A validated miRNA expression profile for response to neoadjuvant therapy in esophageal cancer.

Abstract

4078 Background: For patients with esophageal adenocarcinoma (EAC) preoperative chemoradiation (trimodality therapy) is the preferred approach. We examined several patient cohorts to create and validate a miRNA signature predictive for response to this therapy in esophageal carcinoma. Methods: Pre-treatment tumor specimens from 10 patients with EAC treated with trimodality therapy were examined using a TaqMan Human MicroRNA Card Set (v3.0) for expression of 754 miRNAs, of which 306 were overexpressed. miRNA expression between patients with a pathologic complete response (pCR) and those with a non-pCR was then analyzed. 44 miRNAs differentially expressed between pCR and non-pCR patients were then examined in an additional 42 EAC patientsÕ tumors (model set) using the Fluodigm 48.48 Dynamic Array and were used to generate a miRNA expression profile (MEP) predicting for response in three groups: low, intermediate and high-risk. The MEP was then validated in an additional 72 EAC patients using the Illumina miRNA Expression Array. ROC curves were generated for clinical stage, MEP and a combination of both. Results: For the model set, pCR rates in low risk patients based on the MEP were 75%, 42.9% in intermediate risk and 12.5% in high-risk patients (ptrend=0.008). In the validation set the pCR rates were 66.7%, 36.5% and 15.7% for low, intermediate and high-risk patients respectively (ptrend=0.025). The AUC for the ROC using the MEP was 0.72 (p=7.8x10-12) in the model set, 0.69 (p=1x10-4) in the validation set and 0.7 for the two cohorts combined (p=1.4x10-18). When clinical stage was added to the predictive model, the ROC for each group was 0.79 (p=2.6x10-19), 0.75 (p=8.7x10-27) and 0.77 (p=2x10-41) respectively. Conclusions: The MEP represents the first validated miRNA signature for therapeutic response in esophageal carcinoma and improves significantly on clinical staging. Its use could select patients at high-risk for therapeutic resistance for trial-based therapy, as well as low-risk patients for whom resection may possibly be safely omitted.