Arguments on carcinogen classification

The purpose is to tell how the Secretary will consider scientific and policy arguments when deciding whether to identify or classify a substance as a potential occupational carcinogen. Under 1990.144 the regulation explains that certain categories of evidence (for example, non‑positive human studies, tumors at the site of administration, metabolic differences, high‑dose animal results, benign tumors, and indirect mechanisms) will be considered only if specific criteria are met, and that evaluations will be made based on scientific and policy judgments.

Non-positive human epidemiologic results will be considered only if they meet strict criteria including long exposure and follow-up, a clear prediction of tumor sites, and sufficient sample size. Specifically, 1990.144(a) requires (i) at least 20 years' exposure and 30 years' observation after initial exposure, (ii) documented reasons predicting the site(s) where cancer would occur, and (iii) an exposed group large enough to detect a 50% increase in cancer at the predicted sites.

To use non‑positive epidemiology to set numerical upper limits on risk you must meet the long exposure and site‑prediction criteria and also provide reliable exposure level data. The rule in 1990.144(a) says criteria (i) and (ii) must be met and, in addition, (iv) specific data on exposure levels must be supplied —either periodic direct measurements during the exposure period or other data providing reliable evidence of exposure magnitude.

Such arguments can be considered only if the route of administration is not oral, respiratory, or dermal and the tumors are clearly related to the material's physical configuration or formulation. 1990.144(b) requires (i) the route is not oral, respiratory, or dermal, and (ii) evidence shows local tumors are caused by physical form (for example crystalline dimensions or implant matrix) and that the same material in a different form does not produce tumors.

Only if you can show the tumor was caused by the physical configuration or formulation and not by the substance itself, and the route was not oral, respiratory, or dermal. 1990.144(b) requires documented evidence that the local tumor is related to the physical form (for example abrasive particles, implant matrix) and that changing the formulation or configuration prevents tumor formation.

You must present complete metabolic profiles for both the test animal species and a representative human group, and show the carcinogenic activity in animals is attributable only to metabolites not produced in humans while other shared metabolites have been tested and found non‑carcinogenic. 1990.144(c) lays out four criteria: (i) complete animal metabolic profile, (ii) complete human metabolic profile for an occupationally representative group, (iii) evidence that the carcinogenic metabolite(s) are produced only in the test species, and (iv) evidence that metabolites produced in humans have been adequately tested and shown not carcinogenic.

A complete metabolic profile means identifying the parent compound and its metabolites (including trace metabolites) and documenting pathways and quantities for the species studied. The regulation 1990.144(c) explicitly requires a complete metabolic profile, including identities of trace metabolites, for the experimental animal species (criterion (i)) and for a representative human population group (criterion (ii)).

Such arguments are considered only if you show high doses cause unique metabolites that are the actual carcinogens and those metabolites are not produced in humans at relevant exposures. 1990.144(d) requires (i) evidence that high doses yield metabolites not seen at lower doses, (ii) evidence those high‑dose metabolites are the ultimate carcinogens while low‑dose metabolites are not carcinogenic, and (iii) evidence those high‑dose metabolites are not produced in humans exposed to low doses.

Best support comes from dose‑response metabolic studies, identification of metabolites by analytical chemistry, and mechanistic toxicology showing which metabolite(s) cause cancer. 1990.144(d) requires documented evidence that addresses dose‑dependent metabolite formation (criterion (i)), carcinogenicity attribution to specific metabolites (criterion (ii)), and demonstration those metabolites are not produced in humans at relevant doses (criterion (iii)).

You must provide robust lifetime bioassay data in multiple species and expert pathological review showing tumors are benign and not likely to progress to malignancy. 1990.144(e) requires (i) data from at least two well‑conducted bioassays in each of two mammal species (or equivalent), (ii) lifetime studies, (iii) slides available and diagnoses by a qualified pathologist who examined each slide and provided diagnostic criteria, and (iv) evidence that the tumor types are known not to progress or are at a benign stage with adequate tissue sectioning to rule out invasion and metastasis.

No — the pathologist's review is required, but must include slide examination, specific diagnostic criteria, and sufficient tissue sectioning to rule out invasion or metastasis. 1990.144(e) requires the relevant tissue slides be made available to OSHA or its designee and that at least one qualified pathologist who personally examined each slide provides specific diagnostic descriptions and criteria.

An indirect mechanism is a pathway where the animal tumors arise through a non‑direct mode of action that would not occur in humans, and such arguments will be considered if evidence shows the mechanism is indirect and inapplicable to human exposure. 1990.144(f) states the Secretary will consider evidence that a positive animal result is not relevant to humans if it demonstrates the tumor‑producing mechanism is indirect and would not occur with human exposures.

Arguments and evidence will be evaluated based on scientific and policy judgments, and will be considered only when they meet the specific criteria in the applicable subsection. 1990.144 explains that the Secretary will evaluate arguments on the listed issues (for example metabolic differences, high‑dose artifacts, benign tumors) according to the prescribed criteria and using scientific and policy analysis.

The exposed group must be large enough that a 50% increase above unexposed controls could have been detected at any predicted site. 1990.144(a) sets criterion (iii) requiring sufficient sample size to detect a 50% increase at predicted tumor sites — the rule does not give a fixed number because required sample size depends on background rates and statistical power considerations, so you must show the study had adequate power for the predicted sites.

Not by itself — you must show high‑dose animal metabolites that cause cancer are not produced in humans at relevant exposures and provide mechanistic evidence linking the effect to those metabolites. 1990.144(d) requires documented evidence that the high‑dose metabolites are the ultimate carcinogens and that such metabolites are not produced in humans exposed to low doses (criteria (ii) and (iii)).

No — simple differences in metabolic rates are not enough; you must present full metabolite identification and show the carcinogenic metabolite(s) are unique to the test species while metabolites shared with humans have been adequately tested and found non‑carcinogenic. 1990.144(c) requires complete metabolic profiles for animals and representative humans plus documented evidence attributing carcinogenicity to species‑specific metabolites (criteria (i)–(iv)).