A REVIEW OF THE REPORT

    RAPPORT D'EXPERTISE COLLECTIVE
        INSERM
        

    Effets sur la santé
    des principaux types d'exposition
    à l'amiante

    Rapport établi à la demande
    de la Direction des Relations du Travail
    et de la Direction Générale de la Santé, France

    Paris, juin 1996

    Prepared by
       Graham W. Gibbs, MSc PhD LRSC ROH
    October 1, 1996

    at the request of
    Service du développement minier,
    Ministère des Ressources naturelles
    Gouvernement du Québec

Introduction
The following comments were prepared after a preliminary review of the "rapport de synthèse" and without access to the full technical report. The latter is required to evaluate the scientific basis for some of the choices made by the authors of the report.

While the authors of the report start out with a reasonably careful scientific aproach to describing the available information, insertion of caveats, etc., the criteria on which the impact of asbestos use on health in France is based, appear to receive less rigorous trreatment. There are three main issues, all related to the risk estimation which I believe are the most important. These are...

1. The report does not adequately take into account differences in the risk of mesothelioma associated with the different fibre types.
2. The report does not take into account differences in the lung cancer risks encountered in different industry sectors.
3. The report omits discussion of studies in which workers were exposed at low levels and where no increased lung cancer risks have been reported.

1.   Fibre type is important in mesothelioma
The predicted number of deaths from mesothelioma in workers exposed only to commercial chrysotile is greatly overestimated, because the factor used in the report was not a factor for "chrysotile only".

• The report points out that the K factor used to estimate the number of chrysotile related mesotheliomas was based on exposures "exclusively or principally" to commercial chrysotile.
• The report quite correctly notes that exposure to commercial amphibole fibres carries a considerably higher risk of mesothelioma than exposure to commercial chrysotile. Amphibole exposures are in fact contributing importantly to the risk expressed in the K factor used by INSERM. This means that the predicted number of mesothelioma deaths from exposure to commercial chrysotile is considerably overestimated.

  We know this to be so, because:
  

  1. Dr. Janet Hughes, who derived K factors from 7 Epidemiological studies reportd factors of 0.006 and 0.07 x 10-8 for chrysotile mining and textiles. These are more than 1/10th - 1/100th of the factors used in the INSERM report (1.0 x 10-8).
  2. The K factor applied in the INSERM report appears to be that derived by Peto et al. (1985) for a British textile plant. That textile plant also used crocidolite and in the HEI-AR report (pages 6-15), its limitations for assessing chrysotile mesothelioma risks are noted.
  3. In the Québec cohort of chrysotile miners and millers, with almost 8,000 deaths there have been 28 deaths from mesothelioma in mine and mill workers. This includes dubious cases. On the other hand, the report states that exposure for 40 years at 0.1 f/ml (4 fibre/ml€years) of commercial chrysotile exposure would result in 10 deaths from mesothelioma.
  
  On this basis, it can be estimated that there would be 1000 mesotheliomas deaths in 10,000 men exposed for 40 years at a concentration of 10 fibres/ml. Measurements in the mines and mills, even in the 1970s, averaged well above 10 fibres/ml and miners and millers had cumulative lifetime fibre exposures well above this. It is evident that the mesothelioma risk is being grossly overestimated.

Most of excess mesothelioma risk described in the report is due to amphibole and not chrysotile exposure.

While the present report may reflect an upper end estimate of risk for persons exposed to both commercial chrysotile and amphibole fibres, the report is very misleading in respect to the number of mesotheliomas that would be predicted to occur in chrysotile only industries.

2.   Lung cancer risks at low levels of exposure
All the lung cancer risks in the report are based on linear extrapolation. There are, in fact, several studies showing no increased risk of lung cancer at levels of exposure well above 0.1 fibres/ml. Some of these are...

1. Newhouse and Sullivan (1989), who based on a study of friction manufacturing plant workers concluded that "... with good environmental control chrysotile asbestos may be used in manufacture without causing excess mortality." They found no lung cancer mortality excess.
2. There was no overall excess of lung cancer in asbestos cement plant workers (Ohlsen & Hogstedt, 1985; Gardner et al., 1986; Thomas et al., 1982), all exposed to low levels of chrysotile.
3. A recent study in Québec (Camus & Siemiatycki, 1996) has shown that environmentally and domestically exposed women living in chrysotile mining communities had no excess of lung cancer, even though exposed to an estimated 27 fibre/ml-years of exposure. This estimated exposure is equivalent to 40 years exposure at 0.68 f/ml). The model described in the report would predict 18 excess lung cancer deaths per 10,000 men, based on such exposure from birth.

It is evident that the predicted risks of chrysotile related lung cancer for workers and for the general population are being overestimated. The reason for such an overestimation is the assumption that lung cancer risks associated with mixed commercial fibre and work in textile manufacture apply to workers in all industrial sectors and also to the general population.

3.   Environmental risks of lung cancer
The recent Québec study of women by Siemiatycki et al shows that the general population exposed to chrysotile is not at a detectably increased risk of lung cancer even when exposures have been very high. It should be noted that the exposure of women in Québec was at least 16 times those applied in the INSERM report. The latter concentration (0.025 f/ml) is already at least one and perhpas two orders of magnitude above those reported in buildings including schools in various countries.

A study of persons living in the vicinity of an amosite factory in the USA (Hammond et al. 1979), also failed to find any association between lung cancer mortality and living in the area in which the amosite plant was located.

On this basis, it is highly improbable that the general population is at any detectable increase in the risk of lung cancer due to chrysotile exposures, and if they are, the number of chrysotile related deaths will be orders of magnitude less than predicted in the INSERM report.

4.   Environmental risks of mesothelioma
The HEI-AR report (pages 6-54) reported 0.0005 fibres/ml in schools and 0.0002 fibres/ml for public buildings. They estimated that there would be 10 additional mesothelioma deaths per year if everyone in the USA were to work in asbestos-containing buildings for 20 years at 0.0002 fibres/ml or 13 years in schools at 0.0005 fibres/ml. This increase would not be detectable with existing methods and if real would be only a small fraction of the lifetime background risk of mesothelioma which is about 1 in 5,000.

There are other issues on which you might comment, but their impact is less evident.

5.   Estimated number of deaths from lung cancer
It has been estimated that in the UK, 5-6% of lung cancers are "asbestos-related". Using this factor, the authors of the report estimated the number of lung cancers attributable to "asbestos" exposure in France.

These estimates are crude and assume that the fibre types used, number of factories in the various industry sectors, workforces, smoking habits and exposure levels in France are the same as the UK for example. Is this the case?

6.   Smoking
While there is little doubt that the smoking - asbestos interaction for exposures involving mixed commercial fibre situations is multiplicative. However, the data for chrysotile clearly indicate that the interaction with smoking is less than multiplicative (McDonald et al., 1993).

7.   Substitutes
The report did not consider risks associated with substitutes. This is important, particularly if they are fibrous.

8.   Biopersistence
The report ignores data from the study of synthetic fibres which suggest that fibres of low biopersistence pose a lower lung cancer and mesothelioma risk than those of higher biopersistence.

9.   Different standards for chrysotile and amphiboles
The report suggests that having a less stringent standard for chrysotile would suggest that "chrysotile" is not carcinogenic. This is nonsense. If amphiboles are no longer used in France, it should be possible to deal with the problem of the past using total body protection to handle and remove friable "asbestos" materials. The concern for the future then becomes chrysotile only products.

| Table des matières |