Adjuvant Pembrolizumab Fails to Improve Outcomes in Hepatocellular Carcinoma After Curative Treatment
核心洞察
The phase 3 KEYNOTE-937 trial demonstrated that adjuvant pembrolizumab failed to improve recurrence-free survival compared to placebo in hepatocellular carcinoma (搜索) patients who achieved complete radiological response after surgical resection or local ablation.
Median recurrence-free survival was nearly identical between treatment arms at 46.7 months for pembrolizumab versus 45.5 months for placebo, with no statistical significance.
The study enrolled 959 patients and found no benefit in overall survival or distant metastasis-free survival, while pembrolizumab was associated with higher rates of grade 3-4 treatment-related adverse events.
The phase 3 KEYNOTE-937 trial has delivered disappointing results for adjuvant immunotherapy in hepatocellular carcinoma (搜索) (HCC (搜索)), with pembrolizumab failing to improve recurrence-free survival compared to placebo in patients who achieved complete radiological response after curative treatment. The findings, presented at the 2026 Gastrointestinal Cancers Symposium, represent a significant setback for efforts to prevent HCC recurrence following surgical resection or local ablation.
Trial Design and Patient Population
KEYNOTE-937 was a randomized, double-blind, placebo-controlled trial that enrolled 959 patients with confirmed HCC (搜索) who achieved complete radiological response following resection or local ablation. Patients were required to have an ECOG performance status of 0 or 1 and Child-Pugh class A liver function. The study excluded patients with recent esophageal or gastric variceal bleeding, clinically apparent ascites, or hepatic encephalopathy within six months of enrollment.
Participants were randomly assigned 1:1 to receive pembrolizumab at 200 mg intravenously every three weeks or matching placebo for up to one year. Treatment continued until disease recurrence, unacceptable toxicity, intercurrent illness, or patient withdrawal. Randomization was stratified by geographic region, prior local therapy type, recurrence risk, and alpha-fetoprotein levels at diagnosis.
The enrolled population had a median age of 62-63 years across both arms, with approximately 80% being male and over half of Asian ethnicity. Most patients had hepatitis B (搜索) infection (59-60%), underwent surgical resection as prior therapy (87-88%), and had high-risk disease (71-74%).
Primary Efficacy Results
The trial's primary endpoint of recurrence-free survival showed no benefit for pembrolizumab. Median RFS was 46.7 months (95% CI, 35.6-53.3) in the pembrolizumab arm compared to 45.5 months (95% CI, 35.6-58.0) in the placebo arm (HR, 1.06; 95% CI, 0.88-1.26; P = .719). The 24-month RFS rates were 63% for pembrolizumab versus 61% for placebo, while 48-month rates were identical at 50% in both arms.
"The study did not proceed to final analysis as the RFS hypothesis was not met, and per multiplicity, overall survival was not tested statistically," explained lead study author Stephen Lam Chan, clinical professor in the Department of Clinical Oncology at the Chinese University of Hong Kong.
Secondary Endpoints and Overall Survival
Although overall survival was a co-primary endpoint, the statistical design required meeting the RFS hypothesis first. Since this was not achieved, OS was not formally tested. Descriptive analysis showed median OS was not reached in either arm (HR, 1.08; 95% CI, 0.81-1.43), with 48-month OS rates of 79% for pembrolizumab versus 81% for placebo.
Distant metastasis-free survival outcomes were similarly disappointing, with median DMFS not reached in either group (HR, 0.98; 95% CI, 0.77-1.24). The 48-month DMFS rates were 71% for pembrolizumab compared to 70% for placebo, demonstrating no meaningful difference between treatments.
Safety Profile
Treatment-related adverse events were notably more frequent with pembrolizumab. Grade 3-4 treatment-related adverse events occurred in 14% of pembrolizumab patients compared to 5% receiving placebo. Treatment discontinuation due to adverse events was required in 10% of the pembrolizumab group versus only 1% in the placebo arm.
Any-grade adverse effects were reported in 91% of pembrolizumab patients compared to 82% receiving placebo. Immune-related adverse events and infusion reactions occurred in 26% of pembrolizumab patients versus 7% of placebo patients, with grade 3-4 events in 7% versus 1%, respectively. Importantly, no treatment-related deaths occurred in either group.
Clinical Implications
The negative KEYNOTE-937 results highlight the ongoing challenge of preventing HCC (搜索) recurrence after curative treatment. Despite surgery and ablation offering potential cures for early-stage disease, recurrence rates remain high, and this study demonstrates that adjuvant PD-1 (搜索) inhibition does not address this unmet medical need.
The trial's failure to meet its primary endpoint means that no adjuvant systemic therapy has yet demonstrated clear survival benefit in this setting. The findings suggest that the immune microenvironment in post-treatment HCC (搜索) may not be sufficiently responsive to checkpoint inhibition alone, or that alternative therapeutic approaches may be needed to prevent disease recurrence.
With a median follow-up of 50.7 months and robust statistical design, these results provide definitive evidence that adjuvant pembrolizumab does not offer clinical benefit in patients with HCC (搜索) who achieve complete radiological response after curative treatment.