My main points are that there’s really nothing “new” in the science since the middle 1990s; a comprehensive review ending in 2008 produced no shift; even if there were something new, it has no impact on the need for worker and environmental protection; and that science does support a ban on asbestos because effective protections can’t be implemented.

This is an issue where quantitative treatment of potency can illuminate our understanding of protective measures. The toxic potential, toxic potency, and controls for all forms of asbestos are largely settled questions by now, at least in the developed world (excepting Quebec).

In the developed world, the major problems are legacy of past use, leaving burdens of asbestos removal and illness from past exposures. But continuing use in Eastern Europe and in the developing world is adding to the world’s legacy. The potential of chrysotile asbestos to cause mesothelioma should also be a settled question. Reversing the verdict linking chrysotile to mesothelioma would open an attack on the fraction of asbestos victims in the U.S. who are able to get compensation through the courts.

The “new” interest arises from an EPA effort in the early 2000s to update its Integrated Risk Information System (IRIS) assessment for community exposure to asbestos, probably to address cleanup issues in Libby, Montana. Libby is contaminated with amphibole asbestos from vermiculite mining. The two major forms of asbestos are chrysotile, mined in Quebec and elsewhere; and amphiboles (amosite, crocidolite), mined mostly outside Canada and now rarely used. The EPA effort rekindled longstanding claims that chrysotile asbestos does not cause mesothelioma, and is substantially less potent in causing lung cancer than the amphiboles. These claims were not accepted by any authoritative group.

OSHA RISK ASSESSMENT Following issuance of OSHA’s 1986 standard, which brought the PEL for asbestos from 2 f/ml to 0.2, a lawsuit by the AFL-CIO Building and Construction Trades Department compelled OSHA to reopen the record on a number of issues, including whether the PEL allowed too much disease. On remand, a standard with a PEL of 0.1 f/ml for all forms of asbestos was issued in 1994 based on a quantitative analysis of results of about a dozen mortality studies. The quantitative assessments projected that the risk to workers exposed to the 0.2 f/ml PEL was 6.7 cancers per 1,000 workers, with an additional 5 per 1,000 risk of asbestosis. At 0.1 f/ml the risk was estimated at about 3.4 per 1,000 for cancer, and presumably 2.5 per 1,000 for asbestosis.

Risks above 1 per 1,000 are considered significant in the occupational environment, but lower risks might also be considered significant. (Most OSHA 6(b) PELs, including asbestos, allow much more risk than 1 per 1,000 because of claimed feasibility limitations.)

OSHA’s 1994 explanation for the standard also addressed monitoring. OSHA stated that “the rulemaking reinforces OSHA’s tentative conclusion that workplace asbestos levels of 0.05 f/cc cannot be measured reliably”; OSHA opined that 0.1 f/ml might be the limit of reliable measurement under working conditions. Even though the stated limit of detection for asbestos can be quite a bit lower with high volume or long duration samples in clean environments, the routine detection of significant exposures by current personal sampling techniques is questionable.

Thus, the authoritative statement by OSHA declares a significant risk of cancer and asbestosis at the current PEL, and opines that asbestos exposure can’t reliably be measured by phase contrast microscopy (PCM) at levels approximate to the PEL. Exposures that can’t be measured reliably can’t be controlled reliably. EPA ASSESSMENT EPA’s IRIS cancer risk estimate for asbestos was established in 1993, using data and methods similar to OSHA’s, and projected risks essentially the same as OSHA. EPA focused on much lower risk levels and longer duration of exposure. However, EPA continued pursuing the asbestos risk assessment, commissioning a meta-analysis of epidemiological data initially reported to EPA in 1999 and modified in 2001, 2003, and eventually 2008 in response to comments by the EPA Scientific Advisory Board (SAB). Following a public meeting in 2008, the SAB remained critical of the risk assessment results, and the effort to promulgate a new IRIS assessment for all forms of asbestos was dropped. (Disclosure: I was supported by the American Trial Lawyers Association to prepare comments for the 2008 EPA proceedings on asbestos.)

EPA never adopted the analysis reviewed in 2008. It was published by the contractors, Crump and Berman, in Critical Reviews in Toxicology. Calculation of potency factors from epidemiology studies is a bit of a black box, but what went into the box were several dozen study results for lung cancer and mesothelioma. Lung cancer and mesothelioma potency factors for chrysotile and amphibole asbestos were compared. The controversy was generated by the possibility that EPA might adopt the contractors’ conclusion:

The contractors also noted that potency factors were much more consistent within process type (mining, textile, asbestos cement) than by fiber type.

When these findings were presented for public comment at a meeting of the SAB, most commenters and SAB members rejected the conclusions. The arguments were summarized in the American Journal of Industrial Medicine. Criticisms included the way in which exclusion of particular studies affected the results, and the uncertainty of exposure assessment. These commenters noted that the lung cancer potency factors varied by fifty-fold across individual studies, and mesothelioma potency factors varied by thirty-fold.

The focus on the association or lack of association of chrysotile exposure with mesothelioma has repercussions for lawsuits, but it should not affect public health policy. A study in the British Journal of Cancer established that lung cancer kills at least twice and maybe as many as 10 times as many workers as mesothelioma, and there is no debate that chrysotile exposure is associated with lung cancer. Protection against lung cancer by itself would drive the protective measures now in place. However, it’s also generally believed that mesothelioma victims receive compensation in liability suits, while lung cancer victims are mostly out of luck. So breaking the link between chrysotile and mesothelioma would have been a windfall for defendants.

The association of exposure to all forms of asbestos (including chrysotile) with mesothelioma (and lung cancer) has been stated by at least 12 authoritative or governmental sources. The most recent and prominent assessment is an IARC update published in 2010. In addition to mesothelioma and lung cancer, the working group found sufficient evidence that all forms of asbestos caused cancer of the larynx and ovaries, split on colorectal cancer, and found limited evidence for cancer of the stomach and pharynx. The monograph also summarized laboratory studies. The laboratory studies by inhalation collected by IARC show increased potency (more tumors at comparable exposure levels) of chrysotile compared to amphiboles for lung cancer and mesothelioma. These studies provide no evidence that amphiboles are more potent than chrysotile.

Since the IARC monograph, several dozen additional publications support the IARC conclusions and support the link of chrysotile and mesothelioma. SIGNIFICANT RISK Even if chrysotile is somewhat less potent than amphibole exposure, a significant risk of cancer remains at the current OSHA PEL and the limits of routine measurement. The inability to measure and control dangerous exposures supports a ban on asbestos. The link between chrysotile and mesothelioma has not been broken.