Objective: Previous studies have reported a migration in the occurrence of colorectal cancer (CRC) toward a proximal colonic location. To assess the potential impact of this, we evaluated recent temporal trends in the United States.
Methods: Using the nine population-based cancer registries that constitute the Surveillance, Epidemiology, and End Results Program of the National Cancer Institute, we identified primary CRCs diagnosed between 1978 and 1998. Temporal changes were evaluated for 3-yr time periods. Age-adjusted incidence rates and the proportions of new diagnoses of CRC were calculated by site and by race.
Results: We identified 243,861 individuals with CRC during 1978-1998, 51.1% of whom were men. In whites and in blacks, the proportions of new diagnoses of right CRC rose significantly, from 34% and 37% to 40% and 44%, respectively. In contrast, the proportions of left CRC (in whites and blacks) and of rectosigmoid CRC (in whites) decreased significantly over time. For whites and blacks, the age-adjusted incidence rates for right CRC remained unchanged (in whites) or showed a small increase (in blacks), whereas the age-adjusted incidence rates for left CRC (in whites and blacks) and for rectosigmoid CRC (in whites) declined. During 1996-1998, right CRC was the most common site in the oldest age groups (70-79, 80 yr, and older) in whites and in blacks.
Conclusions: The proximal migration of CRC over time is not attributed to a true increase in the incidence of right CRC. It is explained by a decrease in the incidence of distal CRC coupled with the aging of the population. Older individuals, in whom the burden of CRC is greatest, and in whom right CRC is the most common site, would be the most adversely affected from CRC screening methods that do not assess the total colon.