Abstract This paper integrates two complementary perspectives on the action of carcinogens and risk factors: (i) the macro-epidemiological picture of global incidence and (ii) cohort and mechanistic evidence on tissue-specific effects of exposure. At the global level, GLOBOCAN 2022 estimates 19 976 499 new cancer cases in a population of 7 885 070 781, which corresponds to an annual incidence of 0.253% (“all cancers”, both sexes). Within that relatively narrow annual “baseline flow” of disease, empirically it is seen that strong selectors (e.g., smoking, alcohol, obesity, HPV infection) do not have to multiply the overall volume of incidence by orders of magnitude, but dominantly redistribute the anatomical site of manifestation: in cohorts, for “all cancers” one typically sees an increase on the order of ~1.1–2.5×, while for specific sites (lung, upper aerodigestive tract, liver, endometrium, esophagus–adenocarcinoma, cervix, anus) much larger shifts appear. For example, in a large Japanese cohort (JPHC) smoking is associated with an increased risk for “all cancers” (men HR 1.64, women HR 1.46). In a Danish cohort of patients with alcohol use disorder (AUD), the standardized incidence ratio for “all cancers” is (men SIR 1.9, women SIR 2.5), with pronounced tissue-specific increases (tongue/oral cavity/pharynx/esophagus/liver/larynx/lung) and absence of increase for some other locations (e.g., colorectal, urinary bladder). For BMI, the Million Women Study shows that an increase in BMI carries a relatively modest increase for “all cancers” (RR per +10 BMI units 1.12), but a very large increase for endometrium (RR 2.89) and esophageal adenocarcinoma (RR 2.38). For HPV, a Danish cohort of persons with genital warts (marker of HPV infection) shows a moderate increase for “all cancers” (SIR ~1.3–1.9, depending on the study), but extreme increases for specific sites (anus SIR 21.5 for men, 7.8 for women; vulva SIR 14.8; penis SIR 8.2). Alongside these cohort patterns, the mechanistic level (mutational signatures and biological directions) provides direct evidence of site selection: smoking is associated with signature SBS4 which is dominant in tumors of tissues directly exposed to tobacco smoke; HPV with APOBEC signatures SBS2/SBS13 in HPV-positive tumors. At the same time, there are consistent inverse associations of smoking with endometrial carcinoma (meta-analysis of prospective studies RR 0.81) which can be explained by anti-estrogenic action via AhR–ER interference, which demonstrates that the same “selector” can simultaneously increase risk in one organ and reduce it in another. The paper formalizes a hierarchical model in which (1) the systemic state of population vulnerability determines the baseline annual flow of carcinogenesis, while (2) carcinogens and mass risk factors act as anatomical selectors that channel that flow toward specific tissues through direct exposure, local biotransformation, hormonal and inflammatory axes, and tissue-specific mutational processes. The model is compatible with the observation that overall incidence rates in some registries are longterm relatively stable (e.g., SEER reports stability of age-standardized incidence rates of new cases for “any cancer site” through 2013–2022), while the structure by type can change significantly.
Zakir Causevic (Mon,) studied this question.