At 85.3 months follow-up, the median overall survival with pembrolizumab was 32.7 months vs 15.9 months with ipilimumab.
At a 7-year follow-up, patients with pretreated unresectable stage III/IV melanoma continued to experience superior survival outcomes with pembrolizumab (Keytruda) vs ipilimumab (Yervoy), according to 7-year follow-up data from the phase 3 KEYNOTE-006 (NCT01866319) trial, which included some patients who transitioned to the phase 3 KEYNOTE-587 (NCT03486873) extension study, published in the Journal of Clinical Oncology.1
At a median follow-up of 85.3 months (range, 0.03-90.8), the median overall survival (OS) among patients who received pembrolizumab (n = 556) was 32.7 months (95% CI, 24.5-41.6) compared with 15.9 months (95% CI, 13.3-22.0) for those treated with ipilimumab (n = 278; HR, 0.70; 95% CI, 0.58-0.83). The 7-year OS rates were 37.8% vs 25.3%, respectively. The median modified progression-free survival (PFS) was 9.4 months (95% CI, 6.7-11.6) in the pembrolizumab arm vs 3.8 months (95% CI, 2.9-4.3) in the ipilimumab arm, with 7-year PFS rates of 23.8% vs 13.3%.
In the pembrolizumab arm, the complete response (CR), partial response (PR), and stable disease (SD) rates were 16.7%, 34.4%, and 33.1%, respectively, and the 7-year OS rates from the time of best overall response were 85.2%, 61.8%, and 25.9%.
Following treatment on KEYNOTE-006, 228 patients who received pembrolizumab and 105 patients treated with ipilimumab were eligible to enroll in KEYNOTE-587; 158 patients and 52 patients ultimately enrolled in KEYNOTE-587, respectively. Investigators noted a limitation of the study was the fact that not all eligible patients underwent further follow-up. Patients were eligible to transition to KEYNOTE-587 if they were receiving second-course pembrolizumab or were in survival follow-up per the KEYNOTE-006 study end date of June 3, 2019. Further, those in the survival follow-up phase could receive second-course pembrolizumab for up to a year in KEYNOTE-587.
Patients included in KEYNOTE-006 had received up to 1 prior systemic therapy for advanced disease, excluding CTLA-4, PD-1, or PD-L1 inhibitors. Pembrolizumab was given at a dose of 10 mg/kg once every 2 or 3 weeks for up to 2 years. Patients in the ipilimumab arm were treated at a dose of 3 mg/kg once every 3 weeks for 4 cycles.
Of patients who received second-course pembrolizumab (n = 16), a CR (n = 7), PR (n = 7), or SD (n = 2) was achieved during the first course treatment. During the second course of treatment, the objective response rate was 56% (95% CI, 30%-80%) with patients experiencing a CR (n = 4), PR (n = 5), SD (n = 5), and progressive disease (n = 2). The 2-year PFS rate was 62.5%, and the most common site where progressive disease occurred was the lymph nodes. Although data on these patients were not captured in KEYNOTE-587, response was ongoing for 5 of 7 patients who had a CR/PR, demonstrating that pembrolizumab retreatment may result in benefit for certain patients.
Patients with previously untreated disease who received pembrolizumab (n = 368) experienced a 7-year OS rate of 41.2% compared with 27.6% in the ipilimumab arm (n = 181). The median OS was 38.7 months (95% CI, 27.3-50.9) vs 17.2 months (95% CI, 13.8-26.2), respectively (HR, 0.67; 95% CI, 0.53-0.84). Additionally, the respective 7-year PFS rates were 26.8% vs 15.9% and the median PFS was 12.0 months (95% CI, 8.3-16.6), respectively (HR, 0.62; 95% CI, 0.50-0.76). These findings were consistent with prior analyses.
Previously, improvements in PFS and OS shown in data from KEYNOTE-006 resulted in the FDA expanding the approval of single-agent pembrolizumab to include the frontline treatment of patients with advanced melanoma regardless of BRAF status in December 2015.2
Additional findings from the subgroup analysis demonstrated that pembrolizumab was favorable compared with ipilimumab regardless of BRAF mutation status, lactate dehydrogenase (LDH) level, presence of brain metastases, or tumor size.1
Those with BRAF wild-type disease treated with pembrolizumab (n = 355) experienced a 7-year OS rate of 36.5% vs 25.7% among those treated with ipilimumab (n = 170), and the median OS was 28.1 months (95% CI, 21.1-42.7) and 13.9 months (95% CI,10.7-24.8), respectively (HR, 0.71; 95% CI, 0.56-0.89). Patients with BRAF-mutant disease who were previously treated with a BRAF/MEK inhibitor experienced 7-year OS rates of 28.3% vs 20.0% in the pembrolizumab (n = 87) and ipilimumab (n = 52) arms. The median OS for pembrolizumab in this subgroup was 20.4 months (95% CI, 12.8-35.6) vs 11.9 months (95% CI, 6.0-17.8) for ipilimumab (HR, 0.72; 95% CI, 0.47-1.08). Those with BRAF-mutant disease who were BRAF/MEK inhibitor naïve achieved 7-year OS rates of 49.7% vs 27.7% in the pembrolizumab (n = 87) and ipilimumab (n = 52) arms, with a median OS of 78.5 months (95% CI, 36.1-not evaluable) vs 26.2 months (95% CI 16.0-64.0; HR, 0.58; 95% CI, 0.38-0.89).
Patients with normal LDH levels who received pembrolizumab (n = 369) or ipilimumab (n = 179) achieved 7-year OS rates of 42.0% vs 30.5%, respectively (HR, 0.76; 95% CI, 0.60-0.96). Those with elevated LDH in the pembrolizumab (n = 179) and ipilimumab arms (n = 91) experienced 7-year OS rates of 28.9% vs 14.7%, (HR 0.59; 95% CI, 0.44-0.79).
Patients with brain metastases in the pembrolizumab (n = 51) and ipilimumab arms (n = 29) experienced 7-year OS rates of 50.0% vs 27.6%, respectively (HR, 0.49; 95% CI, 0.27-0.87). The 7-year OS rates for those without brain metastases in the pembrolizumab (n = 500) and ipilimumab arms (n = 248) were 36.8% compared with 25.0%(HR, 0.72; 95% CI, 0.59-0.87).
Finally, patients with tumors 10 cm or greater treated with pembrolizumab (n = 106) or ipilimumab (n = 51) experienced 7-year OS rates of 26.1% vs 15.9%, respectively (HR, 0.67; 95% CI, 0.46-0.99). Those with tumors less than 10 cm in the pembrolizumab (n = 292) and ipilimumab arms (n = 152) achieved 7-year OS rates of 40.7% vs 30.7%(HR, 0.74; 95% CI, 0.58-0.96).