Predictive value of EGFR gene copy number and K-ras mutation for pathological response to preoperative cetuximab, 5FU, and radiation therapy in locally advanced rectal cancer (LARC)

Abstract

4125 Background: Cetuximab improves activity of chemotherapy in metastatic colorectal cancer (mCRC). Gene copy number (GCN) of EGFR has been suggested to be a predictive factor of response to Cetuximab in patients (pts) with mCRC; moreover K-ras mutation has been associated with resistance to cetuximab in mCRC. Methods: We have designed a phase II study with Cetuximab as single agent weekly for 3 weeks followed by the combination with 5-FU and radiation therapy (50 Gy) as neoadjuvant treatment for LARC. The trial included the analysis of EGFR expression, EGFR GCN and K-ras mutation on diagnostic biopsy. EGFR was determined by immunohistochemistry, EGFR GCN by FISH, K-ras mutation (codons 12 and 13) by PCR-based sequencing. High GCN was defined as EGFR/nucleus ratio > 2.9 or a EGFR/CEP7 polysomy > 4 in at least 50 % of the cells. Tumor regression grade (TRG) was evaluated according to Dworak on surgical specimens. Results: Forty pts have been treated. Median age was 59 years (range 35- 74); median number of cetuximab administrations was 8 (range 1–8) with 70 % of the pts completing the planned doses. Two pts did not undergo to surgery: for refusal (1 pt), for progression of disease (1 pt). A TRG 3 and 4 was achieved in 9 (23.1%) and 3 pts (7.7%) respectively. IHC and FISH data are available on 39 pts. So far K-ras mutation status is available on 26 pts. Twenty-one pts (53.8%) had a high EGFR GCN; moreover 22 pts (56.4%) had a polisomy. The concordance between the two tests was 97 %. Eleven of the 21 pts with high GCN (52.4 %) had a TRG 3–4 compared to 1 out of 17 pts (5.6%) with low GCN (p: 0.0016). Moreover 12 of the 22 pts with high polysomy (54.6%) had a TRG 3–4 compared with no response in the 17 pts with a low polysomy: (p=0.0003). Three of the 14 EGFR (IHC) positive pts (21.4%) had a TRG 3–4 vs. 9 of the 25 (36%) negative pts (p: 0.34). K-ras mutation was found in 5 pts (19%). Pts with K-ras mutation had lower rate of high TRG: 20 % vs. 38.1% (p: 0.45). Conclusions: Our analysis confirms that EGFR amplification is predictive of high TRG to neoadjuvant cetuximab plus 5FU and radiotherapy in pts with LARC. Moreover K-ras mutation is associated with a lower rate of pathological tumor regression. No significant financial relationships to disclose.