乌鸦传媒

A year of adjuvant pembrolizumab (Keytruda) reduced the risk of relapse in high-risk melanoma compared with former standards of care, but failed to improve overall survival (OS), a phase III trial found.

Relapse-free survival (RFS) for patients treated with pembrolizumab was significantly improved compared with those who received either ipilimumab (Yervoy) or high-dose interferon (HR 0.74, 95% CI 0.57-0.96), reported Kenneth Grossmann, MD, PhD, of the Huntsman Cancer Institute in Salt Lake City.

For OS, however, the PD-1 inhibitor came up short in both the intent-to-treat (ITT) population (HR 0.84, 95% CI 0.62-1.13) and PD-L1-positive subgroup (HR 0.88, 95% CI 0.60-1.29), according to findings from S1404, a SWOG Cancer Research Network clinical trial, presented at the virtual American Society of Clinical Oncology (ASCO) annual meeting.

"The overall survival impact of pembrolizumab was likely influenced by effective post-relapse treatment, with standard-of-care patients receiving PD-1 blockade," said Grossmann.

Following progression, nearly half of patients in the control arm went on to receive a PD-1 inhibitor alone (39%) or in combination with ipilimumab (6%) or other agents (2%).

But Grossmann concluded that "single-agent anti-PD-1 antibody treatment should be a standard-of-care option for adjuvant treatment of high-risk resected melanoma," and offered that the OS "trend of benefit is there" during a Q&A session following his presentation.

He added that quality-of-life analyses in S1404 and EORTC 1325/KEYNOTE-054 -- a placebo-controlled adjuvant trial where control-arm patients crossed over to pembrolizumab at recurrence -- will "gain significant importance over time, because of the potential impact of relapse on quality of life versus the toxicity you might experience while on therapy."

New data from the EORTC trial on 155 crossover patients, presented during the ASCO session, showed an overall response rate of 39% with pembrolizumab at recurrence (including complete responses in 17.5%), and a median progression-free survival of 8.5 months from the start of therapy. OS data for the trial remain immature.

"The current problem with adjuvant therapy is there's a group of patients that won't recur anyway," said ASCO Discussant Alexander Menzies, MBBS, PhD, of the University of Sydney. "There's a group that recur despite drug, and there's a group in the middle -- perhaps one in five patients -- that truly derive benefit. Yet all are treated, all have the risk of toxicity."

"One year of drug therapy is expensive, and it requires regular clinic visits and infusions," he added.

Menzies also noted that in S1404, no data on subsequent treatment was reported for nearly one-third of patients in the control arm, "so it still doesn't completely answer the question of the impact of post-recurrence treatment on overall survival."

But adjuvant therapy "is here to stay," concluded Menzies. "It's still unclear whether to treat now or later, if we observe and treat at recurrence only a minority have durable survival."

Study Details

From December 2015 to October 2017, S1404 randomized 1,345 patients with resected stage III/IV melanoma to adjuvant therapy with either pembrolizumab (n=640; 200 mg every 3 weeks for 1 year), high-dose interferon (n=146; for 1 year), or ipilimumab (n=421; 10 mg/kg for up to 2 years).

The trial's three primary endpoints were RFS and OS in the ITT population, and OS in the PD-L1 subgroup. The current per-protocol final OS analysis was conducted 3.5 years following the last randomized patient (53% full power).

Patients in the S1404 trial had a median age of 52-54 years, and 60% were men. About 11% had stage IIIA melanoma, half had IIIB disease, a third had IIIC disease, and the remaining were stage IV. Most patients (82%) were PD-L1 positive, and 21-22% had BRAF mutations.

No prior immunotherapy was allowed, and other exclusion criteria included uveal melanoma, autoimmune disease, brain metastases, and a history of hepatitis. Relapsed stage III was allowed.

For safety, grade ≥3 adverse events (AEs) were far less frequent with pembrolizumab, at 31.7% compared with 57.5% with ipilimumab and 71.9% with high-dose interferon. Treatment-related grade ≥3 AEs occurred in 19.5%, 49.2%, and 71.2%, respectively. There were two AE-related deaths in the pembrolizumab arm, and two patients on ipilimumab died due to AEs.

AEs leading to treatment discontinuation occurred in 16.9% of patients on pembrolizumab versus nearly two-thirds of those on ipilimumab and about one-fourth of those on interferon.

Comment