Parallel comparison of chronological risk changes among cancers of the breast, the uterine cervix and the stomach, as tested in 9 female populations of the world from early 1960s to mid 1980s: A stochastic study

The purpose of this study was to investigate comparatively the chronological changes of age-adjusted incidence rate (AAIR) as well as age-specific incidence rate (ASIR) of cancers of the breast (Br), the uterine cervix (Cer) and the stomach (St) from early 1960s to mid 1980s (5 data collections) using the data sets of 9 female populations world wide, the total data set number amounting to 45 per tumor. More specifically, we wanted to see whether or not there was any regular relationship between the differential time trends of the above 3 human neoplasias and their oncogene-tumor suppressor gene balances. Our investigation proceeded in 3 steps as follows: step 1, straight line regression analysis of log AAIR was applied to each of tumor pairs Br(x)-Cer(y), Br(x)-St(y) and Cer(x)-St(y). Step 2, direct successsive elimination test (Gauss) of log AAIR was applied to each of 6 tumor pairs (Br-St, Br-Cer, Cer-St, Cer-Br, St-Br and St-Cer) to assess the fitness of a test tumor (x-partaner of fitness test) to either the oncogene type equilibrium model (r=−1.000) or the tumor suppressor gene type equilibrium model (r=+1.000). For each of 6 tumor pairs, the fitness test was repeated 3 times using i) the original data set in the ordinary (x,y) framework, ii) the rect-type data sets in the rect-(x,y) framework, and iii) the para-type data set in the para-(x,y) framework. The fitness of a test tumor (x partner in the calculation) to an equilibrium system was assessed in terms of the correlation coefficient value r within the range of −1.000 to +1.000 (step 3). Staging of the whole carcinogenesis process was attempted for each of 3 human neoplasias using either the ASIR profiles of a high- and low-risk populations, and/or those of early 1960s and mid 1980s. Results of key importance are given as follows: i) Br with rapidly growing cancer risk, Cer with rapidly declining cancer risk and St with slowly declining cancer risk were clearly distinguished from one another by the single regression analysis. ii) The fitness test demonstrated that Br, Cer and St each were found to have one score in the positivity test of oncogene activation (r=−1.000), whereas Br with 4 scores, Cer with 2 scores and St with 3 scores of tumor suppressor gene inactivation (r=+1.000) were found to have differential openness to the fitness test of Gauss for the detection of tumor suppressor gene inactivation. It is indicated that the differential opennesses of a window to tumor suppressor gene inactivation observed in 3 human neoplasias represent a measure of the time-linked aggressiveness of cancer risk. iii) The comparison of 4 ASIR profiles of Br suggested the presence of 3 stages of carcinogenesis progression, of which the 1st stage was tentatively classified as of heredity-dependent origin, and the 2nd and the 3rd stages were classified as of environment-dependent origin. The St ASIR profile represents a fusion product of very minor heredity-dependent stage I and a major environment-dependent stage II. In contrast, the Cer ASIR profile completely lacked a heredity-dependent stage. The integrity of the above findings is discussed in the light of a number of pioneering achievements along the line of the steroid criminal theory of carcinogenesis in humans and in non-human mammals.