Computer modeling of population exposure to a carcinogen: the example of asbestos and mesothelioma mortality in France

Our plain-English summary of Computer modeling of population exposure to a carcinogen: the example of asbestos and mesothelioma mortality in France, by A. Banaei, B. Auvert, M. Goldberg (Computers and Biomedical Research, an International Journal, 2000).

The study

• Authors: A. Banaei, B. Auvert, M. Goldberg
• Journal: Computers and Biomedical Research, an International Journal
• Published: 2000;33(2):97-109
• Cited by: 3 articles, as counted by NIH iCite
• Read it: PubMed · Publisher

Abstract

The multistage theory of carcinogenesis allows models to be constructed that provide the individual probability of a diagnosed tumor at a given age as a function of the person's past exposure to carcinogenic agents (exposure level versus age), the time since exposure, and age. When exposure to a carcinogenic agent and its impact in terms of morbidity and mortality are modeled on the scale of the general population, individual exposures must often be estimated. If appropriate data do not already exist, this difficult task necessitates expensive and difficult investigations. We propose here a method that allows this global exposure to be modeled without needing to know the individual exposures. The method is used and illustrated in the context of modeling the asbestos exposure of the French male population and calculating its mortality rate from mesothelioma (a type of cancer for which asbestos is the only carcinogen and for which a risk function based on the multistage theory of carcinogenesis exists). This method assumes that the exposure functions (how exposure levels vary with age) for all individuals are the same, with the exception of one parameter. That is, it utilizes a hypothesis that we called the hypothesis of the Standard Exposure Window (SEW). We used two methods to calculate the probability of death from mesothelioma in a representative sample of the French male population for whom individual histories of asbestos exposure are known: we applied the risk function to all the individuals, and we applied the SEW hypothesis. The number of deaths obtained by the two methods are very close and fit the observed mortality data satisfactorily.

In plain English

This was a computer modeling study. The authors used a random sample of 4287 French men, most born between 1905 and 1955, whose recorded work episodes each had an asbestos exposure level in fibers/ml. From this sample they simulated populations, calculated expected mesothelioma deaths among French men aged 50 to 79, and compared these with observed national death counts from 1925 to 1996. They then tested a shortcut, the Standard Exposure Window, which assumes every exposed man started exposure at the same age and was exposed for the same length of time, differing only in exposure level. For the full method, the fit to observed deaths was statistically acceptable for ages 65–69, 70–74 and 75–79 (P = 0.11, 0.13, 0.15) but not for the three youngest age groups (P < 0.01). The authors attributed this probably to particularities in the sample. The shortcut gave results very close to the full method. The authors concluded that it simplified the model without substantially distorting calculated mortality. They noted that it is a strong assumption and that the observed death figures are uncertain.

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: Computer modeling study comparing a reference method (individual exposure histories applied to a risk function from the multistage theory of carcinogenesis) with a simplified method based on the Standard Exposure Window (SEW) hypothesis
• Who was studied: A random sample of the French male population, 4287 men, most born between 1905 and 1955, used to generate simulated populations; results compared with observed mesothelioma deaths among men in France
• Exposure: Asbestos exposure from episodes of exposed work, each with a starting age, duration and exposure level in fibers/ml; fiber types not specified
• Main results:
  • Model-calculated deaths were compared with observed mesothelioma deaths among men in France in 5-year age classes from 50–54 to 75–79, annually from 1925 to 1996
  • Reference method compared with observed deaths: the chi-square test was not significant (P = 0.11, 0.13 and 0.15) for the 65–69, 70–74 and 75–79 age groups, which the authors took as a good fit
  • Reference method compared with observed deaths: the chi-square test was significant (P < 0.01) for the three youngest age groups
  • 30 pairs of standard starting age and exposure duration gave the best fits to the reference method; the authors described the difference between the minimum and maximum of these as negligible
  • The best-fitting pair was starting age 25 with duration 1 year; the starting age was close to the sample mean age at first exposure (roughly 25.5 years among exposed men)
  • The SEW hypothesis reduced the 4536 exposure episodes in the sample (13,608 parameters) to two parameters and one function
• The authors' conclusion: The authors concluded that the SEW hypothesis greatly simplifies the original model and reduces the number of parameters without substantially distorting the calculated mortality, and that it can supply results close to those of the reference method
• Limitations the authors state:
  • Observed mesothelioma deaths were calculated from deaths from pleural cancer, estimated through revisions of the International Classification of Diseases for pre-1978 data, so the figures involve a degree of uncertainty
  • The significant lack of fit in the three youngest age groups was probably due to particularities in the sample, including peaks caused by only nine very highly exposed individuals
  • Because the sample was limited to men born 1905 to 1955, fewer data were available for the older age groups
  • The SEW hypothesis is a strong assumption that does not correspond exactly to reality, and the authors said it should be avoided when a model must describe the phenomenon faithfully

Topics

The National Library of Medicine indexes this study under subjects our site covers: Mesothelioma.

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

Related studies

• Occupations and industries in France at high risk for pleural mesothelioma: A population-based case-control study (1998-2002) — P. Rolland, C. Gramond, et al. (2010)
• War on Carcinogens: industry disputes human relevance of chemicals causing cancer in laboratory animals based on unproven hypotheses, using kidney tumors as an example — R.L. Melnick, J.M. Ward, et al. (2013)
• Modeling Variables With a Spike at Zero: Examples and Practical Recommendations — E. Lorenz, C. Jenkner, et al. (2017)
• Investigation of occupational and environmental causes of respiratory cancers (ICARE): a multicenter, population-based case-control study in France — D. Luce, I. Stücker, et al. (2011)
• Occupational exposure to chlorinated solvents and risk of head and neck cancer in men: a population-based case-control study in France — C. Barul, A. Fayossé, et al. (2017)
• Impact of occupational carcinogens on lung cancer risk in a general population — S. De Matteis, D. Consonni, et al. (2012)

Browse return to medical research.