Purpose: Metachronous colorectal cancer still occurs in a small percentage of patients, despite colonoscopic surveillance. Cancers in hereditary nonpolyposis colorectal cancer for which there is a high risk of metachronous cancer show distinctive DNA changes termed replication errors (RER+). Ten to 20 percent of sporadic colorectal cancers are also RER+. The aim of this study was to identify factors predictive of metachronous colorectal cancer, despite colonoscopic surveillance. Clinicopathologic characteristics and RER status of cancers were examined.
Methods: Colorectal cancer patients, who entered into a surveillance program of being examined with colonoscopy within six months of surgery and then at intervals of three years thereafter, were reviewed. The 433 patients compliant with the protocol who had had more than one colonoscopy had been followed up for a mean of 3.8 +/- 2.2 years. DNA was extracted from archival paraffin-embedded cancer tissue for determination of RER status.
Results: Ten cases of metachronous cancer were identified, giving a rate of 0.61 percent per year. The site of the index cancer in patients who later developed metachronous cancer was predominantly proximal (P = 0.0007), and these cancers were more likely to have mucinous histology (P < 0.0005). Three of 10 (30 percent) index cancers were RER+, which was not significantly different from unselected series of control colorectal cancers in which 20 of 108 (18.5 percent) were RER+.
Discussion: This study documents the rate of metachronous cancer among patients compliant with a defined colonoscopic screening program and suggests that the risk is highest in patients with a proximal mucinous cancer. RER status does not appear to be a very strong predictive factor, and this study does not support its use as a guide to the frequency of surveillance colonoscopy. More data would be required to determine if RER positivity conferred a relative risk of 3.3 or less.