RESUMO
BACKGROUND: Colorectal cancer (CRC) incidence in the USA has increased in adults under age 50. Current CRC surveillance guidelines do not consider age at diagnosis, and there are limited data available on outcomes from surveillance colonoscopies in early-onset CRC (EO-CRC) to guide recommendations on surveillance intervals. AIMS: To compare surveillance outcomes between EO-CRC and traditional-onset colorectal cancer (TO-CRC). METHODS: In a retrospective cohort study in a large tertiary care academic medical center, we collected data on patients with a diagnosis of CRC between 2000 and 2014 who received surgery with curative intent. We used log-rank test and inverse probability of treatment weighted Cox regression analysis to compare the development of metachronous advanced neoplasia (MAN) in patients with EO-CRC (diagnosed ages 18-49) and TO-CRC (diagnosed ages 50-75). RESULTS: Patients with EO-CRC (n = 107) were more likely to present with advanced-stage disease (62% versus 35%, p < 0.0001), rectal tumors (45% versus 27%, p < 0.01), and a family history of CRC (30% versus 16%, p = 0.02) compared to those with TO-CRC (n = 139). Patients with EO-CRC had lower risk of MAN (adjusted HR 0.44, 95% CI 0.22-0.88) than TO-CRC patients. The 5-year event rate for MAN was lower for patients with EO-CRC compared to patients with TO-CRC (5.8% vs. 16.1%, p = 0.07). The presence of synchronous neoplasia or history of diabetes was also predictive of MAN. CONCLUSIONS: EO-CRC was independently associated with a lower risk of developing MAN compared to TO-CRC. Shorter surveillance intervals may not be warranted in EO-CRC; however, large prospective studies are needed.