Recurrence-free survival was similar between adjuvant therapy with the PD-1 inhibitor pembrolizumab and placebo in patients with hepatocellular carcinoma who achieved a complete radiologic response after surgical resection or local ablation, based on the phase III KEYNOTE-937 trial.1
Presented at the 2026 ASCO Gastrointestinal (GI) Cancers Symposium, data from the third interim analysis showed a median duration of recurrence-free survival of 46.7 months (95% confidence interval [CI] = 35.6–53.3 months) with pembrolizumab vs 45.5 months (95% CI = 35.6–58.0 months) with placebo (hazard ratio [HR] = 1.06, 95% CI = 0.88–1.26; P = .719) in the intention-to-treat population. The rates at 24 and 48 months were 63% and 50% with pembrolizumab vs 61% and 50% with placebo, respectively.
Likewise, the preplanned subgroup analysis showed no benefit with pembrolizumab. Although some outliers were observed, “this has to be interpreted cautiously because of the small sample size in the corresponding subgroup,” said presenting author Stephen Lam Chan, MD, FRCP, of the Chinese University of Hong Kong, Hong Kong SAR, China.
“The study did not proceed to the final analysis, as the recurrence-free survival hypothesis was not met, and, per multiplicity, overall survival was not tested statistically,” he remarked.
About KEYNOTE-937
Treatment guidelines recommend surgical resection and local ablation as potentially curative options for patients with newly diagnosed hepatocellular carcinoma.2,3 Five-year survival rates range from 60% to 80% with surgical resection and 40% to 70% with local ablation.4 Despite these outcomes, Dr. Chan explained that tumor recurrence is common, highlighting the unmet need for standard-of-care adjuvant therapies for disease prevention and to improve overall survival.
In this double-blind study, a total of 959 patients with confirmed hepatocellular carcinoma were randomly assigned in a 1:1 ratio to receive intravenous adjuvant therapy with 200 mg of pembrolizumab (476 assigned; 471 treated) or a placebo (483 assigned; 482 treated) every 3 weeks for a year. They were required to be aged at least 18 years, have achieved a complete radiologic response after surgical resection or local ablation, and have an Eastern Cooperative Oncology Group (ECOG) performance status of 0 or 1, as well as a Child-Pugh class A score. Randomization was stratified by region (Asia [not including Japan] vs other), prior local therapy (resection vs ablation), risk of recurrence (intermediate vs high vs very high), and the alpha fetoprotein value at diagnosis (< 200 ng/mL vs ≥ 200 ng/mL).
Dr. Chan highlighted key baseline characteristics of the study population: most patients were Asian (54%–56%), had an ECOG performance status of 0 (94%–95%), had hepatitis B as the underlying etiology (59%–60%), underwent surgical resection (88%), belonged to the high–recurrence risk group (71%–74%), had Barcelona Clinic Liver Cancer stage A disease (83%), and exhibited microvascular invasion (57%–58%).
The coprimary endpoints were recurrence-free survival, assessed by blinded independent central review or pathology, and overall survival. Key secondary endpoints included distant metastasis–free survival, which was evaluated by blinded independent central review or pathology, and safety. At the third interim analysis, follow-up data were provided for a median of 50.7 months.
Other Efficacy Data
Data for the coprimary endpoint of overall survival were immature, Dr. Chan noted, with the median not reached in either treatment arm. The HR of 1.08 (95% CI = 0.81–1.43) was provided solely for descriptive purposes, given the absence of a formal statistical comparison. At 24 months, the overall survival rate was 90% with pembrolizumab and 94% with placebo; these were 79% and 81%, respectively, at 48 months.
Similarly, the median duration of distant metastasis–free survival was not reached in either arm. Dr. Chan reported a HR of 0.98 (95% CI = 0.77–1.24) for descriptive purposes. The 24-month distant metastasis–free survival rate was 82% with pembrolizumab and 83% with placebo, and the 48-month rates were 71% vs 70%, respectively.
Safety Findings
Among patients who received at least one dose of pembrolizumab, 60% experienced a treatment-related adverse event, and 14% experienced an event of grade 3 or 4 in severity. No deaths were attributed to treatment-related adverse events, and 10% of patients discontinued treatment with pembrolizumab because of these toxicities. Pruritus, rash, and hypothyroidism were the most frequently reported treatment-related adverse events in the arm and, according to Dr. Chan, “were comparable with our experience with pembrolizumab in the advanced setting.”
Immune-mediated adverse events and infusion reactions of grade 3 to 4 were documented in 7% of patients receiving pembrolizumab. As noted by Dr. Chan, “again, when you look at the types of the immune-mediated adverse events, we don’t see any new safety signals.” Corticosteroids were used to manage immune-mediated adverse event episodes in 48 patients in the pembrolizumab arm and 1 patient in the placebo arm.
Taking together the findings from this analysis of the KEYNOTE-937 trial, Dr. Chan concluded, “There remains an unmet need for developing treatment, either adjuvant or perioperative treatment, for patients with early-stage hepatocellular carcinoma after surgery or local ablation.”
DISCLOSURE: The study was funded by Merck Sharp & Dohme. Dr. Chan has received honoraria from AstraZeneca, Eisai, Elevar Therapeutics, Ipsen, Merck Sharp & Dohme, and Roche; has held a consulting or advisory role with AstraZeneca, Eisai, Elevar Therapeutics, Exelixis, MSD Oncology, and Roche; has received research funding from Eisai and Merck Sharp & Dohme; and has received reimbursement for travel expenses from Ipsen and Roche.
REFERENCES
1. Chan SL, Bouattour M, Yau T, et al: Adjuvant pembrolizumab for participants with hepatocellular carcinoma and complete radiologic response after surgical resection or local ablation: The phase 3 KEYNOTE-937 study. 2026 ASCO GI Cancers Symposium. Abstract 477. Presented January 9, 2026.
2. Benson AB, D’Angelica MI, Abrams T, et al: NCCN Clinical Practice Guidelines in Oncology: Hepatocellular Carcinoma. Version 2.2025. Available at www.nccn.org. Accessed January 16, 2026.
3. European Association for the Study of the Liver: EASL Clinical Practice Guidelines on the management of hepatocellular carcinoma. J Hepatol 82:315-374, 2025.
4. Llovet JM, Burroughs A, Bruix J: Hepatocellular carcinoma. Lancet 362:1907-1917, 2003.
EXPERT POINT OF VIEW
“Treatment in advanced-stage [hepatocellular carcinoma] has changed dramatically with the advent of various immune-based treatment approaches, [and] the logical next step is to move this into the early stage and intermediate stage,” said invited discussant Tim F. Greten, MD, of the National Cancer Institute, referencing the aim of the KEYNOTE-937 trial of adjuvant treatment with the PD-1 inhibitor pembrolizumab vs placebo in patients who achieved a complete radiologic response after surgical resection or local ablation.1
The study “unfortunately resulted in no benefit in recurrence-free survival,” but “we learn more than we think…,” Dr. Greten emphasized. He highlighted that the 48-month recurrence-free survival rate of approximately 50% is “very positive” and indicates that patients who underwent surgery or radiofrequency ablation are “doing really well.” Although he congratulated the investigators and sponsors for their work, he stressed that “we owe our patients—959 patients to be specific here—to get the most data out of these studies, even if these are negative studies, and I hope there are going to be follow-up studies using this information.”
The findings led Dr. Greten to focus on two discussion points that could guide future research: defining high-risk patients and understanding whether anti–PD-(L)1 therapy works in the absence of tumor antigens.
Defining High-Risk Patients
Dr. Greten highlighted that current recurrence risk markers—tumor size, multiplicity, microvascular invasion, poor differentiation, satellite nodules, and alpha fetoprotein—“are not good enough because we have not been able to identify the patients that are at really high risk.” He suggested that integrating serum markers, circulating tumor DNA, genetic signatures, and immune profiling may help “dissect which patients are at high risk and will benefit from adjuvant treatment.”
Placing KEYNOTE-937 in the context of current adjuvant research, Dr. Greten stated, “Some have already read out…,” highlighting the IMbrave050 trial.2 Several ongoing trials of emerging immunotherapy-based approaches for hepatocellular carcinoma in the adjuvant setting—including bispecific antibodies—“are about to read out in the very near future” and may reshape the treatment landscape, he suggested, referencing EMERALD-2, CheckMate 9DX, SHR-1210-III-325, and COMPASSION-36.
Immunologic Insights
Despite potential advances on the horizon, Dr. Greten reminded the audience that “systemic antitumor immunity requires antigens … These have to be presented by dendritic cells and draining lymph nodes.”
This immunologic principle helps clarify why the timing of therapy matters. Per Dr. Greten, neoadjuvant studies may induce better immune responses and enable tissue-based investigation; however, such treatment may delay surgery or introduce toxicities that affect surgical outcomes, and the role of adjuvant therapy remains unclear. In contrast, he noted that adjuvant studies allow risk stratification, clear clinical staging, and better defined clinical endpoints, but there may be delays in addressing micrometastases and limited information on tumor biology.
Dr. Greten also referenced the CARES-009 trial, which “addressed the question of perioperative treatment” by demonstrating significant improvements with perioperative camrelizumab plus rivoceranib vs surgery alone in patients with resectable disease at intermediate or high risk of recurrence.3 However, he noted an outstanding question in this space: “Does radiofrequency ablation actually do better in the context of the perioperative setting, where we release antigen [before surgery]?”
In conclusion, he emphasized that the results of KEYNOTE-937 “will ultimately lead to more studies in the perioperative field.”
DISCLOSURE: Dr. Greten reported no conflicts of interest.
REFERENCES
1. Chan SL, Bouattour M, Yau T, et al: Adjuvant pembrolizumab for participants with hepatocellular carcinoma and complete radiologic response after surgical resection or local ablation: The phase 3 KEYNOTE-937 study. 2026 ASCO GI Cancers Symposium. Abstract 477. Presented January 9, 2026.
2. Yopp A, Kudo M, Chen M, et al: Updated efficacy and safety data from IMbrave050: Phase 3 study of adjuvant atezolizumab + bevacizumab vs active surveillance in patients with resected or ablated high-risk hepatocellular carcinoma. 2024 ESMO Congress. Abstract LBA39. Presented September 13, 2024.
3. Zhou J, Fan J, Wang Z, et al: Perioperative camrelizumab plus rivoceranib in resectable hepatocellular carcinoma (CARES-009): A randomized, multicenter, phase III trial. 2025 ESMO Congress. Abstract 1470O. Presented October 19, 2025.
