KEYNOTE-054 / EORTC 1325-MG
Does one year of adjuvant pembrolizumab improve recurrence-free survival after resection of high-risk stage III melanoma?
Trial design and population
Population: 1,019 patients with completely resected high-risk stage III melanoma.
Treatment vs comparator: Pembrolizumab 200 mg IV every 3 weeks for 18 doses vs placebo.
Registry: NCT02362594
Key results
At 7 years, RFS 50% vs 36% (HR 0.63, 95% CI 0.53–0.74); DMFS 54% vs 42% (HR 0.64, 95% CI 0.54–0.76). Approximate 7-year NNT 7 for one additional patient recurrence-free, based on rounded rates; this analysis does not establish OS benefit.
Seven-year recurrence-free survival · 7 years
Bars are scaled to the larger reported value. Median survival differences are not NNTs.
Absolute effect
Endpoint/time horizon: Seven-year recurrence-free survival · 7 years. NNT is not calculated from median survival or hazard ratios; use fixed-time absolute risks only. NNH not calculated without a defensible fixed-time excess-harm risk difference.
Safety, burden and QoL
Immune-mediated endocrinopathies, colitis, hepatitis, pneumonitis and skin toxicity may require steroids or lifelong hormone replacement. One-year treatment burden and QoL trade-offs matter after surgery.
QoL and discontinuation percentages are not asserted without full source-table verification; source-level QoL/toxicity audit remains a QC item.
Subgroups, maturity and limitations
RFS/DMFS benefit does not automatically establish OS benefit. Contextualize with current stage, BRAF status and contemporary adjuvant options.
Distinguish prespecified from exploratory subgroup analyses and mature from immature endpoints. Verify current guideline context before applying historical trial findings.
Current-practice context · 21 October 2026
Confirm current guidelines, regulatory status, biomarkers, prior treatment, comorbidities and patient goals. Full clinical freshness and source-level QoL audit remain pending.