Review Occup Environ Med: first published as 10.1136/oemed-2017-104944 on 16 May 2018. Downloaded from Identifying occupational carcinogens: an update from the IARC Monographs Dana Loomis, Neela Guha, Amy L Hall, Kurt Straif IARC Monographs Programme, ABStract have been facilitated by characteristics of the work International Agency for The recognition of occupational carcinogens is important environment that allow cancer occurrence to be Research on Cancer, Lyon, studied, notably well-defined populations that are France for primary prevention, compensation and surveillance of exposed workers, as well as identifying causes of exposed, often at high levels, to agents that can be Correspondence to cancer in the general population. This study updates quantitatively characterised. Analytical methods Professor Dana Loomis, previously published lists of known occupational first developed to study occupational cancer have International Agency for carcinogens while providing additional information also contributed importantly to the development of Research on Cancer, Lyon on cancer type, exposure scenarios and routes, and modern epidemiology.6 69372, France; Identifying occupational carcinogens is an dploomis@ unr. edu discussing trends in the identification of carcinogens over time. Data were extracted from International Agency important research endeavour with broad rele- Received 6 December 2017 for Research on Cancer (IARC) Monographs covering vance to science and public health. Occupational Revised 6 April 2018 the years 1971–2017, using specific criteria to ensure exposure to carcinogens is a major cause of death Accepted 9 April 2018 occupational relevance and provide high confidence in and disability worldwide,7 with an estimated the causality of observed exposure-disease associations. occurrence of 666 000 fatal work-related cancers Selected agents were substances, mixtures or types annually.8 Knowledge of cancer hazards from occu- of radiation classified in IARC Group 1 with ’sufficient pational exposure supports prevention and surveil- evidence of carcinogenicity’ in humans from studies of lance activities, as well as compensation of exposed exposed workers and evidence of occupational exposure workers. However, creating a list of occupational documented in the pertinent monograph. The number carcinogens is not a trivial exercise, as there is of known occupational carcinogens has increased over neither a consensus definition of such agents nor time: 47 agents were identified as known occupational a single, definitive source of all the relevant data. carcinogens in 2017 compared with 28 in 2004. These Doll and Peto3 provided a table of ‘established occu- estimates are conservative and likely underestimate the pational causes of cancer,’ but did not specify the number of carcinogenic agents present in workplaces. methodology by which they were identified. Some Exposure to these agents causes a wide range of 20 years later, Siemiatycki and coauthors9 devel- cancers; cancers of the lung and other respiratory sites, oped a list of ‘definite occupational carcinogens’, followed by skin, account for the largest proportion. drawing on data from the IARC Monographs on The dominant routes of exposure are inhalation and Carcinogenic Risks to Humans published through dermal contact. Important progress has been made 2003 and other sources. The International Agency in identifying occupational carcinogens; nevertheless, for Research on Cancer (IARC) Monographs have http://oem.bmj.com/ there is an ongoing need for research on the causes of been updated since then: more than 120 additional work-related cancer. Most workplace exposures have agents have been evaluated in 36 new volumes; not been evaluated for their carcinogenic potential due furthermore, the methodology for evaluating the to inadequate epidemiologic evidence and a paucity of evidence base has been updated,10 11 and a re-eval- quantitative exposure data. uation of the agents classified as ‘carcinogenic to humans’ in the first 99 volumes has been completed with additional target organ sites identified in the on 22 October 2018 by guest. Protected copyright. process.12 INTRODUCTION Here we provide an updated listing of occu- Historically, much of what was known about pational carcinogens that includes data through the causes of cancer was derived from studies of volume 120 of the IARC Monographs corre- workers. Indeed, an observant 18th-century physi- sponding to the years 1971–2017. We also provide cian’s conclusion that cancer of the scrotum in additional information on tumour type, exposure young chimney sweeps was caused by their occupa- scenarios and exposure routes, identify method- tional exposure to soot, later found to contain poly- ological challenges in compiling such a list from cyclic aromatic hydrocarbons,1 2 is often cited as the available data sources, and discuss trends in the first clear identification of a carcinogen (eg, refs 3 identification of carcinogens over time. 4). With the notable exception of tobacco smoking, most of the other carcinogens that were recognised METHODS during the 19th to mid-20th centuries were discov- As a primary source of data, we used the IARC ered through similar observations.5 Even after Monographs on Carcinogenic Risks to Humans, the To cite: Loomis D, several decades of intensive research beginning world’s most comprehensive encyclopaedia of eval- Guha N, Hall AL, et al. Occup Environ Med Epub in the mid-20th century, nearly half of the ‘estab- uations of carcinogenicity, comprising over 1000 13 ahead of print: [please include lished human carcinogens’ listed in Doll and Peto’s entries. The review and evaluation methods used Day Month Year]. doi:10.1136/ seminal report on the avoidable causes of cancer to develop the IARC Monographs are documented oemed-2017-104944 were occupational in nature.3 These discoveries in the IARC Monographs Preamble.10 Loomis D, et al. Occup Environ Med 2018;0:1–11. doi:10.1136/oemed-2017-104944 1 Review Occup Environ Med: first published as 10.1136/oemed-2017-104944 on 16 May 2018. Downloaded from Briefly, agents are selected for review based on evidence of Review and data extraction human exposure and published scientific data suggestive of Two of us (NG and DL) independently reviewed data for all carcinogenicity. For each agent evaluated, systematic reviews of of the 120 agents classified in Group 1 through October 2017 the available scientific evidence concerning the carcinogenicity in volumes 1–120 of the IARC Monographs to identify entries of the agent in humans and experimental animals are conducted that met the criteria defined above. These determinations were by an international working group of independent experts. reviewed by a third person (KS) and any discrepancies were Each line of evidence is evaluated according to ordered catego- resolved by discussion. For each included agent, we extracted ries that reflect the strength of the evidence of carcinogenicity. data on the cancer sites for which the human evidence was The highest category of ‘sufficient evidence of carcinogenicity’ classified as sufficient, where the classification was established in humans or animals means that a causal relationship between on the basis of epidemiologic studies of workers, and where exposure to the agent and development of cancer has been estab- the occurrence of exposure in workers was documented in the lished. For epidemiological data, ‘sufficient evidence of carcino- monograph. genicity’ is typically based on results from several well-designed, We also summarised agents across six broad classes adapted well-conducted studies where chance, bias and confounding from Cogliano et al12: chemicals; chemical mixtures; metals could be ruled out with reasonable confidence; the conclusion is and metal compounds; airborne particles; airborne complex unlikely to be altered by future studies. Data on human exposure mixtures, and radiation and radionuclides. We grouped arsenic to the agent and toxicological data on pertinent mechanisms of with the metals, although it is now considered to be a metalloid, carcinogenesis are also reviewed. to avoid creating of class containing a single agent. An overall evaluation integrating epidemiological and exper- Information on settings where occupational exposure is imental data is derived according to a structured process that likely to occur, as described in the pertinent monograph, was accounts for the strength of evidence for carcinogenicity in extracted. Primary routes of exposure were also recorded for humans, animals and mechanistic evidence, most notably in agents in categories other than radiation and radionuclides. If exposed humans. Agents with ‘sufficient evidence of carcinoge- the monograph did not provide this information, we consulted nicity’ in humans are assigned by default to the highest category, other sources, most often the NIOSH Pocket Guide to Chemical ‘carcinogenic to humans’ (IARC Group 1) whereas the catego- Hazards.14 ries of ‘probably’ (Group 2A) or ‘possibly’ (Group 2B) carcino- genic to humans, or ‘not classifiable as to its carcinogenicity to humans’ (Group 3) are assigned according to the combined RESULTS AND DISCUSSION strength of the human, animal and mechanistic evidence. Evalu- Counting occupational carcinogens ations may be upgraded to a higher category when the evidence Among the 120 agents classified in IARC Group 1, 70 included for a relevant mechanism of carcinogenesis is sufficiently strong. mention of occupational exposures in