Abstract
Single-cell studies have discovered abundant cancer-associated genetic/phenotypic changes in non-cancerous cells, strikingly contrasting with the infrequency of cancer. Epidemiological data have revealed decades-long plateaus of breast cancer incidence in the contralateral breast and twins/relatives following the first/proband's diagnosis, unlike the well-known continuous increase of population-level incidence with age, the latter ostensibly attributable to the successive accumulation of multiple genetic/epigenetic changes necessary for transformation. Here, we explain these contradicting observations by differentiating cell-level, individual-level, and population-level evidence. First, we show the same decades-long incidence plateau for renal-cell carcinoma in the contralateral kidney following the first diagnosis, expanding the individual-level evidence from breast cancer. We then consider somatic evolution and cell competition in stem-cell compartments and their bounded nature in ageing as a hypothesized mechanism for the abundant cancer-associated cell-level changes and the prolonged constancy of individual-level incidence. Individual-specific propensity with heritable/familial components underlies this process, with which we show congruence between individual-level's constant incidence vs. population-level's increasing incidence. The resulting postulate, an extension of one by Peto and Mack 25 years ago, distinguishes the last multistage "hit" as the critically rate-limiting event in carcinogenesis. Its supporting evidence calls for a reappraisal of the precise nature of multistage carcinogenesis in cancer biology.