General survey of intertumor linkages that connect the chronological changes of age-adjusted incidence rates of 13 neoplasia types from l975 to l993 in Japan.

We attempted a stochastic study of cancer risk change in time using the follow-up data of the age-adjusted incidence rate (AAIR) of cancer in Japan, which covered 13 neoplasia types of both sexes in scope, and ranged from 1975 to 1993 in time. The purpose of our study was to test whether or not there was any mathematical regularity that was to condition cancer risk changes in time in all the 13 human neoplasia types. We investigated the relation between 2 neoplasias as regards log AAIR changes in time by the direct successive elimination method of Gauss, a fitness test of a given pair data to an equilibrium model. The fitness test was repeated in each of 156 tumor pairs [P(13.2)] in both sexes, in each of 3 (x, y) coordinates - the original (x, y) coordinates, the rect- (x, y) coordinates and the para- (x, y) coordinates. Total number of fitness test in this study was estimated to be 156x2x3=936. The rect- (x, y) coordinates and the para- (x, y) coordinates were defined each as an (x, y) framework with its x axis crossed at a right angle to the regression line of the original log AAIR data, and as another framework with its x axis run in parallel with the regression line of the original log AAIR data. The fitness of a given tumor to an equilibrium system was assessed in terms of the correlation coefficient value r within the range of -1.000 (the oncogene-type equilibrium system) to +1.000 (the tumor suppressor gene-type equilibrium system). Results obtained are given as follows: i) the positivity rates of the fitness test to the oncogene-type equilibrium system and the tumor suppressor gene-type system in the male all-cancer population were each 95.5% (149/156 tumor pairs) and 79.5% (124/156 tumor pairs), and those in the female all-cancer population were each 91.0% (142/156 tumor pairs) and 83.3% (130/156 tumor pairs). Evidence was available to indicate that all of the 13 human neoplasia types of both sexes was associated with both oncogene activation and tumor suppressor gene inactivation at the level of individual tumors. In other words, clearance of both oncogene activation and tumor suppressor gene inactivation was the sine qua non premise of carcinogenesis. ii) The positivity score profiles of a given tumor (profile-like presentation of positivity score for each tumor), for each of 2 cancer genes and for each sex, was highly specific for each of the 26 tumor units (13 tumors of both sexes). The presence of a highly specific positivity score pattern might be taken as another expression of complex interaction of 2 cancer genes in carcinogenesis. iii) Evidence was presented to suggest that specified interactions of the oncogene-tumor suppressor gene complexes of both sexes might be causally related to the emergence of sex discrimination of cancer risk, as testified in a set of tumors with both male dominance of cancer risk and female dominance of cancer risk. iv) The significance of one tumor pair that failed to show fitness to both the oncogene-type equilibrium system and the tumor suppressor gene-type equilibrium system was discussed in terms of spacially restricted dissociation of the power center of the oncogene-type equilibrium system from that of the tumor suppressor gene-type equilibrium system in a given tumor pair.