Apples to apples: the origin and magnitude of differences in asbestos cancer risk estimates derived using varying protocols

Our plain-English summary of Apples to apples: the origin and magnitude of differences in asbestos cancer risk estimates derived using varying protocols, by D.W. Berman (Risk Analysis, 2011).

The study

• Author: D.W. Berman
• Journal: Risk Analysis
• Published: 2011;31(8):1308-26
• Cited by: 14 articles, as counted by NIH iCite
• Read it: PubMed · Publisher

Abstract

Given that new protocols for assessing asbestos-related cancer risk have recently been published, questions arise concerning how they compare to the "IRIS" protocol currently used by regulators. The newest protocols incorporate findings from 20 additional years of literature. Thus, differences between the IRIS and newer Berman and Crump protocols are examined to evaluate whether these protocols can be reconciled. Risks estimated by applying these protocols to real exposure data from both laboratory and field studies are also compared to assess the relative health protectiveness of each protocol. The reliability of risks estimated using the two protocols are compared by evaluating the degree with which each potentially reproduces the known epidemiology study risks. Results indicate that the IRIS and Berman and Crump protocols can be reconciled; while environment-specific variation within fiber type is apparently due primarily to size effects (not addressed by IRIS), the 10-fold (average) difference between amphibole asbestos risks estimated using each protocol is attributable to an arbitrary selection of the lowest of available mesothelioma potency factors in the IRIS protocol. Thus, the IRIS protocol may substantially underestimate risk when exposure is primarily to amphibole asbestos. Moreover, while the Berman and Crump protocol is more reliable than the IRIS protocol overall (especially for predicting amphibole risk), evidence is presented suggesting a new fiber-size-related adjustment to the Berman and Crump protocol may ultimately succeed in reconciling the entire epidemiology database. However, additional data need to be developed before the performance of the adjusted protocol can be fully validated.

© 2011 Society for Risk Analysis.

In plain English

This single-author article was a comparison of risk-assessment methods, not a new study of people. It compared the EPA's "IRIS" protocol for estimating asbestos cancer risk with the newer Berman and Crump protocols. Both were applied to fiber-size data from amosite, chrysotile, Korean tremolite, crocidolite and anthophyllite dusts (some from earlier rat inhalation experiments, others reference samples not all studied there) and to samples from six U.S. sites. The article also tested how well each protocol reproduced potency values from published worker studies. For many amphibole dusts, Berman and Crump estimates were more than 10 times the IRIS estimates. At the sites, amphibole estimates were about 5 to 100 times higher, with one exception: an El Dorado County school site. For chrysotile, neither protocol was uniformly higher. The newer metrics fit the mesothelioma data better but improved the lung cancer fit only marginally. The author concluded that the protocols can be reconciled. In the author's view, IRIS may substantially underestimate amphibole risk because it used the lowest available mesothelioma potency factor, and may be overestimating chrysotile risk. A newer fiber-size metric was described as promising but preliminary.

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: Comparative risk-assessment analysis of two cancer-risk protocols (IRIS and Berman and Crump), applied to existing fiber-size data and fitted against potency factors from published epidemiology studies; no new human data collected
• Who was studied: No human subjects. Data came from three sources. The first was 13 dusts characterized for fiber size for a 1995 meta-analysis: some came from the rat inhalation experiments of Davis and co-workers, and the others were additional UICC reference samples, not all of which were studied by Davis et al. The second was samples from six U.S. sites (Diamond XX, CA; Waukegan, IL; Southdown Quarry, NJ; Libby, MT; Klamath Falls, OR; an El Dorado County, CA school yard). The third was lung cancer and mesothelioma potency factors from published occupational cohorts.
• Exposure: Airborne asbestos dusts of chrysotile and amphiboles (amosite, crocidolite, Korean tremolite, anthophyllite), counted under different fiber-size metrics. The PCME metrics count fibers longer than 5 μm and wider than 0.25 or 0.4 μm. The Berman and Crump metrics count long fibers; one candidate metric counts fibers longer than 20 μm and thinner than 1.5 μm.
• Main results:
  • Changing the mortality and smoking input data changed the unit risk factors only slightly, by at most about 70% (less than 2-fold)
  • If IRIS had used the maximum or an upper-bound mesothelioma potency instead of the lowest one, its amphibole risk estimates would have been about an order of magnitude larger. The maximum values (KL 0.048, KM 1.2 × 10 to the minus 7) gave a unit risk factor of 2.88, compared with the IRIS value of 0.31 under the same scenario.
  • Counting fibers between 0.25 and 0.4 μm wide (N-PCME) raised IRIS risk estimates by up to a factor of 4 compared with the I-PCME count
  • For many of the amphibole dusts in the dust data set, Berman and Crump risk estimates were more than 10 times the IRIS (I-PCME) estimates
  • For chrysotile dusts, neither protocol gave uniformly higher risk estimates than the other; the result depended on the dust
  • At the environmental sites, amphibole risks from the 2001 Berman and Crump metric were about 5 to 100 times those from IRIS with N-PCME, with one exception (El Dorado County school site, 0.04)
  • For the El Dorado site, the author's leading explanation, not confirmed, was that the amphiboles there may not be true asbestos
  • Across 22 independent tests, IRIS gave a lower amphibole risk than Berman and Crump in all but one, which the author reported as statistically significant (p = 0.045)
  • For mesothelioma, IRIS fit the study-specific potency values poorly. The thin and all-widths metrics fell within or at the edge of the uncertainty bounds for all studies except Quebec chrysotile mining.
  • For lung cancer, the thin and all-widths metrics improved the fit only marginally over the PCME metrics
  • When the 1986 informal model was refitted to current data, the relative potency of chrysotile differed significantly from that of amphibole (p = 0.025 for the thin metric, p = 0.01 for the all-widths metric). The author gave a lower-bound estimate that amphibole asbestos is at least 200 times more potent than chrysotile for mesothelioma.
  • A candidate metric (fibers longer than 20 μm, thinner than 1.5 μm) gave the best agreement with mesothelioma potencies in the four environments where it could be estimated
• The authors' conclusion: The author concluded that the IRIS and Berman and Crump protocols can be reconciled. In the author's view, IRIS may substantially underestimate risk when exposure is mainly to amphibole asbestos, mainly because of an arbitrary choice of the lowest available mesothelioma potency factor. The author judged the Berman and Crump protocol more reliable overall, especially for amphiboles, and considered that IRIS may be overestimating chrysotile risk.
• Limitations the authors state:
  • Results for the new candidate metric are preliminary because data are not yet sufficient to test it across enough studies
  • Neither protocol has yet fully reconciled the chrysotile epidemiology
  • Comparing the size of risk estimates cannot show when risks are being overestimated
  • Uncertainty for the IRIS chrysotile estimate cannot be defined, because it was extrapolated from amphibole data
  • At the environmental sites, risks could be estimated only for some metrics, because the site data did not support the others

Topics

Medical terms in this study, explained in our Lexicon: Epidemiology, Amphibole Asbestos, Mesothelioma, Amphibole Group.

Related studies

• Occupational exposures and cancer: a review of agents and relative risk estimates — C. Marant Micallef, K.D. Shield, et al. (2018)
• Exposures estimates of the Wittenoom mining workforce and town residents - Implications associated with risk estimation for persons exposed to asbestiform riebeckite — A.J. Rogers (2018)
• Comparative Risks of Cancer from Drywall Finishing Based on Stochastic Modeling of Cumulative Exposures to Respirable Dusts and Chrysotile Asbestos Fibers — F.W. Boelter, Y. Xia, et al. (2015)
• Estimates of historical exposures by phase contrast and transmission electron microscopy in North Carolina USA asbestos textile plants — J.M. Dement, D. Myers, et al. (2009)
• Inhalation unit risk (IUR) of asbestos based on available science — A. Korchevskiy, J.O. Rasmuson, et al. (2020)
• Varying risk comparison elements: effects on public reactions — B.B. Johnson (2004)

Browse more research in Risk Analysis, or return to medical research.