Predicting response to epidermal growth factor receptor inhibitors in colorectal cancer


Session type:

Sabine Tejpar

University Hospital Gasthuisberg, Leuven, Belgium


Predicting response to epidermal growth factor receptor inhibitors in colorectal cancer

In the past 2 years a series of predictive biomarkers for response to EGFR inhibition in CRC have been discovered and incorporated in clinical practice. This process has been extremely interesting, providing us with new clues to the biology of EGFR in CRC, but unresolved questions still hamper the optimal use of these drugs.

It is not yet clear by what mechanism or to what extent CRC is dependent on EGFR signaling. Is this an early and driving event in a subset of tumours with some measure of oncogene dependency, or is this a late event involved in maintenance of established tumours. In the latter case EGFR inhibition may only be effective if combined with other drugs perturbing the robustness of tumour signaling networks. The same insights will aid attempts to overcome secondary resistance.

Mutational activation of the Kras gene has been shown to be a robust negative predictor of efficacy of EGFR targeting monoclonal antibodies in CRC. This can be seen as a downstream activation of the pathway, rendering the tumour independent of the EGF receptor. Our lab has explored the effects of Kras, Braf and Pi3kinase on outcome. We show that not all these putative downstream effectors have the same negative predictive impact on outcome. Recent data has suggested non canonical signaling downstream of Kras in CRC. Also in tumours treated with EGFR inhibitors, we have evidence of heterogeneity downstream of Kras and explore the role of the dual specificity phospatases and guanine nucleotide exchange factor in predicting outcome.

Positive predictors, potentially indicating EGFR dependency of the tumour were explored. Egfr ligands, specifically Epiregulin and Amphiregulin expression strongly correlate with tumour response. We will discuss our work on using the EGFR ligand expression in the primary tumours, alone or in conjunction with the Kras status to predict the outcome of metastatic disease.

Lastly we will discuss the hypothesis generating data from expression profiling studies aimed at detecting novel predictors of EGFR response.