Cancer risk following polyps or cancer of the large bowel in Vaud, Switzerland

We reported a significant excess of colon (but not rectal) cancer following adenomatous polyps of the large intestine in the Canton of Vaud, Switzerland, over the period 1979–1990 (Levi et al., 1993a), with 27 cases observed vs. 10.3 expected. We also found an excess of second primary colorectal cancer and of prostatic cancer following a diagnosis of colorectal cancer (Levi et al., 1993b). Data from several cancer registries (e.g., Connecticut, Hoar et al., 1985; Denmark, Lynge et al., 1985; New South Wales, Australia, McCredie et al., 1997) showed excesses of cancers of the small intestine, breast, uterine corpus, ovary and prostate following cancers of the colon and rectum, pointing to a possible influence of inherited cancer genes and common hormonal, dietary or general lifestyle risk factors. A syndrome of multiple primary adenocarcinomas (predominantly colon, rectum, breast, pancreas, stomach, ovary and endometrium) has been described (Lynch and Smyrk, 1996) as part of hereditary non-polyposis colorectal cancer syndrome, but remains poorly defined (Li, 1996). To provide further quantitative information on the issue, we updated the analysis of the Vaud Cancer Registry over the period 1974–1994. This registry had intermediate colorectal cancer incidence rates (33.9/100,000 males, 22.8/100,000 females, world standard) on a European scale (Levi et al., 1998). It includes data concerning incident cases of malignant neoplasms in the Canton, which has a population, according to the 1990 census, of 601,816 inhabitants. Population-based incidence data on cancer have been available since 1974, whereas data on polyps were available for 1979 and since 1982. The registry is tumour-based, and multiple primaries in the same person are registered separately. Both passive and active follow-up are recorded (Levi et al., 1993a, 1997). After exclusion of 484 colorectal cancer cases detected at death or autopsy and of 101 cases whose colorectal cancer was synchronous (i.e., within 2 months) with another cancer, the present series comprised 5,261 colorectal cancers (World Health Organization, 1976; ICD-9 153–4) diagnosed between 1974 and 1994 and 4,300 polyps (adenomatous, ICD-O 8210/0; villous, 8261/1; and mixed type, 8263/0) diagnosed between 1979 and 1994 (rate of histological verification 97.8% for cancers and 100% for polyps). These persons were followed up to the end of 1994 for the occurrence of any cancer site or type, emigration or death for a total of 41,532 person-years at risk. Only second cancers occurring at a different site of the colon (i.e., right colon including caecum, appendix, ascending colon and hepatic flexure vs. left colon) from the first colon cancer were used to compute incidence. The computation of expected numbers of cases were based on site-, age- and calendar period-specific incidence rates multiplied by the corresponding number of person-years at risk. The significance of the observed-expected ratios (standardized incidence ratio, SIR) and their corresponding 95% confidence intervals were based on the Poisson distribution. Table I gives the observed and expected number of various cancer sites following a diagnosis of adenomatous polyps and colon and/or rectal cancer. After polyps, colon (SIR = 2.03), but not rectal (SIR = 1.13), cancer was observed more frequently than expected. There was a significant excess of cancer of the stomach (20 observed, SIR = 1.86). Two cases of small intestinal cancer were observed vs. 1.1 expected (SIR = 1.91). A significant lack of infiltrating bladder cancer was found (6 observed, SIR = 0.45). For all other cancer sites, the SIRs were not significantly different from unity. Excluding colorectal cancers, a total of 237 cases was observed vs. 243.3 expected (SIR = 0.97). Overall, there were 303 neoplasms observed vs. 284.3 expected (SIR = 1.07). After diagnosis of colon cancer, non-significant excesses were seen for colon (SIR = 1.37) and rectal (SIR = 1.71) cancers. After rectal cancer, non-significant excesses were observed for colon (SIR = 1.52) and kidney (SIR = 1.92). When cancers of the colon and rectum were combined, the increase of colon cancer (SIR = 1.43) was significant, whereas no excess of rectal cancer (SIR = 1.03) was found. SIRs below unity were observed for pancreas (SIR = 0.29) and lung (SIR = 0.70). The present analysis confirms, on the basis of a doubled number of cases (Levi et al., 1993a), that the incidence of cancer of the colon, stomach and, possibly, small intestine (but not rectal cancer) is increased after adenomatous polyp of the large intestine (Atkin et al., 1993; Simons et al., 1992). After a diagnosis of colon cancer, an approximately 50% increased risk of cancer of the large bowel is present, which is similar for colon and rectum. Computation of the expected number of second primary cancer of the colon/rectum is, however, hampered by excision of part of the large intestine in most cases of colon cancer. No appreciable excess was observed for any extra-colonic cancer site. The lack of increases of the commonest adenocarcinomas does not support the inclusion of extra-colonic tumours in the spectrum of the hereditary non-polyposis colorectal cancer syndrome (Lynch and Smyrk, 1996). On account of limited study power, however, increases of less than 50% for breast, corpus uteri or prostate cannot be ruled out. Pancreatic cancer had an SIR significantly below unity, and lung cancer and other tobacco-related neoplasms tended to be, if anything, less frequent than expected after large bowel cancer. The present findings, therefore, are not consistent with the hypothesis that tobacco smoking is associated with colorectal cancer risk (Giovannucci and Martinez, 1996; Tavani et al., 1998). Fabio Levi fabio.levi@inst.hospvd.ch*, Lalao Randimbison*, Carlo La Vecchia , Van-Cong Te*, Silvia Franceschi , * Registre vaudois des tumeurs, Institut universitaire de médecine sociale et préventive, Lausanne, Switzerland, Istituto di Ricerche Farmacologiche “Mario Negri”, and Istituto di Statistica Medica e Biometria, Università di Milano, Milan, Italy, Servizio di Epidemiologia, Centro di Riferimento Oncologico, Aviano, Italy