Although nivolumab did not significantly outperform sorafenib in the CheckMate 459 of patients with advanced hepatocellular carcinoma, positive resuls were shown for the secondary end points.
Compared with sorafenib (Nexavar), nivolumab (Opdivo) does not significantly improve overall survival (OS), despite showing both clinical activity and favorable safety in patients with advanced hepatocellular carcinoma (HCC), according to results of the CheckMate 459 trial published in The Lancet Oncology.
Previous studies established that single agent nivolumab produces durable responses and manageable safety in patients with advanced HCC. However, it’s efficacy compared to sorafenib was unknown.
The phase 3 CheckMate 459 trial (NCT02576509) was designed to compare the efficacy of the 2 agents head-to-head. The trial had an actual enrollment of 743 participants and an estimated completion date of June 2022. The primary end point of the study is OS. Secondary end points include objective response rate, progression-free survival (PFS), and efficacy based on PD-L1 expression.
In total, 371 were assigned to receive nivolumab and 372 were assigned to receive sorafenib. In total, 329 patients in the nivolumab arm discontinued treatment. Reasons for discontinuation included progressive disease (245), adverse events uncreated to drug (37), study drug toxicity (32), patient request (8), other reasons (2), withdrew consent (2), maximum clinical benefit (1), no longer meeting study criteria (1), and death (1). In total, 371 were analyzed for efficacy and 367 for safety.
In the sorafenib arm, 355 patients discontinued treatment. Reasons for discontinuation included disease progression (240), study drug toxicity (41), adverse events unrelated to the study drug (40), patient request (17), withdrawal of consent (7), other reasons (7), poor compliance (1), lost to follow-up (1), and death (1). In total, 372 patients were evaluated for efficacy, and 363 for safety.
The median age of patients in the nivolumab group is 65 (range, 57-71), 85% were male, and 54% were white. Seventy-three percent of patients had an ECOG performance status of 0. Barcelona Clinical Liver Cancer Stages included A (4%), B (14%), and C (82%). Sixty percent of patients experienced extrahepatic spread and 33% experienced vascular invasion. Ten percent experienced tumor invasion of the liver and 98% of patients had a Child-Pugh score A. Fifty-four percent of patients underwent surgery and 52% of patients underwent non-systemic treatment.
The median age of patients in the sorafenib group was also 65 (range, 58-72), 53% were male, and 53% were white. Seventy percent had an ECOG performance status of 0. Barcelona Clinical Liver Cancer stages included A (5%), B (17%), and C (78%). Fifty-six percent of patients were experiencing extrahepatic spread and 32% were experiencing vascular invasion. Forty-seven percent of patients had more than 3 liver modules and 9% had tumor invasion in the liver. Ninety-six percent of patients had Child-Pugh score A and 56% of patients underwent locoregional therapy.
Fifteen percent of patients in the nivolumab arm experienced an objective response, compared with 7% of patients in the sorafenib group. In the nivolumab group, 4% of patients experienced a complete response, 12% experienced a partial response, 35% achieved stable disease, 37% experienced progressive disease, and 55% achieved disease control. The median duration of disease control was 7.5 months (range, 6.5-10.7). The median time to response was 3.3 months (range, 1.9-3.8). The median duration of response was 23.3 months (range, 17-31.5).
In the sorafenib group, 1% of patients experienced a complete response, 6% experienced a partial response, 48% achieved stable disease, and 28% experienced progressive disease. Fifty-eight percent of patients achieved disease control. The median duration of disease control was 5.7 months (range, 5.6-7.4), and the median time to response was 3.7 months (range, 2.1-5.3). The median duration response was 23.4 months (range, 8.5-NE).
In the nivolumab group, the 12-month OS was 60%, the 18-month OS was 47%, and the 34-month OS was 37%. In the sorafenib group, the 12-month OS was 55%, the 18-month OS was 44%, and the 24-month OS was 33%.
In terms of PFS, the 12-month PFS rate in the nivolumab group was 22%, the 18-month rate was 17%, and the 24-month rate was 14%. In the sorafenib group, the 12-month PFS was 14%, the 18-month rate was 9%, and the 24-month rate was 6%.
In both groups, adverse events were overwhelmingly grade 1 through 3 in nature and very rarely resulted in discontinuation. Common grade 1/2 adverse events for both the nivolumab and the sorafenib groups, respectively, included fatigue (14% vs. 22%), rash (11% vs. 12%), and diarrhea (8% vs. 42%).
“The proportion of responders in the nivolumab group was also higher, although not significantly higher, in patients with tumor cell PD-L1 expression of 1% and greater compared with patients with PD-L1 expression less than 1%, consistent with findings from a biomarker analysis in the nivolumab CheckMate 040 study. Tumor cell PD-L1 expression was not predictive for overall survival or progression-free survival in the current study,” wrote study authors.