Aims. This study assessed the possibility to build a prognosis predictor, based on non-neoplastic mucosa microarray gene expression measures, in stage II colon cancer patients. Materials and Methods. Non-neoplastic colonic mucosa mRNA samples from 24 patients (10 with a metachronous metastasis, 14 with no recurrence) were profiled using the Affymetrix HGU133A GeneChip. The k-nearest neighbor method was used for prognosis prediction using microarray gene expression measures. Leave-one-out cross-validation was used to select the number of neighbors and number of informative genes to include in the predictor. Based on this information, a prognosis predictor was proposed and its accuracy estimated by double cross-validation. Results. In leave-one-out cross-validation, the lowest number of informative genes giving the lowest number of false predictions (3 out of 24) was 65. A 65-gene prognosis predictor was then built, with an estimated accuracy of 79%. Genes included in this predictor suggested branching signal transduction pathways with possible extensive networks between individual pathways. It also included genes coding for proteins involved in immune surveillance. Conclusion. This study suggests that one can build an accurate prognosis predictor for stage II colon cancer patients, based on non-neoplastic mucosa microarray gene expression measures.


Disease Modeling | Medical Specialties | Multivariate Analysis