A Retrospective Analysis of Safety and Efficacy of Ruxolitinib in CALR-Positive Patients with Myelofibrosis

BACKGROUND

In the phase 3 COMFORT-II study, ruxolitinib, a potent JAK1/JAK2 inhibitor, demonstrated durable reductions in splenomegaly and symptoms, as well as an overall survival benefit compared with best available therapy (BAT) in patients with myelofibrosis (MF). Clinical responses were seen with ruxolitinib independent of JAK2 mutation status. Recently, somatic mutations in the calreticulin gene (CALR), represented by deletions (type 1; p.L367fs*46 and similar) or insertions (type 2; p.K385fs*47 and similar), were identified in most patients lacking JAK2 mutations. Overexpression of CALR mutations resulted in cytokine-independent proliferation of Ba/F3 cells and activation of the JAK/STAT pathway. Here we present safety and efficacy data from a retrospective analysis of patients with CALR mutations (CALR+) in COMFORT-II.

METHODS

Patients (N = 219) with MF and palpable splenomegaly were assigned (2:1) to receive ruxolitinib or BAT. Patients randomized to BAT could cross over to receive ruxolitinib per protocol. The primary endpoint was the percentage of patients with a ≥ 35% reduction in spleen volume at week 48. Mutation status was determined using Sanger sequencing, and DNA fragment analysis by capillary electrophoresis was used to measure allele burden.

RESULTS

Of 166 patients with available baseline mutation status assessments (data cutoff, 01 Dec 2012), 29 (17.5%) were CALR+ (ruxolitinib, n = 20; BAT, n = 9); 18 (62.0%) had type 1 mutations, 6 (21.0%) had type 2, and 5 (17%) had other mutations. Most of these patients had primary MF (ruxolitinib, 70%; BAT, 67%), and the rest had post–essential thrombocytopenia MF; 70% in the ruxolitinib arm and 100% in the BAT arm were classified as high risk. Baseline median spleen volume was 2355 cm³ (range, 459.5-5244 cm³) in the ruxolitinib arm and 2543 cm³ (range, 1131-4527 cm³) in the BAT arm. After 3 years of treatment, 30% of ruxolitinib-randomized patients (6/20) remained on treatment. In comparison, no patients continued to receive BAT; 8 of 9 BAT-randomized patients (89%) crossed over to receive ruxolitinib. Of the 8 patients who crossed over to receive ruxolitinib, 6 patients discontinued, mostly due to disease progression and adverse events.

Response rates were consistent with those in the overall COMFORT-II study; the primary endpoint was achieved by 20% of CALR+ patients in the ruxolitinib arm compared with no patients in the BAT arm. The majority of patients (16/18) in the ruxolitinib arm had decreases in spleen volume at week 48 (median, −24.5% [range, −62.3% to +10.2%]), whereas a majority of those in the BAT arm had increases (3/4; median, 14.4% [range, −17.7% to +18.9%]). Overall, there were 5 deaths (25%) in the ruxolitinib arm and 3 deaths (33%) in the BAT arm. The Kaplan-Meier–estimated probability of survival at 144 weeks was 0.76 (range, 0.49-0.90) in the ruxolitinib arm vs 0.50 (range, 0.11-0.80) in the BAT arm.

Consistent with the overall COMFORT-II study, anemia and thrombocytopenia were common in CALR+ patients receiving ruxolitinib. In the first 48 weeks of treatment, 66.7% of patients (10/15) in the ruxolitinib arm developed anemia compared with 25% in the BAT arm (1/4); 33.3% (5/15) and 25% (1/4) of ruxolitinib and BAT patients developed thrombocytopenia. Analyses on the effect of ruxolitinib on CALR allele burden will also be presented.

CONCLUSION

In this small cohort of CALR+ patients, ruxolitinib showed an efficacy and safety profile consistent with that seen in the overall COMFORT-II population. Ruxolitinib was associated with meaningful decreases in splenomegaly and an improvement in overall survival compared with BAT. Additionally, these results show an adverse event profile similar to what has been seen previously in the overall COMFORT-II population. Taken together, these findings suggest that patients with mutations in CALR identified in COMFORT-II showed a similar response and tolerability profile with ruxolitinib treatment as patients with other mutations, including JAK2 V617F.

^a In patients with assessments at baseline and week 48.