Predictive value of proteomic markers for advanced rectal cancer with neoadjuvant chemoradiotherapy.

Background: Preoperative neoadjuvant chemoradiation (nCRT) has been the standard treatment for locally advanced rectal cancer. Serum biomarkers to stratify patients with respect to prognosis and response to nCRT are needed due to the diverse response to the therapy.

Methods: Thirteen paired pre- and post-nCRT sera from rectal cancer patients were analyzed by isobaric tags for relative and absolute quantitation (iTRAQ) method. Twenty-five proteins were selected for validation by parallel reaction monitoring (PRM) in ninety-one patients.

Results: Totally, 310 proteins were identified and quantified in sera samples. Reactome pathway analysis showed that the immune activation-related pathways were enriched in response to nCRT. Twenty-five proteins were selected for further validation. PRM result showed that the level of PZP was higher in pathological complete response (pCR) patients than non-pCR patients. The Random Forest algorithm identified a prediction model composed of 10 protein markers, which allowed discrimination between pCR patients and non-pCR patients (area under the curve (AUC) = 0.886 on testing set). Higher HEP2 and GELS or lower S10A8 in baseline sera were associated with better prognosis. Higher APOA1 in post nCRT sera was associated with better disease-free survival (DFS).

Conclusions: We identified and confirmed a 10-protein panel for nCRT response prediction and four potential biomarkers HEP2, GELS, S10A8 and APOA1 for prognosis of rectal cancer based on iTRAQ-based comparative proteomics screening and PRM-based targeted proteomic validation.