Genetic classification of colorectal cancer based on chromosomal loss and microsatellite instability predicts survival

Abstract

Purpose: Colorectal cancers harbor one of two distinct alterations, unilateral chromosomal loss as evidenced by a loss of heterozygosity (LOH) and microsatellite instability (MSI), as represented by the widespread insertion or deletion of simple repeat nucleotides. We investigated the relationships between the clinicopathological features and microsatellite alterations (LOH and MSI) of 168 colorectal cancers.

Experimental design: The concerted and individual effects of various chromosomal losses on survival were comparatively analyzed using a reference panel of 40 microsatellite markers in eight cancer-related chromosomes, 3p, 4p, 5q, 8p, 9p, 13q, 17p, and 18q.

Results: Of the 168 colorectal cancers tested, 29 (17%) with high-frequency MSI were associated with good survival (P < 0.05). The extent of LOH detected in 139 (83%) cases without MSI was classified as low level involving three or fewer arms (35%), moderate level involving four arms (22%), or high level involving five or more arms (43%). High-level loss correlated with earlier onset, lymphatic invasion, and rectal location, whereas low-level loss was more common in proximal colon and stages I and II (P < 0.05). The survival curve and multivariate analysis identified high- and low-level chromosomal loss as the most significant predictor of poor and good survival, respectively (log-rank test, P < 0.0001), in patients with stage II (hazard ratio, 6.27; 95% confidence interval, 1.99-19.7; P = 0.0017) and those with stage III (hazard ratio, 10.89; 95% confidence interval, 2.54-46.77; P = 0.0013). Moderate chromosomal loss showed dual prognostic values associated with favorable stage II and unfavorable stage III. Single chromosomal losses tended to play a role as a part of the concerted chromosomal function.

Conclusion: The classification of colorectal cancer based on chromosomal loss and MSI provides a prognostic index that reflects tumor pathobiology.