MET amplification in metastatic colorectal cancer: an acquired response to EGFR inhibition, not a de novo phenomenon.

BACKGROUND: MET amplification appears to be a predictive biomarker for MET inhibition. Prior studies reported a MET amplification rate of 9-18% in metastatic colorectal cancer (mCRC) but do not differentiate increased gene copy numbers due to chromosomal level aberrations from focal gene amplifications. Validation of MET amplification rate in mCRC is critical to this field. RESULTS: In tumor tissue-based analyses, overall MET amplification rate was 1.7% (10/590). MET amplification was seen in 0/103 (0%), 4/208 (1.9%) and 6/279 (2.2%) cases, in cohorts 1, 2 and 3, respectively. Rate of MET amplification in cfDNA of cohort 4 patients refractory to anti-EGFR therapy (n = 53) was 22.6% (12/53) and was significantly higher compared to patients not exposed to anti-EGFR therapy (p < 0.001). MATERIALS AND METHODS: We analyzed MET amplification in mCRC (n = 795) using different methods across multiple cohorts. Cohort 1 (n = 103) and 2 (n = 208) included resected liver metastases and tumor biopsies, respectively, tested for MET amplification using fluorescence in-situ hybridization [amplification: MET/CEP7 ratio ≥ 2.0]. Using another tissue-based approach, cohort 3 (n = 279) included tumor biopsies sequenced with HiSeq (Illumina) with full exome coverage for MET [amplification: ≥ 4 copies identified by an in-house algorithm]. Using a blood-based approach by contrast, cohort 4 (n = 205) included patients in whom the full exome of MET in circulating-free DNA (cfDNA) was sequenced with HiSeq. CONCLUSIONS: Contrary to prior reports, in this large cohort, MET amplification was a rare event in mCRC tissues. In plasma by stark contrast, MET amplification identified by cfDNA occurred in a sizable subset of patients that are refractory to anti-EGFR therapy.

A prospective pilot study of target-guided personalized chemotherapy with intensity-modulated radiotherapy in patients with early rectal cancer.

PURPOSE: To investigate the feasibility of personalizing chemotherapy in patients with rectal cancer. METHODS: Patients with cT3 or cN1 and cM0 rectal cancer were eligible. A set of 6 molecular markers including KRAS, BRAF, and PI3K mutations and expression of topoisomerase-1 (Topo-1), ERCC-1, and thymidylate synthase (TS) using immunohistochemistry were performed in a tumor biopsy. All patients were treated with capecitabine 625 to 825 mg/m/12 h M-F in combination with either irinotecan or oxaliplatin based on Topo-1 and ERCC-1 expression plus either bevacizumab or cetuximab based on the mutation status. All patients received intensity-modulated radiation therapy. A surgery was performed 6 to 8 weeks after the treatment. RESULTS: Fifteen patients (94%) had T3 tumor and 10 (62%) N+ disease of 16 patients enrolled. In all patients, the full set of markers was analyzed within 10 days. Seven patients had K-ras mutation, and 4, 5, and 10 expressed Topo-1, ERRC-1 and TS, respectively. All patients had wild-type BRAF and PI3K tumors. The median time from obtaining informed consent to the treatment period was 18 days and all patients completed the chemoradiation treatment. Fifty percent achieved a complete pathologic response to treatment. Four patients (25%) developed grade 3 proctitis or diarrhea. There were no relevant surgical complications. Sixty-nine percent of the patients received adjuvant XELOX. CONCLUSIONS: The individualization of neoadjuvant chemotherapy in patients with rectal cancer is feasible and leads to a high rate of pathologic response.