Analysis of HER-3, insulin growth factor-1, nuclear factor-kB and epidermal growth factor receptor gene copy number in the prediction of clinical outcome for K-RAS wild-type colorectal cancer patients receiving irinotecan–cetuximab

Background: A large proportion of colorectal cancer patients does not benefit from the use of anti-epidermal growth factor receptor (EGFR) treatment although in the absence of a mutation of the K-RAS gene. Preliminary observations suggested that HER-3, insulin-like growth factor-1 (IGF-1), nuclear factor-kB (NF-kB) and EGFR gene copy number (GCN) might identify patients not likely to benefit from anti-EGFR therapy. We tested the interaction between HER-3, IGF-1, NF-kB, EGFR GCN and K-RAS mutational analysis to verify the relative ability of these variables to identify a subgroup of patients more likely to benefit from EGFR-targeted treatment among those harbouring a K-RAS wildtype status. Patients and methods: We retrospectively collected tumours from 168 patients with metastatic colorectal cancer treated with irinotecan‐cetuximab. K-RAS was assessed with direct sequencing, EGFR amplification was assessed by chromogenic in situ hybridisation (CISH) and HER-3, IGF-1 and NF-kB were assessed by immunohistochemistry. Results: In patients with K-RAS wild-type tumours, the following molecular factors resulted independently associated with response rate: HER-3 [odds ratio (OR) = 4.6, 95% confidence interval (CI) 1.8‐13.6, P = 0.02], IGF-1 (OR = 4.2, 95% CI 2‐10.2, P = 0.003) and EGFR GCN (OR = 4.1, 95% CI 1.9‐26.2, P = 0.04). These factors also independently correlated with overall survival as follows: HER-3 [hazard ratio (HR) = 0.4, 95% CI 0.28‐0.85, P = 0.008], IGF-1 (HR = 0.47, 95% CI 0.24‐0.76, P < 0.0001) and EGFR GCN (HR = 0.59, 95% CI 0.22‐0.89, P = 0.04). Discussion: We believe that our data may help further composing the molecular mosaic of EGFR-resistant tumours. The role of HER-3, IGF-1 and CISH EGFR GCN should be prospectively validated in clinical trials investigating anti