Patients with high-risk stage III melanoma who received one year of pembrolizumab after surgery saw their chance of cancer returning drop by roughly 43 percent compared with those given a placebo, according to five-year data from the EORTC 1325-MG/KEYNOTE-054 trial. The U.S. Food and Drug Administration has approved pembrolizumab for this adjuvant setting, and a separate long-running trial of nivolumab, CheckMate 238, has now reported nine years of follow-up showing sustained benefits from a different checkpoint inhibitor. Together, these results are reshaping how oncologists treat melanoma patients whose tumors have been surgically removed but still carry a high probability of spreading.
Why five-year recurrence data changed the treatment calculus
Melanoma that has reached regional lymph nodes, classified as stage III, returns in a majority of patients within a few years of surgery alone. The question driving both KEYNOTE-054 and CheckMate 238 was whether giving an immune checkpoint inhibitor right after complete resection could meaningfully delay or prevent that recurrence. KEYNOTE-054 answered with a clear separation: at five years, 55.4 percent of patients randomized to pembrolizumab remained free of recurrence, versus 38.3 percent on placebo. That 17-point gap held across follow-up updates published in the Journal of Clinical Oncology, which also broke results down by disease substage and biomarker status.
The trial’s design, a randomized, double-blind comparison registered on ClinicalTrials.gov, pre-specified both recurrence-free survival and distant metastasis-free survival as primary endpoints. Distant metastasis-free survival matters because melanoma that spreads to organs such as the brain or lungs is far harder to treat than a local recurrence near the original site. Separate results published in The Lancet Oncology confirmed that pembrolizumab also reduced the rate of distant spread, reinforcing that the drug was not simply catching local relapses but preventing the kind of metastatic disease that kills.
The regulatory impact of these findings was immediate. The FDA’s decision to authorize pembrolizumab as an adjuvant option for resected stage III melanoma relied on the magnitude and durability of the recurrence-free survival benefit, as summarized in the agency’s communication on adjuvant treatment. For clinicians, that approval effectively moved pembrolizumab from a drug reserved for advanced disease into a standard consideration right after surgery for patients at high risk of relapse.
A hypothesis worth tracking is whether certain biomarker subgroups benefit even more dramatically. Patients whose tumors show both high tumor-mutational burden and PD-L1 expression above 50 percent are biologically primed to respond to checkpoint blockade. If that subgroup achieves at least a 15-percentage-point greater absolute reduction in distant metastasis at seven years than the trial-wide average, it would argue for even more targeted use of adjuvant therapy. Published subgroup analyses from KEYNOTE-054 hint at stronger effects in PD-L1–positive tumors, but seven-year biomarker-stratified distant metastasis data have not yet appeared in the peer-reviewed record. That gap limits how confidently oncologists can promise individual patients a specific degree of benefit based on their tumor profile.
CheckMate 238 and the nine-year follow-up for nivolumab
KEYNOTE-054 is not the only source of long-term evidence. CheckMate 238 randomized patients with resected stage IIIB through IV melanoma to receive either nivolumab or the older immunotherapy ipilimumab. At five years of follow-up, recurrence-free survival for nivolumab reached approximately 50 percent, with a corresponding advantage in distant metastasis-free survival over the ipilimumab arm. Detailed five-year outcomes and exploratory biomarker work are described in the open-access report on long-term nivolumab, which examined tumor PD-L1 status and quantitative immune markers to add granularity to the question of who benefits most.
The trial then extended its observation window to nine years, with results reported in the New England Journal of Medicine. That duration of follow-up is rare in adjuvant immunotherapy and provides some of the longest survival curves available for this patient population. The sustained separation between the nivolumab and ipilimumab arms over nearly a decade suggests the early benefit is not simply delaying recurrence but may be preventing it outright in a subset of patients. A summary published by the American Cancer Society’s journal characterized the nine-year data as showing a “sustained benefit” for nivolumab in high-risk resected melanoma, echoing the impression that many recurrences are being averted rather than merely postponed.
Safety has also been a central concern, since adjuvant therapy is given to patients who may already be cured by surgery. In CheckMate 238, nivolumab showed a more favorable tolerability profile than ipilimumab, with fewer high-grade immune-related adverse events and lower discontinuation rates. Over nine years of follow-up, no new late-onset safety signals emerged in the nivolumab arm, which reassures clinicians that the risk of long-term toxicity appears limited. This contrasts with the higher burden of side effects historically seen with CTLA-4 blockade, supporting the shift toward PD-1–targeted agents in the adjuvant setting.
Comparing pembrolizumab and nivolumab in practice
Although KEYNOTE-054 and CheckMate 238 were not designed to be compared head-to-head, their broadly similar patient populations and endpoints allow some cautious cross-trial observations. Both trials enrolled patients with completely resected stage III or IV melanoma at high risk of recurrence. Both demonstrated that a year of PD-1 blockade could roughly halve the rate of relapse compared with either placebo (for pembrolizumab) or an older immunotherapy standard (for nivolumab). In each case, the benefit extended to distant metastasis-free survival, the outcome most tightly linked to eventual mortality.
For oncologists, the choice between pembrolizumab and nivolumab often comes down to familiarity, dosing preferences, and institutional pathways rather than clear efficacy differences. Pembrolizumab was tested against placebo, providing a clean estimate of absolute benefit over observation, while nivolumab’s comparator, ipilimumab, was itself an active drug with known though smaller adjuvant benefit. As a result, some clinicians view the pembrolizumab data as more directly quantifying how much risk reduction adjuvant PD-1 therapy can provide over doing nothing, and the nivolumab data as demonstrating superiority over a prior immunotherapy benchmark.
Both agents share similar mechanisms, targeting the PD-1 receptor on T cells to release inhibitory brakes that tumors exploit to evade immune attack. Their toxicity profiles overlap as well, with immune-related side effects affecting organs such as the thyroid, lungs, liver, and colon. In adjuvant use, the key question is whether the long-term reduction in recurrence justifies exposing patients to these risks for a year. The five- and nine-year follow-up data suggest that, for many high-risk patients, the answer is yes, but the decision remains individualized, particularly for those with comorbidities or marginal performance status.
What long-term data mean for patients and future research
The durability of benefit seen in both KEYNOTE-054 and CheckMate 238 has practical implications for patient counseling. Oncologists can now explain that the advantage of adjuvant PD-1 therapy is not confined to the first couple of years after surgery. Instead, the curves remain separated at five and even nine years, indicating that many patients who avoid early recurrence may never see their melanoma return. That prospect of long-term remission or functional cure is a powerful motivator for patients weighing the burdens of treatment against the fear of relapse.
At the same time, the trials highlight unresolved questions. Not all patients benefit equally, and some still recur despite adjuvant therapy. Better biomarkers are needed to identify who can safely forgo treatment, who should receive standard PD-1 blockade, and who might require more intensive combinations. Ongoing studies are testing dual checkpoint inhibition, targeted therapy combinations for BRAF-mutant disease, and shorter or response-adapted treatment durations informed by circulating tumor DNA. The long-term datasets from pembrolizumab and nivolumab provide the benchmark against which these new strategies will be measured.
For now, the message from the five- and nine-year follow-ups is clear: adjuvant PD-1 blockade has transformed the outlook for patients with high-risk resected melanoma. What was once a disease marked by frequent, early recurrence after surgery can now, for many, be held in check for years. As more mature data accumulate and biomarker science advances, the hope is that these gains can be extended further, while sparing patients unlikely to benefit from unnecessary toxicity.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
