Estimating preclinical efficacy targets utilizing cetuximab efficacy in KRAS mutant and wild-type colorectal cancer models.

Background: Clinically relevant targets for developmental drug efficacy in animal models of cancer are critical yet understudied parameters. Materials and Methods: Cetuximab, a chimeric antibody to epidermal growth factor receptor (EGFR), was administered to athymic mice bearing subcutaneous tumors established with 13 human colorectal cancer cell lines of varying biomarker status, defined by DNA sequencing and RT-PCR. Results: If tumor growth inhibition is taken as a target, as is commonly done, then in contrast to the clinical situation where KRAS mutation strongly predicts for a lack of clinically meaningful benefit in colorectal cancer patients, cetuximab alone and in combination with irinotecan-based chemotherapy were efficacious in a similar proportion of KRAS wild-type and mutant models. It was only when tumor regression was utilized to define relevant efficacy that cetuximab monotherapy was efficacious in KRAS wild-type, but not mutant models. Adding cytotoxic therapy to cetuximab treatment increased tumor regression frequency in both genotypes to the point that once again the response was similar for KRAS wild-type and mutant models. Conclusion: Our data support shifting the threshold for claiming clinically relevant targeted therapy efficacy in subcutaneous xenograft models towards tumor regression, rather than tumor growth inhibition, focusing on the evaluation of tumor cells that are addicted to the pathways being targeted.