Researchers have identified higher levels of hepatocytes positive for pERK immunostaining and greater microvascular invasion as independent prognostic factors of recurrence in patients treated with sorafenib Nexavar for hepatocellular carcinoma.
Roser Pinyol, PhD
According to results reported at the 2017 International Liver Congress in Amsterdam, researchers have identified higher levels of hepatocytes positive for pERK immunostaining and greater microvascular invasion as independent prognostic factors of recurrence in patients treated with sorafenib (Nexavar) for hepatocellular carcinoma (HCC). “The is an unmet need for adjuvant treatment for HCC since HCC recurrence after resection or ablation occurs in 70% of patients within 5 years,” said Roser Pinyol, PhD, a senior post-doctoral fellow, Liver Cancer Translational Research Laboratory, BCLC-IDIBAPS, Liver Unit, Universitat de Barcelona.
The analysis was based on data from the phase III randomized STORM trial that compared adjuvant sorafenib to placebo in patients who had undergone surgical resection or local ablation with curative intent and had an intermediate or high risk of recurrence. The STORM trial randomized 1114 patients to sorafenib or placebo; however, no differences in recurrence-free survival (RFS;P= .26), time to recurrence (TTR;P= .12), or overall survival (OS;P= .48) between treatment arms were observed.
“We conducted BIOSTORM, a biomarker companion study using the STORM study population, to identify biomarkers that independently associated with HCC recurrence and also to identify predictors of recurrence prevention following sorafenib treatment of HCC,” explained Pinyol.
BIOSTORM included samples obtained from HCC patients after resection; of these, 83 patients were on sorafenib and 105 received placebo. Recurrence occurred in 70 cases.
The investigators evaluated the predictive value of an activated MAPK pathway to identify variables that were independently associated with HCC recurrence; 20 samples (11.4%) were positive for pERK staining in hepatocytes.
On univariate and multivariate analysis, only positive versus negative hepatocyte pERK staining (P= .01 andP= .02, respectively), and microscopic vascular invasion (P = .01 in both) emerged as prognostic factors of RFS. Rates of RFS relative to hepatocyte pERK positivity were significantly shorter with sorafenib than placebo. Two patients with pERK-positive hepatocyte staining versus 37 patients with pERK-negative hepatocyte staining were recurrence-free at 40 months post-treatment (HR, 2.46; 95% CI, 1.36 4.44;P= .003).
Positive hepatocyte pERK staining was associated with proliferation gene signatures (P<.001), poor differentiation (P= .01), and Ki-67 positivity (P<.05).
No other variables evaluated, including the presence of liver cirrhosis, Hepatitis C infection, or positive staining for pVEGFR2 were independently associated with RFS.
The study did not identify candidate biomarkers predicting the efficacy of sorafenib to prevent HCC recurrence. “None of the biomarker candidates tested, including pERK, VEGA, pVEGFR2, or previously reported gene signatures, predicted benefit to sorafenib in terms of prevention of HCC recurrence,” said Pinyol.
However, the investigators generated a multi-gene signature that was enriched in sorafenib responders and associated with improved RFS with sorafenib compared to placebo. The molecular characteristics of sorafenib responders included enrichment of JAK2 and other genes involved in KRAS signalling, among others.
Genes indicating a poorer prognosis that were more highly expressed in sorafenib non-responders included genes in the mTOR pathway, the NOTCH pathway, MAKK14, and DDX17.
“The novel 146-gene signature was able to identify 30% of the HCC resected patients as candidates likely to benefit from adjuvant sorafenib in terms of HCC recurrence prevention,” Pinyol commented.
Plots of the predictive value of the 146-gene signature demonstrated improved RFS with sorafenib over placebo (HR, 0.25; 95% CI, 0.05-1.15; P = .075). The opposite was observed in non-responders where RFS was shorter with sorafenib (HR, 1.40; 95% CI, 0.73-2.69; P = .312). The P value of interaction was .040 for RFS in responders versus non-responders. “The 146-gene signature predicts response to sorafenib and requires external validation in HCC cohorts with similar characteristics,” said Pinyol.
Pinyol R, Montal R, Takayama T, et al. Molecular predictors of recurrence prevention with sorafenib as adjuvant therapy in hepatocellular carcinoma: Biomarker study of the STORM phase III trial. Presented at: 2017 International Liver Congress; April 19-23, 2017; Amersterdam, Netherlands. Abstract PS-022.