372. ASSESSMENT OF TUMOR REGRESSION GRADE IN ESOPHAGEAL SQUAMOUS CELL CARCINOMA AFTER CHEMORADIOTHERAPY: COMPARISON OF 3 COMMONLY USED SCORING SYSTEMS

Abstract

Abstract Tumor regression grade (TRG) is a measure of histopathological response of cancer to neoadjuvant chemoradiotherapy (nCRT) and is associated with outcomes. Several TRG system are used in esophageal cancer: Becker (0%,1 ~ 10%,10 ~ 50%, >50% vital residual tumor cell [VRTC]), Chirieac (0%,1 ~ 50%, >50% VRTC) and Japan TRG (0%,1 ~ 33%,33 ~ 66%, >66% VRTC). Although these methods are generally accepted, currently there is no common standard. We compared the application of three systems for assessment of TRG. Hematoxylin and eosin–stained slides from 482 resection specimens of esophageal squamous cell carcinoma following nCRT were independently reviewed by three pathologists. Interobserver agreement was analyzed by kappa statistics. The prognostic strength of each TRG system was assessed using the concordance index. Interobserver agreement for the Chirieac system showed the best kappa value (0.93) followed by the Becker (0.90) and JTRG (0.87) system. All three models predicted survival (P < 0.0001) independent of yp-stage in multivariate analysis. Concordance indices of the Chirieac, Becker and JTRG were: 0.627, 0.627 and 0.633, respectively (higher number = better prediction). When comparing with Chirieac system, the JTRG system more accurately predicted survival (p = 0.03) while the Becker system did not (p = 0.12). Chirieac system is most reproducible for the evaluation of esophageal cancer response to nCRT. However, the JTRG system seemed to be more accurate in predicting survival.