Clinical and socio-demographic predictors of second colorectal malignancies in 239,527 colorectal cancer survivors: A 25-year retrospective analysis (1973-1998)

Background: The occurrence of a second colorectal cancer (CRC) is of clinical interest because of the increasing numbers of CRC survivors, necessitating a comprehensive analysis of clinical and demographic risk factors for second CRC among colorectal cancer survivors on a large cohort of colorectal patients.

Methods: A retrospective analysis was conducted using data from the National Cancer Institute's Surveillance, Epidemiology, and End Results Program data for the years 1973-1998. Kaplan-Meier survival curves were used to examine the probability of developing a second CRC. Cox proportional hazards models were used to determine the predictors of increased risk of a second CRC.

Results: Of the 239,527 patients, 4.03% had a synchronous or metachronous CRC. The estimated cumulative probability of a second CRC tumor at 1 year is 2.65% rising to 4.17% at 5 years and 5.65% at 10 years. The Cox models indicate that female sex, being married, residence in Atlanta, Hawaii, Utah or New Mexico, having the first primary CRC between 1984-1993 and a “distant” clinical staging were all significantly associated with decreased risk of a second colorectal tumor. An original tumor in the transverse colon, splenic flexure or descending colon was associated with higher incidence of a second CRC. Older age at first diagnosis and residing in Iowa were also associated with a higher risk. Median income of county of residence was not significantly associated with increased risk.

Conclusions: Demographic and disease-specific factors modify the risk of a second CRC suggesting an interplay of lifestyle and biological mechanisms.