Reply: Erlotinib might be a double-edged sword in HCC

We appreciate the interest of Dong et al. in our recent article.¹ As they point out, hepatocellular carcinoma (HCC) cell lines are largely resistant to erlotinib at clinically relevant doses. These results are consistent with HCC clinical trials where only a minority of patients treated with erlotinib exhibited disease control.^2-3 Our data support these findings, as we observed no effect of erlotinib on tumors in our diethylnitrosamine model as measured by several endpoints including phopsho-EGFR levels, phospho-ERK levels, Ki67 staining, tumor histology and gene-expression signature analysis. Instead, erlotinib had an effect on the surrounding cirrhotic tissue, where suppressed phospho-EGFR levels were associated with reduced hepatocyte proliferation, fewer activated hepatic stellate cells, diminished fibrosis progression and decreased HCC development. Our results suggest that erlotinib does not have appreciable direct anti-tumor effects and instead holds potential for chemoprevention through its effects on cirrhotic and fibrotic liver.

Dong et al. also raise concern over potential resistance of HCC to EGFR inhibition, presumably based on their misfounded notion that the primary effect of erlotinib in our models is on the HCC. We have previously shown that epithelial-to-mesenchymal transition is associated with HCC resistance to erlotinib.⁴ However, various mechanisms of HCC resistance to erlotinib have little if any impact on a strategy to modulate the underlying cirrhotic liver. The EGFR mutations observed in a subset of non-small cell lung cancers are somatic mutations (and rare in HCC⁵). More importantly, germline mutations that could affect hepatocytes and stellate cells are extraordinarily rare⁶. In direct response to the referenced questions, we have not generated any data sets on erlotinib treatment of HCC in animals or HCC cell lines that harbor an EGFR mutation.

The largest clinical data sets that inform erlotinib toxicity profiles are from cancer therapy trials. As pointed out by Dong et al., erlotinib treatment is associated with increased risk of interstitial lung disease⁷ and case reports of fatal complications due to liver toxicity have been described.⁸ The side effects of erlotinib in doses typically used for oncology indications (150 mg/day) renders this dose unsuitable for long term use in chemoprevention. Accordingly, we have launched a clinical trial to determine the minimum effective dose of daily erlotinib in cirrhotic patients (i.e. the minimum dose at which EGFR phosporylation is inhibited in the liver). Other clinical trials examining erlotinib as a prevention strategy have been initiated both as a single agent at chemotherapeutic doses to reduce the incidence of oral cancer in the setting of oral leukoplakia (NCT00402779) and also in combination with sulindac at lower doses to regress duodenal and colorectal adenomas in familial adenomatous polyposis (NCT01187901). We remain cautiously optimistic that erlotinib may be an effective chemoprevention strategy in cirrhosis patients at high-risk for HCC.