Invited Perspective: Prioritizing Chemical Testing and Evaluation Using Validated in Vitro Assays Relevant to Key Characteristics

By interfering with hormone action, endocrine-disrupting chemicals (EDCs) can increase the risk of various adverse health outcomes, including cancer and reproductive impairment (La Merrill et al. 2020). In their article, Cardona and Rudel (2021) have identified nearly 300 chemicals that increased estradiol, progesterone, or both in an in vitro steroidogenesis assay that is internationally validated for use in regulatory contexts. They screened publicly available testing data for more than 2,000 chemicals tested in the ToxCast™ high-throughput in vitro steroidogenesis assay in cultured human H295R adrenocarcinoma cells. This U.S. Environmental Protection Agency Tier 1 assay has been used to study chemical impacts on 13 hormones involved in the steroidogenic pathway, including estrogens and progestogens (Haggard et al. 2018; Karmaus et al. 2016). Cardona and Rudel focused specifically on estradiol and progesterone. For the active agents, they systematically compiled available in vivo evidence from databases, authoritative evaluations, and published studies. An important consideration was whether in vivo effects were reported in the rodent mammary gland. This well-studied model system has supported classifications of reproductive toxicants and carcinogens (Rudel et al. 2011). The authors also leveraged information on exposure to provide context for their findings.

In total, 33 of the chemicals that increased estradiol and progesterone in their screening approach have been classified as carcinogenic (Group 1), probably carcinogenic (Group 2A), or possibly carcinogenic (Group 2B) to humans by the International Agency for Research on Cancer (IARC). Notably, this carcinogenicity classification includes five Group 1 agents [diethylstilbesterol, benzidine, 2-napthylamine, nitrosomethylamino-1-(3-pyridyl)-1-butanone (NNK), and pentachlorophenol], five Group 2A agents (captafol, 4-chloro-2-methylaniline, diazinon, malathion, and 3,3′,5,5′-tetrabromobisphenol A), and 23 Group 2B agents (including catechol, 2,4-diaminotoluene, 3,3′-dimethoxybenzidine, methyleugenol, and parathion). Although diethylstilbesterol is a recognized risk factor for breast cancer (IARC 2012a), no information is available on breast carcinogenicity for the other 32 chemicals. Interestingly, some of these carcinogens are linked to different types of cancer in humans [e.g., bladder cancer, non-Hodgkin lymphoma, or prostate cancer (IARC Monographs 2021)]. However, many of the active chemicals have not been tested in vivo or evaluated by IARC, and in total authoritative carcinogenicity evaluations were lacking for 56 and 63 chemicals that increased estradiol or progesterone, respectively. Several agents (e.g., anthracene, bisphenol A, carbaryl, chlorpyrifos, diphenylamine, permethrin) have been recommended for an upcoming IARC evaluation, although few of these recommendations were based on notable evidence regarding breast cancer in humans (Marques et al. 2019).

Further effort in identifying breast carcinogens is urgently needed because female breast cancer has become the most commonly diagnosed cancer type in the world and the leading cause of cancer death in women globally (Ferlay et al. 2021). The chemicals prioritized for further testing and evaluation by Cardona and Rudel comprise a range of use scenarios relevant to exposures in women. Epidemiological studies are an important avenue for elucidating any potential links with female breast cancer. A large percentage of the agents classified by IARC in Group 1 have been identified by epidemiological studies of cancer in exposed workers (Loomis et al. 2018). However, occupational epidemiological studies of cancer in women have proved challenging to conduct and interpret, due to small sample sizes and limitations in exposure characterization. The very few IARC Group 1 agents with “sufficient” evidence of human breast cancer were instead identified wholly or primarily through studies of dietary, medical, or pharmaceutical exposures (IARC 2012a, 2012b, 2012c). In total, only a small fraction of the 528 carcinogens classified by IARC have “sufficient” (6 agents) or “limited” (7 agents) evidence of human breast cancer.

Evidence on carcinogen mechanisms holds promise for shedding light on causes of breast and other cancers. The known or suspected breast carcinogens identified by IARC from epidemiological studies exhibit a range of the 10 key characteristics of carcinogens (Smith et al. 2016). Some modulate receptor-mediated effects (e.g., estrogen–progestogen contraceptives, diethylstilbesterol), others are genotoxic (e.g., ethylene oxide), and several others exhibit many key characteristics of carcinogens (ionizing radiation, tobacco smoking, alcoholic beverages). Cross-sectional and panel studies in occupationally exposed women using end points relevant to key characteristics of carcinogens provide opportunities to identify new carcinogens, although they may not explicitly identify breast carcinogenicity. Indeed, data relevant to key characteristics are directly applicable in hazard identification of carcinogens (IARC 2019; Samet et al. 2020), EDCs (La Merrill et al. 2020), and male and female reproductive toxicants (Arzuaga et al. 2019; Luderer et al. 2019). Expansion of the key characteristics approach to other types of toxicants has been recommended (National Academies of Sciences, Engineering, and Medicine 2017). Efforts are underway to integrate findings relevant to key characteristics from test guideline studies to inform regulatory contexts (Madia et al. 2021).

The approach used by Cardona and Rudel provides an important example of applying evidence from validated in vitro assays relevant to key characteristics to aid prioritization of chemicals for further study and evaluation. Overall, they provide motivation for including such results to help fill gaps in evidence, as an aid to interpreting in vivo findings, to stimulate further research, and for integration with relevant in vivo and epidemiological studies in hazard assessment.

Refers to https://doi.org/10.1289/EHP8608