Second-line pembrolizumab improves survival outcomes in advanced HCC

Pembrolizumab improved progression-free survival (PFS) and overall survival (OS) over best supportive care in patients with advanced hepatocellular carcinoma (HCC), according to the phase III KEYNOTE-240* trial, supporting its role as a treatment option for HCC in the second-line setting.

“Pembrolizumab reduced the risk of death by 22 percent and improved PFS over placebo in patients with advanced HCC,” said Dr Chee Cheng Ean from the National University Cancer Institute Singapore, who presented the findings at BASCO 2019.

A total of 413 patients with HCC and ECOG performance status 0–1, BCLC** stage B or C, and Child Pugh class A who were intolerant to or progressed on sorafenib therapy were randomized to receive pembrolizumab (200 mg Q3W; n=278, median age 67 years, 81.3 percent male) or placebo (n=135, median age 65 years, 83 percent male) in addition to best supportive care for up to 35 cycles.

OS was longer in patients on pembrolizumab than those on placebo (median 13.9 vs 10.6 months, hazard ratio [HR], 0.781, 95 percent confidence interval [CI], 0.611–0.998; p=0.0238). [BASCO 2019, abstract 4004]

In the primary analysis, patients on pembrolizumab had longer PFS than those on placebo (median 3.0 vs 2.8 months, HR, 0.775, 95 percent CI, 0.609–0.987; p=0.0186).

Despite these improvements with pembrolizumab, KEYNOTE-240 did not meet the statistical criteria for OS and PFS (p=0.0174 and p=0.0020, respectively), said the researchers. However, an improvement was noted in the final PFS analysis (HR, 0.718, 95 percent CI, 0.570–0.904; p=0.0022).

Subgroup analysis for OS and PFS showed a trend favouring pembrolizumab in all subgroups analysed.

The objective response rate was greater in patients on pembrolizumab than those on placebo (18.3 percent vs 4.4 percent), with a median 13.8-month duration of response in the pembrolizumab group (vs not reached in the placebo group). Six patients on pembrolizumab had a complete response, while 45 and 122 patients had a partial response and stable disease, respectively. In comparison, no placebo recipients had a complete response, with six and 66 patients having partial response and stable disease, respectively.

Grade 3–4 adverse events (AEs) occurred in 52.7 and 46.3 percent of pembrolizumab and placebo recipients, respectively, while 18.6 and 7.5 percent, respectively, experienced grade 3–4 treatment-related AEs (TRAEs). Grade 3–4 TRAEs led to treatment discontinuation in 6.5 and 0.7 percent of patients, respectively. Grade 3–4 immune-mediated AEs occurred in 7.2 and 0.7 percent of patients, respectively; 3.6 percent of patients in the pembrolizumab group and none in the placebo group discontinued treatment due to immune-mediated AEs.

The most common (≥15 percent) grade 3–4 AEs were elevated aspartate aminotransferase (AST; 13.3 and 7.5 percent of pembrolizumab and placebo recipients, respectively), bilirubin (7.5 and 5.2 percent), and alanine aminotransferase levels (ALT; 6.1 and 3.0 percent). The most common (≥5 percent) grade 3–4 TRAEs were elevated AST (5.4 and 1.5 percent) and ALT levels (3.6 and 1.5 percent).

“The safety profile, including incidence of immune-mediated events and hepatitis, was similar to that of pembrolizumab in other tumour types; no hepatitis B or hepatitis C viral flares were identified,” said the researchers.

Following the study, 41.7 and 47.4 percent of patients in the pembrolizumab and placebo groups, respectively, had subsequent anti-cancer treatment.

The results of KEYNOTE-240 are consistent with those of the phase II KEYNOTE-224*** study. However, there were more Asian patients in KEYNOTE-240 than KEYNOTE-224 and results of the ongoing KEYNOTE-394^# study will inform of the efficacy of pembrolizumab in Asian patients, said Chee.

One potential reason for the longer OS in the placebo arm in this trial compared with previous ones could be the exclusion of patients with main portal vein invasion, said Chee. She also noted that 10 percent of placebo recipients went on to receive anti-PD1 or anti-PD-L1 agents.

Despite the current results, anti-PD1 therapy remains an option for certain patient populations, particularly those intolerant to tyrosine kinase inhibitors, said Chee.