The distribution of cancer within the large bowel

The distribution of cancer of the anatomical sub-sites of the colon and rectum in Alameda County (California), Bombay, Denmark, Kingston and St Andrew (Jamaica), Miyagi Prefecture (Japan), New Zealand, Norway, Puerto Rico, Saskatchewan (Canada). Singapore, The South Metropolitan Cancer Registry (SMCR) (England), and Sweden were analysed. In this series, the highest rates for large-bowel cancer are found in North America and New Zealand; the lowest in the East (Japan, Bombay and Singapore) and the Caribbean, while intermediate rates prevail in Scandinavia and England. The pattern of sub-site distribution of the cancers throughout the large bowel is similar in areas of high and intermediate risk; namely, there is a decreasing incidence from ascending colon towards the descending colon with a sharp increase in incidence at the sigmoid colon, the rates for rectum being in turn higher than those for the sigmoid. In low-incidence areas, the general pattern is the same, but there may be a deficit of sigmoid cancers. This remains to be proved. Denmark shows a significant excess of lower rectum cancer, particularly in males. The age-standardized incidence rates for rectum cancer in this registry are higher than in any of the others, although the rates for other parts of the large intestine in Denmark are in the intermediate range. This finding should be exploited. With advancing age, the sex ratio for colon cancer as a whole shows a higher male risk; this male excess is, however, confined to the descending and sigmoid colon. The sex ratios for ascending and transverse colon cancer do not change with advancing age, while those for the sigmoid, and to a lesser extent the descending colon, do. Rectal cancer is commoner in males, the male preponderance becoming more pronounced after 65 years of age. For descending, sigmoid and rectal cancers, incidence increases more steeply with age in males than in females, the incidence-age curves on a log-log scale tending to have a downward curvature for the older age-groups. The slopes for ascending and transverse colon cancer do not differ between the sexes and their age-incidence curves have an upward curvature, particularly in women. It makes little difference to the incidence rates whether cancers of the recto-sigmoid are included in the rectum or the colon. The findings are discussed in the light of current aetiological theories. The results are consistent with the sub-site risk being related to the rate of transportation of colonic content. Transportation rates in the variety of risk situations need to be investigated, with particular attention being paid to the sigmoid in low-risk areas