Nivolumab elicited a prolonged clinical benefit in patients with advanced hepatocellular carcinoma regardless of prior sorafenib.
Nivolumab (Opdivo) elicited a prolonged clinical benefit in patients with advanced hepatocellular carcinoma (HCC) regardless of prior sorafenib (Nexavar) according to 5-year follow-up data from 2 cohorts of the phase 1/2 CheckMate040 trial (NCT01658878).
Patients were stratified into cohorts based on prior sorafenib exposure and were enrolled regardless of HCC etiology and PD-L1 expression status, according to Jörg Trojan, MD, head of the Gastrointestinal Oncology Unit at Goethe University Hospital and Cancer Center in Frankfurt, Germany, who presented that data during the 2021 International Liver Cancer Association Conference.1
At data cutoff December 2020, patients in the sorafenib-naïve cohort (n = 80) had a median follow-up of 62.9 months (range, 60-94) and a reported objective response rate (ORR) of 20%. Of the responders, 4% had a complete response (CR), 16% had partial response (PR). Stable disease (SD) was reported in 33%, and 40% had progressive disease (PD). The median duration of response was 22.6 months (range, 4.2-56.6).
For patients in the sorafenib cohort (n = 154), the median follow-up was 62.8 months (range, 60-86). The ORR was 14%, which included a CR and PR rate of 3% and 11%, respectively. SD and PD were reported for 42%, and 38%, respectively. The median duration of response in this cohort was 39.7 months (range, 3.2-79.8).1
“We have seen a clinical benefit from nivolumab regardless of the etiology in both groups,” Trojan said. “We found a similar benefit independent of the level of PD-L1 expression.”
In the sorafenib-naïve cohort, the ORR was 21% for uninfected patients (n = 47), 20% for those with hepatitis C virus (HCV; n = 25), and 13% for those with hepatitis B virus (HBV; n = 8). The ORRs were 27% and 20% for patients with PD-L1 expression of at least 1% (n = 27) and less than 1% (n = 20), respectively.
For the sorafenib-experienced patient cohort, the ORRs for uninfected patients (n = 13), those with HCV (n = 19), and those with HBV (n = 13) were 13%, 19%, and 13%, respectively. The ORRS were 27% and 13% for patients with PD-L1 expression of at least 1% (n = 27) and less than 1% (n = 13), respectively.
Despite a variety of options for patients in the first and second line, patients with advanced HCC have a poor prognosis and high mortality, Trojan explained. Sorafenib, the first systemic therapy approved by the FDA for this patient population in the first-line setting has recently been joined by lenvatinib (Lenvima) and the immuno‐oncology–based combination therapy of atezolizumab (Tecentriq) plus bevacizumab (Avastin), which has emerged as the standard-of-care therapy.2
Nivolumab was granted an accelerated approval by the FDA in 2017 based on data from CheckMate040 for the treatment of patients with HCC who have been previously treated with sorafenib. However, in July 2021, Bristol Myers Squibb voluntarily withdrew the indication for nivolumab from the US market following the FDA’s evaluation of confirmatory data for checkpoint inhibitors that have not met their post-marketing requirements.3 The FDA’s Oncologic Drug Advisory Committee voted 5 to 4 against the continued accelerated approval.
In the open-label, multicohort CheckMate040 study, 11 patients in the sorafenib-naïve cohort received nivolumab monotherapy at escalating dose levels (0.1-10.0 mg/kg) once every 2 weeks and 69 patients received nivolumab 3 mg/kg once every 2 weeks. Patients in the prior-sorafenib cohort received nivolumab monotherapy at expansion dose levels (3 mg/kg) once every 2 weeks.1
The median age for the sorafenib-naive cohort was 65 (range, 20-83) and 63 (range, 19-81) for that of the sorafenib-experienced cohort. In both populations, most patients had BCLC stage 3 HCC (90% and 90%, respectively), extrahepatic spread (60% and 71%, respectively), Child-Pugh score 5 (73% and 60%, respectively), and tumor cell PD-L1 expression less than 1% (70% and 71%, respectively). Of the patients in the sorafenib-experienced cohort, 91% had previously progressed on or after treatment with sorafenib.
A secondary end point in the follow-up analysis was OS, which was comparable between both treatment groups. In the sorafenib-naive group, the median OS was 26.6 months (range, 95% CI, 16.6-30.6). The sorafenib-experienced group had a median OS of 15.2 months (95% CI, 13.0-18.2).
Additionally, the 3-year and 5-year OS rates were 28% (95% CI, 18%-38%). and 14% (95% CI, 7%-23%) in the sorafenib-naïve cohort and 20% (95% CI, 14%-27%) and 12% (95% CI, 7%-18%) in the sorafenib-experienced cohort, respectively.
In an analysis of survival by best overall response (CR/PR) the median OS was 39.6 months (95% CI, 30.6-not estimable [NE]) in the sorafenib-naïve cohort. The median OS was not reached (95% CI, 28.1-NE) for those with best overall response in the sorafenib-experienced cohort.
Efficacy ranged for patients with differing baseline characteristics. Extrahepatic and alpha-fetoprotein (AFP) levels at least 400ug/L, for example, were associated with shorter OS, and tumor cell PD-L1 expression of at least 1 was associated with longer OS regardless of prior sorafenib, Trojan noted.
Specifically, of the patients in the sorafenib-naïve cohort who had a median OS of less than 1 year (n = 25), 76% had extrahepatic spread, 32% had high AFP levels, and 68% had PD-L1 expression less than 1%. Similar trends were observed for those with a median OS of less than 1 year in the sorafenib-experienced cohort with 74% of patients having extrahepatic spread, 42% with high AFP levels, and 71% with PD-L1 expression less than 1%.
Of note, higher levels of CD8 expression at baseline were associated with longer median OS in the sorafenib-naïve cohort; the median percentage of CD8-positive cells via immunohistochemistry in 16 patients with a median OS of 3 years or more was 6.5 (range, 1.2-17.8).
Investigators also reported survival data stratified by albumin and bilirubin (ALBI) grade and Child-Pugh score. Patients in the sorafenib-naïve cohort with ALBI grade 2 (n = 47) had a shorter median OS than those with ALBI grade 1 (n = 33) at 16.6 months (range, 1.0-66.3) and 26.7 months (range, 1.5-65.6), respectively. The median OS for patients in the sorafenib-experienced cohort with ALBI grade 2 (n = 76) was 13.1 months (range, 0.4-86.5) and 17.9 months (range, 0.4-64.9) for those with ALBI grade 1 disease (n = 78).
Baseline Child-Pugh score did not affect OS in patients who were sorafenib naïve at 20.1 months for those with Child-Pugh score 5 (n = 58; range, 1.5-66.3) and 20.1 months (range, 1.0-62.7) for those with Child-Pugh score 6 (n = 19). However, for patients who received prior sorafenib, those with worse liver status (Child-Pugh score 6; n = 48) had a shorter median OS at 11.5 months (range, 0.4-64.0) compared with 15.4 months (range, 0.4-86.5) for those who had Child-Pugh score 5 disease (n = 104).
Regarding safety, grade 3/4 treatment-related adverse events (TRAEs) were observed in 33% of patients in the sorafenib-naïve cohort and in 21% of patients in the sorafenib-experienced cohort. TRAEs leading to discontinuation occurred in 3% (grade 3/4) and 6% (any grade) of in patients who were sorafenib naïve and in 2% (grade 3/4) and 3% (any grade) of those who had prior sorafenib. No new safety signals were identified in the follow-up, Trojan said.
Immune-meditated adverse events with nivolumab were mostly grade 1/2, according to Trojan, with any-grade rash (16% and 10%, respectively), hepatitis (6% and 5%, respectively), and hypothyroidism/thyroiditis (6% and 5%, respectively) most commonly reported in the sorafenib-naïve and sorafenib-experienced cohorts, respectively.
“In summary, these results demonstrate a long-term benefit of nivolumab in a high proportion of patients with advanced HCC,” Trojan concluded.