Outcomes Are Similar for Combination Interferon and Ruxolitinib Versus Ruxolitinib in Myelofibrosis: A Propensity-Score Matched Study

Background: Combining interferon (rIFNα), namely peginterferon alpha-2a, with ruxolitinib (rIFNα-RUX) in the treatment (Rx) of myelofibrosis (MF) is thought to enhance clinical benefit and deplete malignant clones. High response rates were observed in two single-arm rIFNα-RUX trials in both rIFNα-refractory and rIFNα-RUX-naïve patients (pts)^1,2. However, it is not known if the combination improves longer-term outcomes, including progression-free survival (PFS) and overall survival (OS). We conducted this single-center, retrospective, propensity-score matched study to compare clinical response, PFS/OS outcomes and adverse events (AEs) between rIFNα-RUX and RUX in pts with MF.

Methods: Medical records of pts with MF treated with rIFNα-RUX were identified and matched to a control group of RUX monotherapy pts using propensity-scores based on the Dynamic International Prognostic Scoring System Plus (DIPSS+) variables of age, hemoglobin (HGB), white blood count (WBC), platelets (PLT) and peripheral blasts (PB) at Rx initiation, as well as sex and MF type (primary vs secondary). Clinical and lab data were collected as previously described³. MF diagnoses met WHO & ICC 2022 criteria^4,5 and clinical response was assessed using the IWG-MRT response criteria⁵. Spleen response (SR) by physical exam, molecular response (MR) in driver mutation allele frequency, and serial bone marrow response (BMR) was assessed as available. OS and PFS were compared using the Kaplan-Meier method. Fisher’s test was used to compare response rates, AEs and discontinuation rates.

Results: 31 MF pts treated with rIFNα-RUX were matched 1:1 to 31 RUX monotherapy pts (total 62). Groups were similar in age (both median of 67 years (yr)), sex, race, HGB, WBC, PLT or PB at the start of Rx, MF type, driver mutation (JAK2 in 77% rIFNα-RUX vs 74% RUX), and DIPSS+ risk (majority intermediate-1: 65% rIFNα-RUX vs 58% RUX). Median follow-up (5.9 yr rIFNα-RUX vs. 6.6 yr RUX, p=0.08) and times from diagnosis to Rx were similar (median 0.9 yr rIFNα-RUX vs 0.9 yr RUX). Of the rIFNα-RUX pts, 16 received rIFNα first to which RUX was added, 13 received RUX first and 2 initiated both simultaneously. rIFNα-RUX was initiated to improve clinical response (n=20), for hypothesized disease-modifying benefit (n=8) or both (n=3). Median RUX doses and durations were comparable (median 1.6 yr rIFNα-RUX vs 1.7 yr RUX). Discontinuation rates were not significantly different (68% rIFNα-RUX vs 58% RUX). The most common reasons for discontinuation were Rx-related AEs (8 vs 9 pts) and lack of response/progression (4 vs 5).

Median PFS was not reached in either cohort, and there was no significant difference in PFS (p=0.78). Median OS was 9.7 yrs, with no significant advantage for combination Rx (p=0.75). Clinical response rates by IWG-MRT response were comparable between groups at all timepoints. Eleven pts (35%) in each group had clinical benefit (clinical improvement, partial or complete response).

There were no significant differences in SR, MR or BMR between groups. Median SR was 33% for rIFNα-RUX (n=21) vs 29% for RUX (n=20). Best MR was similar between Rx groups for the 26 pts with serial molecular testing. There was no significant difference in reticulin grade reduction in the 28 pts with serial marrow biopsies during Rx.

rIFNα-RUX was well tolerated, with no significant differences in AEs compared to RUX. The most common Rx-emergent AEs were constitutional symptoms (55%), anemia (31%) and thrombocytopenia (32%).

Conclusion: This matched, retrospective analysis of 62 MF pts identified no significant differences in OS or PFS between pts treated with rIFNα-RUX compared to RUX. Despite high clinical responses in earlier single-arm studies, our analysis did not find the combination improved clinical response rates at the doses used. AEs and discontinuation rates were comparable, indicating similar efficacy and tolerability. These results suggest that while rIFNα-RUX is safe, its clinical and survival outcomes compared to RUX do not yet justify the financial burden of combination drug costs. Randomized trials are needed to define any advantages of rIFNα-RUX to monotherapy. Until then, the combination should be reserved for carefully selected pts.

References:

1) Sørensen et al. Haematologica, 2020.

2) Kiladjian et al. Blood, 2022.

3) Abu-Zeinah et al. Leukemia, 2021.

4) Khoury et al. Leukemia, 2022.

5) Arber et al. Blood, 2022

6) Tefferi et al. Blood, 2013.