Highly Prognostic Signature for Stage II Colon Cancer from the TOSCA Trial: the PROSIT Study.

3524 Background: The predictive and prognostic value of current biomarkers for guiding the treatment of stage II colon cancer (CC) is very limited. Here, we aimed to develop a prognostic transcriptional signature for high-risk stage II CC patients undergoing adjuvant treatment with either 3 or 6 months of FOLFOX4/CAPOX within the framework of the phase III TOSCA Italian trial. Methods: We conducted an observational retrospective study based on the TOSCA trial database, matching 62 patients with high-risk stage II CC who relapsed with 62 who did not. The matching criteria included the center of recruitment, pT stage, adjuvant treatment type and duration (3 or 6 months). Transcriptome sequencing was performed on material from centrally assessed FFPE blocks on an Illumina Novaseq 6000 instrument. Generated data were processed using the nf-core/rnaseq (v 3.14.0) pipeline, aligned to the human genome GRCh37/hg19, and analyzed for differential gene expression, requiring a minimum absolute log fold change of 1 and false discovery rate below 0.1. Enrichment analysis was performed using GSVA package with the z-score method. Results: Out of the 124 stage II CC cases considered for the present analysis, 60% received 3 months and 40% 6 months of adjuvant chemotherapy. The median age was 64 years (IQR 55.2-71.1), with 40% of the cancers being right-sided. Fifty-seven percent of the patients received FOLFOX, while the remaining were treated with XELOX. Of the 124 matched samples, 72 were of sufficient quality for RNA-seq.We identified a signature among the top 17 differentially expressed genes (absolute log fold-change > 1, p-value < 0.001, FDR < 0.1, see Table). Patients exhibiting a lower prognostic signature (PROSIT-L) showed worse disease-free survival (DFS) compared to those with a high prognostic signature (PROSIT-H) (HR 3.5; 95% CI 1.53 – 7.9; p = 0.002), indicating its strong and independent prognostic value, regardless of the treatment received. Conclusions: The new transcriptomic signature may help in identifying stage II CC patients who are likely to relapse despite receiving adjuvant oxaliplatin-based treatment. [Table: see text]