Identification of chemosensitivity markers for postoperative adjuvant chemotherapy using cellular, molecular, and immunohistochemical profiles of advanced carcinomas from gastrointestinal tract.

Abstract

55 Background: The prediction of chemotherapeutic efficacy is a fundamental problem. We have developed an effective system to identify prediction markers based on an in vitro approach using a quantitative matrix of chemosensitivity assay (CSA) and protein expression by reverse-phase protein lysate microarrays (RPA), followed by tissue microarrays (TMA). In the present study, we attempted to identify predictive markers by extracting candidate proteins from the “chemosensitivity x protein expression matrix” of 12 cancer cell lines and 12 chemotherapeutic agents, followed by examination using TMA in post-operative adjuvant chemotherapy for advanced gastric and colorectal carcinomas. Methods: The “chemosensitivity x protein expression matrix” was constructed using quantitative data from CSA and RPA. TMA was constructed from parafin-embedded, formalin-fixed surgically removed tissues from 30 gastric and 49 colorectal carcinomas. All subjects had received post-operative adjuvant chemotherapy. Expression and/or localization of candidate marker proteins on TMA were evaluated for the comparison of time to relapse (TTR) between the groups based on the score of each candidate marker protein. Results: TTR was compared in each binary score for a candidate marker that showed a moderate to high correlation coefficient of the “chemosensitivity x protein expression matrix” of 144 combinations. Among these candidate combinations, it was found that the presence/absence of NFkB protein nuclear localization and the TTR of those patients who received 5-FU based adjuvant chemotherapy differed significantly. The three-year relapse rate was 28% in NFkB absent cases, whereas it was 84% in the present cases. Conclusions: We identified NFkB nuclear localization as an effective 5-FU based chemotherapeutic prediction marker in the context of gastrointestinal post-operative adjuvant chemotherapy. The combination of CSA, RPA, and TMA technologies may allow rapid identification of chemotherapeutic prediction markers. No significant financial relationships to disclose.