The clinical utility of a tumour microenvironment-based histopathological score in patients with primary operable colorectal cancer.

Abstract

664 Background: The tumor microenvironment is increasingly recognized as an important determinant of disease progression and outcome in colorectal cancer (CRC). The aim of the present study was to examine the clinical utility of a novel histopathological prognostic score based on both the tumor inflammatory cell infiltrate and the tumor stroma percentage, termed the Glasgow Microenvironment Score (GMS), in patients with primary operable CRC. Methods: Using routine H&E pathological sections, Klintrup-Mäkinen (KM) grade was graded as strong or weak, and tumor stroma percentage (TSP) was assessed as high (>50%) or low (≤50%) retrospectively in 319 patients undergoing resection of stage II-III CRC. The relationship between a cumulative score based on these factors and cancer-specific survival (CSS) was examined. Results: Median survival of survivors was 122 months with 106 cancer deaths. 63% of patients had stage II CRC. 5-year CSS of patients with stage II and stage III CRC was 82% and 58% respectively (p<0.001). 33% of patients were GMS=0 (strong KM/high or low TSP), 47% were GMS=1 (weak KM/low TSP), and 20% GMS=2 (weak KM/high TSP), with 5-year CSS of 86%, 74%, and 48% (p<0.001) respectively. On multivariate analysis, GMS was associated with CSS (HR 1.77, p<0.001), independent of age, emergency presentation, TNM stage and venous invasion (all p<0.01), and peritoneal involvement (p<0.05). The combination of TNM and GMS stratified 5-year CSS from 92% (stage II, GMS=0) to 35% (stage III, GMS=2) (Table). Conclusions: The present study shows the clinical utility of a novel, routinely available assessment of the tumor microenvironment in patients undergoing curative resection of CRC. [Table: see text]