Value of primary tumor gene signatures in colon cancer when national quality standards are adhered to: preliminary results of an international prospective multicenter trial.

ABSTRACT

BACKGROUND: The purpose of this study was to determine if gene signatures are informative in colon cancer (CC) when National Quality Standards (NQS) are adhered to. Several studies have demonstrated the prognostic potential of gene signatures in primary CC. This has never been evaluated prospectively with adherence to NQS. METHODS: This was a prospective, international, multicenter trial. Eligibility criteria were no distant metastasis, ≥12 lymph nodes (LNs), and no adjuvant chemotherapy for LN-negative CC. RNA from frozen tumor samples was considered reliable if RNA Integrity Number >9. Using an Agilent whole human genome array, 44,000 genes were analyzed in primary tumors for differential gene expression (DGE). ANOVA applied at 2-fold expression level was performed in at least 8 experiments to obtain the DGEs. RESULTS: Molecular analysis was completed in 113 of 128 patients. With median follow-up of 27 months, 11.5 % recurred within 3 years after surgery. Significant DGE was identified in recurrent tumors reflected by upregulation (UR) in cellular proliferation and by downregulation (DR) in prodifferentiating panel of 9 genes, independent of T or N classification. By multivariate analysis 3-year disease-free survival was 12.5 % in the UR/DR group versus 93.4 % in the non-UR/DR group (p < .0001; HR = 24.2; 95 % CI 4.8-120.4). CONCLUSIONS: This is the first prospective trial to evaluate gene signatures in CC with adherence to a 12-node minimum quality standard. Certain molecular pathways may be prognostically relevant if both surgery and pathology are standardized, regardless of T or N classification. Careful consideration should be made to include surgical quality measures when planning clinical trials to evaluate the true effect of molecular markers in CC.