During a preplanned interim analysis, an independent data monitoring committee recommended the continuation of the phase III METIV-HCC trial, which is assessing the safety and efficacy of tivantinib in patients with hepatocellular carcinoma (HCC) with high MET expression.
The ongoing phase III METIV-HCC study is currently exploring tivantinib in patients with MET-high inoperable HCC who have received prior treatment with sorafenib. The study randomized 368 patients to tivantinib or placebo in a 2:1 ratio. The primary endpoint of the study was overall survival. The interim analysis was conducted following 60% of required events. The next planned analysis will be conducted after 100% of target events have been reported (NCT01755767).
The METIV-HCC study was based on an analysis of the phase II ARQ-197-215 trial, which showed that the risk of death was reduced by 62% with tivantinib compared with placebo in patients with MET-high HCC (HR, 0.38; 95% CI, 0.18-0.81). Conversely, in MET-low tumors, placebo outperformed tivantinib (HR, 1.33; 95% CI, 0.58-3.03). In this analysis, immunohistochemistry (IHC) was an effective test for MET expression, if strict testing criteria were followed.
In the phase II study that was the basis for the biomarker analysis, 107 patients were randomized to tivantinib (n = 71) or placebo (n = 36). This analysis did not restrict enrollment to MET-high individuals but did assess tumor samples for MET expression, with a high expression level defined as ≥2+ in ≥50% of tumor cells.
In the full population of the study, the median time to progression in patients treated with tivantinib was 6.9 weeks compared with 6.0 weeks with placebo (HR, 0.64;P= .04). The median progression-free survival was 1.7 months for the tivantinib group versus 1.5 months for the placebo arm (HR, 0.67;P= .06).
In the biomarker analysis, 77 samples were available for analysis. MET was overexpression more commonly in patients who received prior sorafenib. In this group, MET-high status was seen in 82% of patients compared with 40% for samples taken prior to the administration of sorafenib. Additionally, following sorafenib treatment, just 3 patients (18%) were MET-low.
As a prognostic factor, MET overexpression was associated with significantly worse outcomes. For patients treated with placebo, those with MET-low expression had a median overall survival (OS) of 9 months compared with 3.8 months in the MET-high group (HR, 0.34; 95% CI, 0.13-0.86;P= .02).
A significant interaction was seen between tivantinib and tumor MET levels as a predictive factor for OS (P= .039). The median OS in the placebo arm for patients with MET-low status was 9 months versus 7.2 months in the MET-high tivantinib group (HR, 0.72; 95 % CI, 0.30-1.70;P= .45).