Time-related factors in quantitative risk assessment

Our plain-English summary of Time-related factors in quantitative risk assessment, by K.S. Crump, B.C. Allen, R.B. Howe, et al (Journal of Chronic Diseases, 1987).

The study

• Authors: K.S. Crump, B.C. Allen, R.B. Howe, P.W. Crockett
• Journal: Journal of Chronic Diseases
• Published: 1987;40 Suppl 2:101S-111S
• Cited by: 17 articles, as counted by NIH iCite
• Read it: PubMed · Publisher
• In J-M's library: Johns-Manville's company medical library held a copy (J-M no. 000957).

Abstract

In regulatory or decision-making contexts related to carcinogenic hazards, one would like to know the extra risks associated with various levels, durations, and ages of exposure to a carcinogen. To supply that information, quantitative risk assessments are required that make extrapolations on variables related to dose levels, timing of exposure, and age. Quantitative models that express age-specific mortality rates as functions of the exposure pattern and that allow such extrapolations to be made are presented. The uncertainty inherently associated with those extrapolations is discussed and is found to be exacerbated by small data sets and inadequate data availability. Specific examples are provided that involve asbestos-induced mesothelioma and leukemia caused by benzene exposure.

In plain English

This methods paper described models for estimating lifetime cancer risk from different levels, durations and ages of exposure, using asbestos and benzene as examples. The authors analyzed Seidman's follow-up of World War II amosite workers exposed for an average of 1.5 years. Mesothelioma mortality was 5.1 per 1000 person-years 30 to 35 years after first exposure, falling to 1.8 per 1000 person-years at 35 to 40 years. The authors judged that the regulatory agencies' model, in which risk keeps rising, appeared to overestimate risk after 40 years. For hypothetical school exposure from age 5 to 10 at 0.001 fibers per cubic centimeter of air, lifetime mesothelioma risk was 34 per million with the agency model. Using Seidman-based rates, and assuming a constant rate of 0.11 per million per year beyond 50 years after exposure ended, the risk was 6.2-fold smaller. The authors advised against applying these figures to current schoolchildren. Their main reason was that building asbestos is predominantly chrysotile, which appears less potent for mesothelioma than amosite. The authors concluded that time-related extrapolations carry inherent uncertainty, worsened by small, insufficiently detailed data sets.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Methods paper presenting quantitative risk-assessment models, illustrated by reanalysis of existing cohort data (asbestos-induced mesothelioma and benzene-related leukemia)
• Who was studied: Asbestos example: a cohort of amosite workers exposed during World War II (Seidman cohort), followed about 40 years since exposure. Data from the Selikoff et al. insulation worker cohort were also compared. Benzene example: pooled cohorts studied by Rinsky et al., Ott et al. and Wong et al.
• Exposure: Amosite asbestos, with exposure for an average of 1.5 years and a maximum of 4 years. A hypothetical school exposure was also modeled: age 5 to 10, at 0.001 fibers per cubic centimeter of air. The Ontario Royal Commission had estimated that concentration as a high average for buildings containing asbestos. Occupational benzene exposure was used for the leukemia example.
• Main results:
  • In the Seidman amosite cohort, mesothelioma mortality dropped to 1.8 per 1000 person-years 35 to 40 years after first exposure (2 mesotheliomas in 1086 person-years). This was about one third of the rate of 5.1 per 1000 person-years seen 30 to 35 years after first exposure (8 mesotheliomas in 1577 person-years).
  • Estimated average exposures in the 30 to 35 year and 35 to 40 year groups were almost identical: 47 and 45 fiber-years per cubic centimeter. The authors stated that differences in exposure did not explain the drop.
  • The authors judged that the regulatory agencies' model appeared to overestimate risk occurring more than 40 years after first exposure.
  • A similar reduction appeared in the Selikoff et al. cohort, beginning 50 years after first exposure, compared with 30 years in the Seidman data.
  • For hypothetical school exposure from age 5 to 10 at 0.001 fibers per cubic centimeter, lifetime mesothelioma risk was 34 per million using the agency model (equation 1). Using the Seidman-based rates, plus an assumed constant mortality rate of 0.11 per million beyond 50 years from the end of exposure, the risk was 6.2-fold smaller. The resulting figure is damaged by scanning in the article; see text_problems.
  • If no risk was assumed beyond 50 years from the end of exposure, the risk was 13 times less than with the agency model.
  • In the benzene example, the models using cumulative exposure gave the largest leukemia risk estimates among the eight models compared.
• The authors' conclusion: The authors concluded that, if the reduction in mesothelioma mortality is real, the agency model would considerably overestimate mesothelioma risk from exposures early in life. They also concluded that time-related extrapolations carry uncertainty, compounded by small cohorts and insufficiently detailed published data.
• Limitations the authors state:
  • The authors did not recommend using the school calculations to estimate risk to currently exposed schoolchildren. Their main reason was that building asbestos is predominantly chrysotile, which appears less potent in causing mesothelioma than the amosite in the Seidman cohort.
  • Average exposures for the Seidman cohort were not given directly. The authors had to estimate them using an assumption.
  • The mesothelioma rates showed considerable statistical variability.
  • Fitting a multistage model to the mesothelioma data required ages and timing of exposure and ages at death, which were not available.
  • The benzene cohorts were small, with only 16 leukemias in total. As a result, the models could not be distinguished by how well they fit the data.

Topics

Medical terms in this study, explained in our Lexicon: Carcinogenic, Mesothelioma.

Related studies

• De Novo Solid Tumors After Kidney Transplantation: Is It Time for a Patient-Tailored Risk Assessment? Experience From a Single Center — A. Rossetto, P. Tulissi, et al. (2015)
• Quantitative risk assessment of environmental hazards — S.L. Brown (1985)
• Mesothelioma mortality in asbestos workers: implications for models of carcinogenesis and risk assessment — J. Peto, H. Seidman, et al. (1982)
• Asbestos exposure--quantitative assessment of risk — J.M. Hughes, H. Weill (1986)
• Cancer risks in a historical UK cohort of benzene exposed workers — T. Sorahan, L.J. Kinlen, et al. (2005)
• Chronic infections and inflammatory processes as cancer risk factors: possible role of nitric oxide in carcinogenesis — H. Ohshima, H. Bartsch (1994)

Browse return to medical research.