Assaying for gefitinib responder mutations in EGFR (+) colon cancer: Molecular implications on targeted therapeutics

3391

Epidermal growth factor receptor (EGFR) is the target of several recently developed therapeutic agents undergoing clinical trials. One mechanism of action is through inhibition of the receptor’s tyrosine kinase activity. Gefitinib is a small molecule EGFR tyrosine kinase inhibitor that was recently shown to be associated with a remarkable clinical response in a subset of patients with non-small cell lung cancer. Molecular analysis identified mutations in exons 19 and 21 in those patients that responded to the administration of gefitnib. A number of other tumors also express EGFR including adenocarcinoma of the colon, suggesting the possiblity it can and should be treated with gefitinib. However, it is currently unknown if “responder” mutations in exons 19 and 21 exist in colon cancer. Molecular analysis of EGFR(+) colon cancers for these mutations is warranted to identify whether potential responders exist, and to identify if the administration of gefitinib is rational in this population. Seven cases of invasive adenocarcinoma of the colon that expressed EGFR by immunohistochemistry (Dako, Carpenteria, CA) were identified from the archival files of the Roswell Park Cancer Institute. A second corresponding section of tissue was cut and placed on a slide, then deparaffinized, stained with hematoxylin and dehydrated. EGFR expressing tumor cells were identified from the immunohistochemically stained slide and procured from the unstained slide using laser capture microdissection (Arcturus, MountainView, CA). The procured cells were digested and the DNA isolated using spin columns (Zymo Research, Tustin, CA). Primers specific for exons 19 and 21 of EGFR were used for a polymerase chain reaction (PCR). The amplified product was excised from an agarose gel and isolated using spin columns (Qiagen, Valencia, CA). The PCR product was sequenced on an ABI Prism 3100 using Big Dye Terminator kit v3.1. The sequences were then compared by a pairwise BLAST search to the wild type sequence (GI accession #11494376). Amplifiable DNA was recovered in five of the seven cases. All five cases showed conservation of the wild type sequence for both exons. These results show that clinical material that is formalin fixed and paraffin embedded can be analyzed at the molecular level to help define the potential applicability of certain targeted therapeutic agents in specified patient populations. The absence of responder mutations suggest adenocarcinoma of the colon may not respond to gefitinib. In this age of molecular diagnostics, empiric medicine by itself could potentially lead to costly, ineffective therapy. Targeted molecular therapeutics should be preceded by molecular evaluation of targeted sequences.