Srdan Verstovsek, MD, PhD
Patients with intermediate/high-risk myelofibrosis (MF) had improved survival when treated long term with the Janus kinase (JAK) inhibitor ruxolitinib (Jakafi), pooled data from 2 randomized trials showed.
An intention-to-treat analysis yielded a 5-year survival of 57.5% with ruxolitinib treatment versus 48.5% for patients who received best available treatment (BAT) or placebo. The median overall survival (OS) was 5.3 years with ruxolitinib and 3.8 years with the control treatment.
After correction for crossover to ruxolitinib from the control groups, patients initially assigned to ruxolitinib lived more than twice as long as those assigned to control treatment, according to a presentation at the American Society of Hematology meeting in San Diego.
“Because most patients in the control group eventually crossed over to ruxolitinib, the primary difference between the two groups illustrated the immediate versus delayed treatment effect with ruxolitinib,” Srdan Verstovsek, MD, PhD, chief of myeloproliferative neoplasms at the University of Texas MD Anderson Cancer Center, and colleagues concluded. “These findings suggest that earlier treatment with ruxolitinib may improve the survival advantage for patients with myelofibrosis.”
Both the FDA and the European Medicines Agency approved ruxolitinib for the treatment of intermediate/high-risk MF, including patients with primary MF, post-polycythemia vera MF, and post-essential thrombocythemia MF. Principal support for the approvals came from 2 pivotal phase III trials (COMFORT-I and II). Both trials included patients with intermediate-2/high-risk MF. The results showed that treatment with ruxolitinib led to significant improvement in OS, associated with reductions in splenomegaly and MF-related symptoms and quality of life as compared with control therapy.
Long-term follow-up continued for both trials. Verstovsek and colleagues performed an exploratory analysis of pooled data for OS in the 2 COMFORT trials. COMFORT-I was placebo controlled, and the control arm of COMFORT-II was BAT. Patients were randomized 1:1 in COMFORT-I and 2:1 ruxolitinib versus control in COMFORT-II. The pooled analysis included 301 patients randomized to ruxolitinib and 227 assigned to the control groups.
Patients in the control arms could cross over to ruxolitinib for prespecified worsening of splenomegaly, lack of effect on spleen length, or occurrence of protocol-defined events. Crossover was mandatory after unblinding in COMFORT-I. All patients still on randomized control treatment crossed over to ruxolitinib after 3 years of follow-up.
Median duration of ruxolitinib exposure was 2.5 to 3.0 years in the 2 trials, as compared with <1 year in the control groups. In COMFORT-I, 73.5% of patients in the control group crossed over to ruxolitinib, as did 61.6% of patients on the COMFORT-II control arm. Overall, 156 of 227 (69.6%) patients assigned to control crossed over.
The 1.5-year absolute difference in median OS demonstrated by the ITT analysis represented a 30% reduction in the hazard for death in favor of ruxolitinib (95% CI 0.54-0.91, P
After correction for crossover, the median OS was 5.3 years with ruxolitinib and 2.3 years with control treatment. The difference translated into a 65% reduction in the hazard for death (95% CI 0.23-0.59).
Further analysis showed that OS had yet to be reached for ruxolitinib-treated patients with intermediate-2 disease (versus 4.2 years for high-risk patients, HR 2.86, P
< 0.0001). The estimated median OS for the intermediate-2 subgroup was 8.5 years. A subgroup of patients with intermediate-2 primary MF (n = 58) originally randomized to ruxolitinib also lived much longer as compared with high-risk patients (n = 89) assigned to ruxolitinib (HR 2.55, P
The 5-year survival analysis showed that patients with intermediate-2 primary MF had not yet reached median OS (estimated median of 5.8 years), which compared favorably with a median OS of 4.0 years for a historical control group with intermediate-2 primary MF. Median OS for the ruxolitinib-treated patients with high-risk primary MF was 2.8 years, as compared with a median of 2.3 years for historical controls.
A subgroup analysis demonstrated a consistent survival benefit with ruxolitinib versus control therapy (HR 0.85 to HR 0.55). The analysis included patients ≤65 or >65, men, women, all 3 principal types of myeloproliferative neoplasms, spleen length ≤10 cm or >10 cm, presence or absence of anemia, baseline platelet count, and presence or absence of the JAK2V617F