The KEYNOTE studies, which have advanced the use of pembrolizumab in metastatic melanoma, now report on long-term (KEYNOTE-001) and mature data (KEYNOTE-006) that support the approval of pembrolizumab in the treatment of refractory and newly diagnosed metastatic melanoma. Data from KEYNOTE-029 provide support for the efficacy of pembrolizumab in combination with ipilimumab. The studies were presented on June 6 during the ASCO Annual Meeting.
Data from KEYNOTE-001 indicated that pembrolizumab treatment provided 2-year and 3-year overall survival (OS) rates of 50% and 40%, respectively (Abstract 9503). Final data from KEYNOTE-006 showed that OS rate was significantly higher for patients on pembrolizumab 10 mg/kg given every 2 or 3 weeks compared with ipilimumab (55% vs 43%; p = 0.0008) given at its approved dosage (Abstract 9504). For the expanded cohort in KEYNOTE-029, the combination of pembrolizumab and ipilimumab demonstrated an overall response rate (ORR) of 57% (Abstract 9506).
“Before the advent of immune therapies, patients with metastatic melanoma had a survival of less than one year. Now some patients are alive for 5 and 10 years,” Caroline Robert, MD, PhD, of Gustave Roussy Cancer Center and Paris-Sud University, who reported on the KEYNOTE-001 study, told the ASCO Daily News. “We have real hope for our patients,” she added.
For the first time, we are hearing the word “cure” in regard to some patients with metastatic melanoma. All KEYNOTE speakers told the ASCO Daily News that long duration of response and high progression-free survival (PFS) rates will be key to identifying whether a therapy may be associated with a cure.
“These studies represent more than 2,500 patients, and, together with published clinical studies, we have significant data and experience with using PD-1 inhibitors,” discussant Marc S. Ernstoff, MD, of Roswell Park Cancer Center, said. His discussion included CheckMate 067, which evaluated nivolumab in combination with ipilimumab for the treatment of melanoma.
KEYNOTE-001: Pembrolizumab in Refractory/First-Line Metastatic Melanoma
In KEYNOTE-001, 655 patients were enrolled across several cohorts—one nonrandomized and three randomized—that included patients who were treatment naive, ipilimumab naive, or ipilimumab treated.
Patients received pembrolizumab 2 mg/kg or 10 mg/kg every 3 weeks or pembrolizumab 10 mg/kg every 2 weeks until unacceptable toxicity or progression. Response was evaluated every 12 weeks using RECIST version 1.1, and patients who discontinued treatment were contacted every 3 months to evaluate survival.
Median follow up was 32 months. Of 655 patients, 41% discontinued treatment for progressive disease and 21% were still on treatment. Median exposure to pembrolizumab was 5.6 months.
OS was similar across the two doses, Dr. Robert said. She showed that for a population with advanced disease, median OS of 24.4 months for the overall study population and 32.2 months for ipilimumab-naive patients. For the overall population, 2- and 3-year OS rates were 50% and 40%, respectively. For the approved 2 mg/kg every 3 weeks dosage, median OS was 23.5 months; 2-year OS was 49%; and 3-year OS was 38%.
The initial overall response rate (ORR) of 33% was sustained in this long-term follow up, as was the disease control rate of 51% (complete response, partial response, or stable disease). After 2 and 3 years, responses were ongoing in 73% and 57% of patients, respectively, which was impressive, Dr. Robert said. “Responses from this long-term analysis remained similar to those from the previous analysis,” she said.
At the time of this analysis, 19% of the 95 patients with complete responses remain on pembrolizumab, 17% discontinued treatment because of adverse events or progressive disease, and 64% stopped pembrolizumab after a complete response but continue to be observed. “The majority of patients with complete response have stopped treatment, and only two of them have experienced disease progression,” Dr. Robert said.
With 40% of patients alive at 3 years, Dr. Robert concluded that pembrolizumab continues to demonstrate safety and tolerability over long-term follow-up. Responses were durable in approximately two-thirds of patients who achieved a response. Even after treatment discontinuation for complete response, 97% of these patients are still disease free.
KEYNOTE-006: Pembrolizumab in the First-Line Treatment of Metastatic Melanoma
Jacob Schachter, MD, of the Ella Institute for Research and Treatment of Melanoma, Sheba Medical Center, Israel, reported on the phase III KEYNOTE-006 study, in which 834 patients were randomly assigned to receive pembrolizumab 10 mg/kg every 2 weeks (279 patients) or every 3 weeks (277 patients) for 24 months or to the approved dose of ipilimumab 3 mg/kg—four doses given every 3 weeks (278 patients). Primary efficacy endpoints were OS and PFS.
At the median follow up of 23 months, 201 and 213 patients receiving pembrolizumab every 2 weeks and every 3 weeks, respectively, discontinued therapy; and 110 patients receiving ipilimumab therapy discontinued therapy. Treatment was ongoing in 70 and 58 patients receiving pembrolizumab every 2 weeks and every 3 weeks, respectively, at the 23-month follow-up mark.
Dr. Schachter reported that median OS was not reached for patients on any dosage of pembrolizumab compared with 16 months for ipilimumab. Two-year OS was 55% with either dosages of pembrolizumab and 43% with ipilimumab. With a hazard ratio of 0.68 for any dosage of pembrolizumab, patients on pembrolizumab were at a 32% reduced risk for death compared with patients on ipilimumab (p = 0.0008).
Reported 2-year PFS rates were 31%, 28%, and 14% for pembrolizumab every 2 weeks, pembrolizumab every 3 weeks, and ipilimumab, respectively, with patients on either dose of pembrolizumab at a 39% reduced risk for progression compared with ipilimumab (p = 0.0001).
ORR were also significant for pembrolizumab—37% for the 2-week dosage, 36% for the 3-week dosage, and 13% for ipilimumab (p < 0.001); of patients who responded, complete responses were reported in 12% and 13% on the every 2 weeks and every 3 weeks dosage, and in 5% of patients on ipilimumab.
Dr. Schachter indicated that after a median of 22.8 months, of patients who responded, 67% and 60% of patients had an ongoing response on pembrolizumab every 2 weeks and every 3 weeks, respectively, compared with 62% of patients on ipilimumab.
Any-grade and grade 3/4 adverse event rates were similar across all treatment groups: for pembrolizumab every 2 weeks—82% and 17%, respectively; for pembrolizumab every 3 weeks—77% and 17%, respectively; and for ipilimumab—74% and 20%, respectively.
“Data from this study showed that pembrolizumab provides superior OS over ipilimumab,” Dr. Schachter said. He said that responses were durable and continued to accrue. “Data confirm pembrolizumab as a standard of care for patients with advanced melanoma,” he concluded.
KEYNOTE-029: Pembrolizumab With Ipilimumab in the Treatment of Advanced Melanoma
Georgina V. Long, BSc, PhD, MBBS, FRACP, of the Melanoma Institute Australia and the University of Sydney, reported on the phase 1b KEYNOTE-029 study; 153 patients with metastatic melanoma received the combination of pembrolizumab 2 mg/kg and ipilimumab 1 mg/kg every 3 weeks for 4 weeks and then pembrolizumab 2 mg/kg every 3 weeks.
Unlike the CheckMate 067 study, which examined the combination of nivolumab and ipilimumab, patients enrolled on KEYNOTE-029 were allowed to continue single-agent pembrolizumab after a high-grade event during the combination period, Dr. Long said. Notably, in KEYNOTE-029 ipilimumab was given at a low dose—1 mg/kg, not the full dose of 3 mg/kg as in CheckMate 067.
Median follow up was 10 months, and all evaluable patients had a minimum follow up of 6 months. In the expansion cohort, 24% had elevated LDH levels; 56% had M1c disease; 36% had BRAF V600E-mutant disease; 87% were treatment naive; 83% had PD-L1-positive tumors.
For the primary safety endpoint, Dr. Long said that immune-related adverse events of any grade were reported in 58% of patients; and 25% were grade 3/4. She pointed out that treatment was maintained in a high percent of patients—72% received all ipilimumab doses; 10% discontinued ipilimumab only; 7% discontinued pembrolizumab only; and 10% discontinued both drugs. Management of immune-related adverse events is key to retaining patients on treatment. Most patients were treated with steroids (57%) and few with infliximab (7%) or other methods (4%), with grade 3/4 events resolving in 81% of patients.
The ORR was 57%—10% of patients with a complete response, and 47% with a partial response. Stable disease was reported in 22% of patients. With immunotherapy, stable disease may in time convert to a partial response, Dr. Long noted. With 98% of patients maintaining the response at data cut and response duration between of 6+ months and 43+ months, these results are compelling, she said. Median OS at 6 months was 93%, and median 6-month PFS was 70%.
“We know that the approved combination of nivolumab and full-dose ipilimumab is toxic. We need to explore low-dose ipilimumab with anti–PD-1 therapy in a randomized trial to see if the combination can maintain efficacy with lower toxicity,” Dr. Long said.
Hints of a Cure and How It May Be Achieved
The ASCO Daily News interviewed the KEYNOTE presenters, and they discussed what the results mean with respect to the future potential for a cure for the disease.“Metastatic melanoma is not a disease that is hopeless to treat anymore,” Dr. Robert said. “We don’t know how to predict it yet, but we may have some patients who will be cured.”
“If we want to cure patients, we need to see a durability of response and high PFS rates,” Dr. Long said. The more patients we have progression free at 1, 2, and 3 years, the better are the chances of a cure, she said. “We need to use more efficient clinical trial endpoints,” she added.
“Duration of response is important if patients are not to relapse,” Dr. Schachter said. “We have the potential for a cure for the first time in stage IV melanoma.”
Dr. Ernstoff agreed. He said that response rates, which persist even in patients who have discontinued therapy following a complete response, surpass the historic remission rates seen with IL-2 therapy.
The Future of Immunotherapies
Data from KEYNOTE 006 and CheckMate 067 indicate that both pembrolizumab and nivolumab are superior to ipilimumab in clinical studies, Dr. Ernstoff said. It is highly unlikely that we will have head-to-head trials, but the two PD-1 inhibitors are remarkably similar. As single-agent therapy, each provides similar OS and PFS. In combination with ipilimumab, ORRs are also similar; any difference is insignificant, Dr. Ernstoff said. “If one is going to use single-agent therapy in our patients first, it is likely going to be a PD-1 inhibitor,” Dr. Ernstoff said.
He discussed the need for better biomarkers. Current biomarkers such as BRAF and PD-L1 are not optimal to determine unequivocally whether single-agent or combination immunotherapy is a better choice for patients, Dr. Ernstoff said.
With the plethora of choices, several questions arise in the clinical management of melanoma. How is one to sequence immunotherapies with targeted therapies? Dr. Ernstoff encouraged physicians to enroll patients on ECOG6134, which is designed to answer how to best sequence immunotherapy and targeted therapy.
He noted that with immunotherapies providing extended survival, as a community, we have to address the management of chronic low-grade toxicities, such as diabetes, which arise with these agents.
Several questions remain to be resolved, Dr. Ernstoff said. The efficacy of adding ipilimumab or using it as single agent after PD-1 failure is yet unknown. Do we continue treating past progression? The optimal length of therapy, the dose of ipilimumab, and the number of ipilimumab doses are all issues that arise in clinical management that need to be addressed, he said.
For Dr. Robert, the goal of treatment is to understand how to increase the response rates and see that the disease does not progress anymore. “We need to paralyze the disease,” she said. With a view to increase response rates and duration of response, future studies will study new drug combinations, she said.
Dr. Ernstoff pointed out that the U.S. National Cancer Institute’s PDQ Clinical Trial Database shows a list of 76 trials of PD-1 inhibitors in combination with drugs across several classes.
But cost always factors into a discussion of treating patients with immunotherapy. Dr. Ernstoff provided a wholesale acquisition cost of therapy with each of the immunotherapies and for the approved combination of ipilimumab and nivolumab and concluded that the headline for immunotherapy has changed dramatically over 3 years: From the “breakthrough of the year” in 2013 to “brutally selective, hugely expensive, and lifesaving” in 2016.
-Alexander Castellino, PhD