing treatment. Fedratinib (FEDR), approved in August 2019, is active against wild-type and mutationally activated JAK2 and receptor tyrosine kinase FLT3, but its real-world effectiveness in patients discontinuing RUX had not been evaluated.

Aims

To compare baseline characteristics and survival outcomes for patients with MF receiving FEDR versus those not receiving FEDR after discontinuing RUX in routine US clinical practice.

Methods

Patients receiving RUX for primary MF were identified using Flatiron Health’s nationwide electronic health record-derived database. Eligible patients were those with ≥2 recorded visits in the database between Jan 1, 2011, and Oct 31, 2020, ≥18 years of age at the index date, with data ≥1 month before and after index, and no record of receiving unclassified clinical study drugs prior to index. Patients were stratified by treatment with FEDR post-RUX (FEDR and non-FEDR group); the non-FEDR group was further stratified by time of RUX discontinuation to enable a contemporaneous comparison with the FEDR group. Demographics and clinical characteristics were assessed at index. Overall survival (OS) was defined as time from index until death or censoring and assessed by Kaplan–Meier analysis; landmark survival was defined as the proportion of patients who survived at a given point. Associations of baseline variables and survival were assessed by Cox proportional hazards model.

Results

A total of 229 patients were evaluated (FEDR group: n=70; non-FEDR group: n=159). Median age at index was 71.0 for the FEDR group and 70.0 years for the non-FEDR group. Baseline demographic characteristics were broadly similar for the 2 groups. Median follow-up from index in the FEDR and non-FEDR group was 7.0 and 6.0 months, respectively. At index, 90.2% (46/51) and 74.3% (84/113) of patients had an ECOG performance status (PS) score of 0–1, respectively. Median duration of FEDR therapy in the FEDR group was 3.7 (range, 0–12.2) months, and 47.1% (33/70) received 400 mg FEDR once/day. Median OS was not reached in the FEDR group (vs 17 months in the non-FEDR group). Landmark survival in the FEDR and non-FEDR groups was 71.6% and 53.5% at 12 months, respectively. In the non-FEDR subgroup B, the corresponding survival rate at 12 months was 47.9%. The Cox proportional hazards model suggested that being male, of ‘other’ race, and having Charlson comorbidity index ≥1, ECOG PS 2–4, and body mass index <18.5 kg/m2 were significantly associated with poorer survival. Based on the risk-adjusted analysis, we witnessed an apparent trend toward increased OS with FEDR, but the observed differences did not achieve statistical significance (HR [95% CI]: 0.6 [0.3–1.2]; P=0.1466).