After 4 years of follow-up, a majority of patients with polycythemia vera (PV) that responded to treatment with ruxolitinib (Jakafi) maintained their responses, results from a randomized trial showed.
Three-fourths of patients who had a primary response had persistence of that response at 208 weeks. Additionally, 70% of patients who had clinicohematologic (CLHM) response maintained that status after 208 weeks.
The safety profile remained unchanged from a previous analysis after 80 weeks of follow-up, according to a presentation at the 2017 ASH Annual Meeting in Atlanta.
“Patients with polycythemia vera who were resistant to or intolerant of hydroxyurea [Hydrea] had durable hematocrit control and clinocohematologic response,” said Jean-Jacques Kiladjian, MD, PhD, head of clinical investigation at Saint Louis Hospital in Paris. “Taken together (with the safety data), these results support ruxolitinib as an effective long-term treatment option for patients with polycythemia vera who have had an inadequate response to or are intolerant of hydroxyurea.”
The findings came from an updated analysis of the randomized RESPONSE trial of ruxolitinib for patients with PV and hydroxyurea resistance or intolerance. Patients were randomized to ruxolitinib or best available therapy. The trial had a composite primary endpoint of hematocrit control plus at least a 35% reduction in spleen volume at week 32.2
At the primary analysis, 21% of 110 patients assigned to ruxolitinib met the primary endpoint, as compared with 1% of patients assigned to best available care (P<.001). Response to ruxolitinib was similar in patients who had hyroxyurea intolerance (22%) or resistance (20%).
The RESPONSE trial design specified a long-term analysis of safety and efficacy at 208 weeks. Follow-up will continue, said Kiladjian.
Investigators in RESPONSE randomized 222 patients to ruxolitinib or the control group. At 208 weeks, 41 patients (37%) originally randomized to ruxolitinib remained on treatment (median exposure 225 weeks), as compared with none of the patients allocated to the control group (median exposure 34 weeks). Kiladjian reported that 37 of 98 patients (38%) who crossed over from the control arm to ruxolitinib after 32 weeks remained on treatment. Additionally, 32 patients (29%) assigned to ruxolitinib completed 32 weeks of treatment versus 1 patient (1%) in the control group.
The principal outcome of interest for the 208-week analysis was durability of efficacy, including the primary endpoint and the prespecified secondary endpoint of CLHM, defined as hematocrit control, platelet count ≤400 x 109/L, and white blood cell (WBC) count ≤10 x 109/L. Overall survival was an exploratory endpoint.
At the 208-week landmark, 19 of 25 patients who attained primary responses to ruxolitinib at 32 weeks maintained those responses. The absolute numbers translated into an estimated probability of 0.73 maintaining primary response. Analysis of the separate components of the primary endpoint yielded an estimated probability 0.73 (95% CI, 0.60-0.83) for maintaining hematocrit control and 0.86 (95% CI, 0.61-0.95) for maintaining spleen-volume reduction.
Among patients who attained a primary response at 32 weeks, the median duration of response had yet to be reached, said Kiladjian.
The estimated probability of maintaining CLHM remission at 208 weeks was 0.54 (95% CI, 0.31-0.72). Of 87 patients with WBC >10 x 109/L at baseline, 42 (48.3%) achieved WBC count ≤10 x 109/L at 208 weeks. Of 54 patients who had a platelet count >400 x 109/L, 26 (48.1%) had a platelet count ≤400 x 109/L at 208 weeks.
The 70 patients who attained CLHM at 32 weeks had a probability of maintaining that response of 0.67 at 208 weeks. Kiladjian reported that 21 of the 70 patients had progressed during long-term follow-up. The median duration of CLHM response had yet to be reached, he added.
The exploratory intention-to-treat analysis of estimated 5-year survival yielded values of 90.6% (95% CI, 80.1%-95.7%) for ruxolitinib and 87.7% (95% CI, 74.8%-94.3%) for the control arm, not accounting for crossover to the ruxolitinib arm from the control group. Kiladjian pointed out that no patient in the control group remained on assigned therapy at the time of the 80-week analysis.
The most frequent hematologic adverse events (AEs; all grades, irrespective of causality) in the ruxolitinib arm were anemia (9.3%; 1.0% grade 3/4) and thrombocytopenia (4.6%; 1.0% grade 3/4). Infection was the most common type of nonhematologic AE, occurring in 19.6% of patients in the ruxolitinib arm (3.7% grade 3/4). The next most common types of events were pruritus (7.3%), diarrhea (7.1%), headache (6.1%), arthralgia (5.9%), weight gain (5.6%), muscle spasms (5.2%), and fatigue (5.1%).
Since the week-80 analysis, 2 on-treatment deaths occurred in the ruxolitinib arm (gastric cancer and unspecified malignant neoplasm, neither considered related to treatment). Additionally, 4 on-treatment fatal AEs occurred among patients who crossed over to ruxolitinib, but none were considered related to treatment with ruxolitinib.