DNA methylation-based classifier to predict pathological complete response to neoadjuvant chemoradiotherapy in locally advanced rectal carcinomas.

Locally advanced rectal cancer (LARC) patients have been treated with a pre-operative multimodal regimen based on 5-fluorouracil and radiotherapy (nCRT) followed by surgery. Patients that achieve pathological complete response (pCR) present better overall survival and lower rates of recurrence. Substantial evidence indicate that these patients could be spared of surgery, while near 70% of cases present incomplete response (pIR), varying from partial response to stable-disease/progression. Markers capable of distinguishing pIR from pCR have the potential to address alternative treatment strategies. Herein, we evaluated the ability of predicting response to nCRT by DNA methylation analysis in pre-treatment biopsies of LARC. The methylation profiling (EPIC BeadChip platform, Illumina) of 32 LARC biopsies (11 cases with pathological complete response (pCR) to neoadjuvant chemoradiotherapy (nCRT) and 21 with incomplete response, pIR) were compared to five normal rectal tissues. The differentially methylated CpGs were assessed to develop a predictive classifier based on the response to nCRT using diagonal linear discriminant analysis. A classifier based on the DNA methylation status of three CpGs was developed, which is easily applicable to the clinical practice and capable of identifying pCR patients prior to surgery, supporting the decision making on LARC patients management.