Long-Term KEYNOTE-006 Data Support Pembrolizumab as SOC in Advanced Melanoma

Pembrolizumab (Keytruda) continued to demonstrate improved survival benefits compared with ipilimumab (Yervoy) in patients with unresectable stage III or IV melanoma, according to 10-year follow-up data from the phase 3 KEYNOTE-006 study (NCT01866319) presented during the 2024 ESMO Congress.^1,2

The trial, which was launched to compare the safety and efficacy of the immunotherapy agents in this population, concluded on June 3, 2019.2 Eligible patients were able to transition to the open-label phase 3 KEYNOTE-587 extension study (NCT03486873), which was designed to continue collecting long-term efficacy in those with solid tumors who received prior pembrolizumab in a clinical trial. The median time from KEYNOTE-006 entry to KEYNOTE-587 data cutoff was 123.7 months (range, 122.0-127.3).

Patients who received pembrolizumab (n = 159) experienced a median overall survival (OS) of 32.7 months (95% CI, 24.5-41.6) vs 15.9 months (95% CI, 13.3-22.0) with ipilimumab (n = 52; HR, 0.71; 95% CI, 0.60-0.85). The 8- and 10-year OS rates in the pembrolizumab arm were 36.9% and 34.0%; in the ipilimumab arm, these rates were 24.8% and 23.6%. Those in KEYNOTE-006 who did not transition to the extension study were censored at the date last known to be alive or the data cutoff of KEYNOTE-006, which was July 31, 2019; those without an overall survival (OS) event were censored at the date last known to be alive or the data cutoff of KEYNOTE-587, which was May 1, 2024.

The modified median progression-free survival (PFS) with pembrolizumab was 9.4 months (95% CI, 6.7-11.6) vs 3.8 months (95% CI, 2.9-4.3) with ipilimumab (HR, 0.64; 95% CI, 0.54-0.75). The PFS rates at 8 and 10 years in the pembrolizumab arm were 23.4% and 22.0%; in the ipilimumab arm, these respective rates were 12.8% and 12.8%. As with OS, the patients who did not transition from one trial to the other were censored. The median melanoma-specific survival (MSS) with pembrolizumab was 51.9 months (95% CI, 30.0-114.7) vs 17.2 months (95% CI, 13.9-25.9) with ipilimumab (HR, 0.66; 95% CI, 0.55-0.81). The 8- and 10-year MSS rates in the pembrolizumab arm were 46.5% and 45.2%; these rates were 32.1% and 31.3% in the ipilimumab arm.

“With this 10-year follow-up, we have an OS rate of 34% but a MSS rate of 45.2%. We have a very good outcome for patients who responded or who had the benefit to the first course, and again, it shows that response is actually the best marker for long-term outcome. Unfortunately, we cannot know before [treatment] who is going to respond, but the patients who respond really do very well,” Caroline Robert, MD, PhD, of Gustave Roussy and Paris-Saclay University, in Villejuif, France, and a 2018 Giant of Cancer Care® inductee in melanoma, said in a presentation of the data. “We [also] show that we have an additional activity that is quite significant for patients who are rechallenged if they had the benefit in the first course of pembrolizumab. I also would like to remind you of the safety. We all know that the grade 3 and above adverse effects with a single[-agent] regimen is around 15% to 20%, which is far below what we can see with a combination.”

KEYNOTE-006 enrolled patients with unresectable stage III or IV melanoma not amenable to local therapy who had at least 1 measurable lesion by RECIST 1.1 criteria.2 Patients were at least 18 years of age, had an ECOG performance status of 0 or 1, and known BRAF mutational status. They could have received up to 1 prior systemic treatment as long as it was not an anti–CTLA-4, PD-1, PD-L1, or PD-L1 agent. Those with uveal or ocular melanoma or with known central nervous system metastases were excluded. Notably, patients with previously treated, stable brain metastases were permitted. Any patients enrolled in this trial who received study treatment or were in the survival follow-up phase of the research were able to transition to the extension study.

In KEYNOTE-006, participants were randomly assigned 1:1:1 to receive pembrolizumab at 10 mg/kg every 2 weeks up to 2 years, pembrolizumab 10 mg/kg every 3 weeks up to 2 years, or ipilimumab at 3 mg/kg every 3 weeks for 4 cycles.

Those who experienced a complete response (CR), which had been confirmed on 2 scans at least 4 weeks apart, were able to stop treatment, as long as they received pembrolizumab for 6 months or more and had received 2 or more doses of the agent beyond achievement of CR. Patients who experienced stable disease (SD) or better during the first course of treatment with pembrolizumab were able to receive a second course at 200 mg every 3 weeks for up to 1 year or approximately 17 cycles while in either trial if they experienced progressive disease after discontinuing therapy.

In KEYNOTE-006, investigators evaluated tumor response via CT/MRI at week 12; this was subsequently done every 6 weeks through week 48, and every 12 weeks thereafter. Once treatment was stopped, participants were monitored for survival every 12 weeks until the end of the trial. Those who transitioned to KEYNOTE-587 were then followed for long-term survival, progressive disease, and initiation of next anti-cancer treatment. Moreover, those who receive a second course of pembrolizumab in the extension trial were also followed for assessment.

For KEYNOTE-587, the primary end point was OS. Exploratory end points included PFS and PFS on the second course of pembrolizumab. Investigators conducted post-hoc analyses that examined MSS in the overall population, OS in subsets, OS and modified PFS from week 94 in those who received at least 94 weeks of pembrolizumab, and more.

Of those assigned to pembrolizumab treatment in KEYNOTE-006 (n = 556), 228 were eligible to transition to KEYNOTE-587, and 159 were enrolled in the trial. Of the 278 patients assigned to the ipilimumab arm on KEYNOTE-006, 105 were able to enroll in the extension trial and 52 did so.

Pembrolizumab continued to show superiority over ipilimumab with regard to OS outcomes across all subgroups analyzed, including those with elevated lactate dehydrogenase levels (n = 270; HR, 0.60; 95% CI, 0.44-0.80), those with large tumors defined as at least 10 cm (n = 186; HR, 0.64; 95% CI, 0.45-0.91), and those with brain metastases (n = 80; HR, 0.56; 95% CI, 0.32-0.98).

In those who completed at least 94 weeks of treatment with pembrolizumab (n = 103), and with a median follow-up duration of 122.9 months (range, 56.2-127.2), the median OS from week 94 was not reached (NR; 95% CI, NR-NR); the 6- and 8-year OS rates from week 94 were 91.8% and 80.8%, respectively. The median modified PFS in those who completed 94 weeks or more of pembrolizumab was also NR (95% CI, NR-NR); the 6- and 8-year modified PFS rates from week 94 were 73.2% and 64.8%, respectively.

Of the patients who received a second course of pembrolizumab in either trial (n = 16), 5 achieved a CR, 5 experienced a partial response, and 4 had SD, and 2 experienced disease progression. The median modified PFS from the start of the second course of pembrolizumab was 51.8 months (95% CI, 11.0-NR); 4- and 6-year modified PFS rates from the start of the second course were 56.3% and 49.2%, respectively.

Of the overall population (n = 556), 66.2% of those who received pembrolizumab and 65.1% of those given ipilimumab received the agents in the first-line setting. In these patients, the median OS with pembrolizumab or ipilimumab was 38.7 months (95% CI, 27.3-50.9) and 17.2 months (95% CI, 13.8-26.2; HR, 0.68; 95% CI, 0.55-0.86). The respective 8-year OS rates were 40.1% and 27.0%; the 10-year OS rates were 36.9% and 25.0%. The median modified PFS in these patients who received pembrolizumab or ipilimumab as first-line treatment was 12.0 months (95% CI, 8.3-16.6) and 4.1 months (95% CI, 2.9-5.2; HR, 0.62; 95% CI, 0.50-0.76). The 8- and 10-year modified PFS rates were 27.0% and 25.5% with pembrolizumab and 15.1% and 15.1% with ipilimumab.

“These results confirm that pembrolizumab provides long-term benefit in patients with advanced melanoma, supporting it as a standard of care in this setting,” the study authors conclude.

Disclosures: Dr Robert disclosed receipt of consulting fees from Bristol Myers Squibb, Roche, Pierre Fabre, Novartis, Sanofi, Pfizer, MSD, Merck, Sun Pharma, Ultimovacs, Regeneron, Egle, Philogen, MaaT Pharma, and IO Biotech. She received honoraria from Pierre Fabre, Sanofi, BMS, MSD, and Novartis and support for attending meetings and/or travel from Pierre Fabre. She serves on the data safety monitoring board or advisory board for Bristol Myers Squibb, Roche, Pierre Fabre, Novartis, Sanofi, Pfizer, MSD, Merck, Sun Pharma, Ultimovacs, Regeneron, Egle, Philogen, and MaaT Pharma. Study funding was provided by Merck Sharp & Dohme LLC, a subsidiary of Merck & Co. Medical writing and/or editing assistance was provided by Jemimah Walker, PhD, and Mehak Aggarwal, PharmD, of ApotheCom and the assistance was funded by Merck Sharp & Dohme LLC.

References

1. Robert C, Carlino MS, McNeil C, et al. Pembrolizumab vs ipilimumab for advanced melanoma: 10-year follow-up of the phase 3 KEYNOTE-006 study. Presented at: 2024 ESMO Congress; September 13-17, 2024; Barcelona, Spain. Presentation LBA44.
2. Long GV, Carolino MS, McNeil C, et al. Pembrolizumab versus ipilimumab for advanced melanoma: 10-year follow-up of the phase 3 KEYNOTE-006 study. Ann Oncol. Published online September 15, 2024. doi:10.1016/j.annonc.2024.08.2330