Asbestos is regarded as a hazardous substance because it causes cancer. This post explains the medical and epidemiological foundations behind that: which diseases asbestos causes and how they differ (Section 1), why what counts for the risk is not the mere presence of fibres but the cumulative dose over time (Section 2), and why "white asbestos" and "blue asbestos" are not equally dangerous (Section 3). This is the general, supra-regional part. The Austrian findings, the mesothelioma clusters and the question of undercounting are covered by a second part; the concrete case is documented on the Burgenland page. Every statement is backed by primary sources; the full citations are at the end.
Mesothelioma and lung cancer: two diseases, one substance
Asbestos causes several diseases. At the centre of this post are two cancers that are often confused although they differ fundamentally: mesothelioma and lung cancer. Beyond these, asbestos also causes non-malignant conditions, above all asbestosis, a scarring (fibrosis) of the lung tissue after high exposure, and benign changes of the pleura (pleural plaques); the evidence is also regarded as sufficient that asbestos causes cancer of the larynx and the ovary (Wolff et al. 2015). These further conditions are only named here, not treated in depth.
Mesothelioma is a tumour of the mesothelium, the thin cell layer that lines the chest and abdominal cavities. It arises overwhelmingly at the pleura, the lining of the chest, and more rarely at the peritoneum, the lining of the abdomen (Robinson et al. 2005). It is the signature disease of asbestos. Few common cancers have so direct a causal relation to a defined carcinogen as mesothelioma has to asbestos, a closer one even than lung cancer to tobacco smoke (Robinson et al. 2005). It was first described as a distinct, asbestos-related disease in 1960, on the basis of 33 histologically confirmed cases from the north-western Cape Province of South Africa, almost all linked to the local crocidolite deposits (Wagner et al. 1960). Mesothelioma is not curable. For a long time the median survival from diagnosis was about 9 to 12 months (Robinson et al. 2005); the then-standard chemotherapy of pemetrexed and cisplatin extended survival in its approval trial by barely three months, at a response rate of 41 % (Vogelzang et al. 2003). The immunotherapy of nivolumab and ipilimumab, now used first-line, raised median survival in the CheckMate 743 trial to 18.1 months against 14.1 months under chemotherapy (Baas et al. 2021); it too does not cure the disease. Unlike lung cancer, mesothelioma does not depend on smoking, presumably because the carcinogens in tobacco smoke do not reach the mesothelium (Robinson et al. 2005).
Asbestos-related lung cancer, by contrast, is an ordinary bronchial carcinoma that differs neither in tissue nor in course from one caused by smoking. Lung cancer is multifactorial: by far the most common cause is smoking, and asbestos is one of several possible additional causes. The increase in risk from asbestos is dose-dependent, and asbestos and smoking reinforce one another. Exactly how this interplay is to be described is not conclusively settled (see the fold-out below). For attributing a particular lung cancer to asbestos exposure, the Helsinki criteria have become established: as a guideline value, a cumulative exposure of 25 fibre-years doubles the lung-cancer risk (Wolff et al. 2015).
The difference is thus twofold. First, location and specificity: mesothelioma arises in the mesothelium and is nearly conclusive evidence of a past asbestos exposure, whereas lung cancer arises in the lung tissue and has many causes, of which asbestos is only one. Second, and less well known, the magnitude: in public perception mesothelioma stands in the foreground, yet asbestos causes more lung-cancer deaths than mesothelioma deaths overall. From the follow-up of 55 cohorts of asbestos-exposed workers, McCormack et al. (2012) estimated the ratio of additional asbestos-related lung-cancer deaths to mesothelioma deaths, by fibre type, at about 0.7 for crocidolite, about 4 for amosite and, with great estimation uncertainty, about 6 for chrysotile; across all fibre types except crocidolite, asbestos claims at least twice as many lives through lung cancer as through mesothelioma (McCormack et al. 2012). The high ratio for chrysotile follows from the small number of chrysotile-related mesotheliomas in the denominator, not from a particularly high lung-cancer potency of chrysotile. Mesothelioma is thus the more specific consequence, lung cancer, except under crocidolite exposure, the quantitatively weightier one.
For experts (or the expert-curious): how asbestos transforms the mesothelium, and why the latency is so long
How a mineral fibre triggers cancer in the mesothelium cannot be traced to a single mechanism. Robinson et al. (2005) name four plausible routes that do not exclude one another. First, mechanical irritation: long, thin fibres penetrate deep into the lung, can pierce the lung epithelium and damage the mesothelial surface in repeated cycles of injury, repair and local inflammation, which can lead to scarring (plaques) or to cancer. Second, disturbance of mitosis: fibres can sever or pierce the mitotic spindle and so produce aneuploidy and further chromosomal damage that characterise mesothelioma. Third, the formation of reactive, iron-mediated oxygen species that cause DNA damage and strand breaks. Fourth, persistent kinase-mediated signalling, for instance the activation of the mitogen-activated protein kinases and the extracellular signal-regulated kinases 1 and 2 (Robinson et al. 2005).
At the molecular-pathological level, mesothelioma is unusual. What dominates is not the loss of the tumour suppressors p53 or Rb that is common in many tumours, but the loss of P16INK4A, P14ARF and NF2 (Robinson et al. 2005). This specific signature of tumour-suppressor loss is regarded as characteristic of mesothelioma development.
The latency, that is, the time from first exposure to disease, is extraordinarily long. In the first roughly 10 to 15 years after first exposure the risk is practically zero; thereafter it keeps rising for life, with typical latencies of more than 30 years (Robinson et al. 2005). A prospective British cohort determined a median latency of 22.8 years (95 % confidence interval 16.0 to 27.2 years), in women about 29 % longer than in men (Frost 2013). The less obvious conclusion: because mesothelioma is unrelated to smoking, because the risk does not recede after exposure and rises over decades, the disease is an almost pure image of past asbestos exposure.
Taking Burgenland as an example: because decades lie between exposure and disease, a present-day burden from asbestos-bearing rock will only show up as possible illness in the distant future, and conversely cases observed today reflect the exposures of earlier decades. The second part sets out what is already measurable in Austria in terms of disease, and what follows from it.
For experts (or the expert-curious): the dose-response in lung cancer, the 25-fibre-year threshold, and the dispute over the smoking synergy
Unlike mesothelioma, a single lung cancer cannot be identified as asbestos-related from the tissue. Attribution is therefore made via the exposure. The Helsinki criteria define, as a guideline value, that a cumulative exposure of 25 fibre-years doubles the lung-cancer risk (Wolff et al. 2015); this measure, the fibre-year, is the subject of the following section.
Asbestos and smoking act over-additively together, that is, more strongly than the mere sum of their individual effects. Exactly how is disputed. An analysis of 23 studies on the joint effect found no general departure from the multiplicative model (Lee 2001), in which the relative risks multiply. A more recent meta-analysis of 17 studies, by contrast, draws an over-additive but not strictly multiplicative picture: the additive synergy index was 1.44 (95 % confidence interval 1.26 to 1.77), the multiplicative index 0.91 (0.63 to 1.30); for the cohort studies the values were 1.11 and 0.51 respectively (Ngamwong et al. 2015). What remains undisputed is both: the combination is markedly more dangerous than either factor alone, and its exact mathematical form is not conclusively settled.
The quantitative burden thus shifts toward lung cancer. In a population with 2 mesothelioma deaths per 1000 deaths at ages 40 to 84, for instance US men, McCormack et al. (2012) estimated the share of all lung-cancer deaths attributable to mixed asbestos, the population attributable fraction, at 4.0 %. For chrysotile, the asbestos most used worldwide today, the number of asbestos-related lung cancers can, however, be estimated only uncertainly from the few mesothelioma deaths (McCormack et al. 2012).
Taking Burgenland as an example: whether and how strongly an environmental burden counts is a question of the cumulative dose and not of the mere presence of individual fibres. The measure for it, the fibre-year, and the question of what low background burdens mean, are the subject of the following section.
Acute and background: why the dose matters, not the single fibre
Asbestos exposure occurs at very different levels. At one end stands the acute, high exposure at the workplace, for instance of insulation workers, miners or remediation workers who handled dense fibre dust over years. At the other end stands the low background burden of the environment: asbestos is ubiquitous in industrialised areas, and almost everyone who lives there carries asbestos fibres in the lung (Robinson et al. 2005). The decisive question is therefore rarely whether fibres are present at all, but how many were inhaled over what time.
To make this cumulative dose measurable, occupational medicine uses the fibre-year. A fibre-year is the product of fibre concentration and exposure duration: it corresponds to the exposure to one million fibres per cubic metre (10⁶ F/m³, that is one fibre per cubic centimetre) over a working year, namely eight hours on 240 working days (DGUV 2013). The fibre-year is thus the asbestos counterpart to the "pack-year" in smoking: not the snapshot of a concentration, but the burden summed over time.
In the occupational-disease framework a particular cumulative dose serves as a threshold. A lung cancer is recognised as occupational disease No. 4104, among other conditions, when exposure to a cumulative asbestos-fibre-dust dose of at least 25 fibre-years is demonstrated (DGUV 2013). This value is a doubling dose: at 25 fibre-years the lung-cancer risk is doubled (Wolff et al. 2015). A German case-control study confirmed this empirically; its results are compatible with a doubling of the lung-cancer risk at 25 fibre-years (Pohlabeln et al. 2002).
Below this recognition threshold the risk does not, however, vanish. For chrysotile-caused lung cancer the exposure-response relationship is approximately linear, and no indication of a threshold value was found (Stayner et al. 1997). Because no safe effect threshold can be demonstrated, precautionary regulation does not work with a "harmless" limit value but on a risk basis: it assigns an acceptance risk and a tolerance risk each a substance-specific air concentration (TRGS 910). The 25 fibre-years are accordingly a doubling dose for recognition, not a value below which asbestos would be harmless.
For experts (or the expert-curious): the fibre-year, occupational disease BK 4104, and the limits of a high-dose measure
The fibre-year is defined as a product, formally 25 fibre-years = 25 · 10⁶ [(fibres/m³) · years] (DGUV 2013). A cumulative dose of one fibre-year can therefore be reached in various ways, for instance through one working year at 10⁶ F/m³ or through ten working years at 10⁵ F/m³; what is decisive is solely the product of concentration and duration. The occupational fibre-year is thereby normalised to eight hours on 240 days; a continuous environmental burden over a full calendar year yields, at the same concentration, about four and a half times as much, which must be kept in mind when transferring to local-resident exposures. Only respirable fibres that have been released from the material are counted: by the recognised criteria, fibres with a length over 5 µm, a diameter under 3 µm and a ratio of length to diameter greater than 3:1.
The empirical basis of the 25-fibre-year threshold is provided, among others, by the two-phase case-control study of Pohlabeln et al. (2002): the odds ratios adjusted for the fibre-year categories (under 1, 1 to under 10, over 10) were 0.86, 1.33 and 1.94 in the two-phase model, with the value for the highest category reaching statistical significance. The result is compatible with the doubling at 25 fibre-years (Pohlabeln et al. 2002).
That there is no safe threshold below it is shown by the exposure-response analysis of a cohort of chrysotile textile workers: for lung cancer the relationship proved approximately linear with no sign of a threshold value, whereas for asbestosis it ran non-linearly (Stayner et al. 1997). From this absence of a threshold follows the risk-based approach to prevention with acceptance and tolerance concentrations (TRGS 910).
The less obvious conclusion concerns the reach of the measure. The fibre-year and the dose-response curves underlying it come from occupational cohorts with high exposure. The transfer to very low, long-term environmental burdens is subject to considerable uncertainty, because this dose range is barely covered by the underlying studies. The measure describes the occupational world precisely; at its lower edge it becomes extrapolation.
Taking Burgenland as an example: for local residents the question is low, chronic background burden, a dose range far below the occupational cohorts. What the Austrian air measurements there show, and how robust statements in this lower range are at all, is the subject of the second part.
White and blue: why not all asbestos is equally dangerous
"Asbestos" is not a single substance but a collective term for fibrous minerals from two groups. The serpentine group provides, in chrysotile, the "white asbestos", by far the most common and most used asbestos type worldwide. The amphibole group comprises crocidolite ("blue asbestos") and amosite ("brown asbestos") as well as the non-commercially mined but naturally occurring tremolite, actinolite and anthophyllite. All are regarded as carcinogenic to humans and are classified by the International Agency for Research on Cancer in Group 1 (Gualtieri et al. 2018; IARC 2012). In their potency, however, they differ considerably.
For mesothelioma the amphiboles are far more potent than chrysotile. In an evaluation of occupational cohorts Hodgson and Darnton (2000) estimated the relative mesothelioma potency of chrysotile, amosite and crocidolite at a ratio of about 1:100:500; these potency estimates are, however, subject to considerable uncertainty. For lung cancer the potency difference between chrysotile and the amphiboles is, by contrast, markedly smaller, lying roughly in the range of 1:10 to 1:50, yet remaining substantial (Hodgson and Darnton 2000).
From this finding arose the "amphibole hypothesis": the assumption that the mesotheliomas observed in chrysotile-exposed workers go back to contamination with amphiboles and that chrysotile is markedly less dangerous. Stayner et al. (1996) examined it critically. Their finding: mechanistic and lung-burden studies do not convincingly support the hypothesis; chrysotile is associated with an increased risk of lung cancer and mesothelioma; it may be less potent for inducing mesotheliomas than some amphiboles, but for a lower lung-cancer risk there is little evidence. Their conclusion: it is prudent to treat chrysotile with nearly the same concern as the amphibole forms (Stayner et al. 1996). Lower mesothelioma potency therefore does not mean harmlessness.
The reason for the potency difference lies to a substantial part in the biopersistence, the durability of the fibre in the tissue. Chrysotile dissolves in lung fluid markedly faster than amphibole asbestos. In comparative dissolution experiments Gualtieri et al. (2018) determined, for a 0.25 µm thick fibre, a dissolution time of 94 to 177 days for chrysotile, against 49 to 245 years for amphibole fibres; the longer a fibre remains in the lower respiratory tract, the greater the probability that it causes damage (Gualtieri et al. 2018). We cover the geochemical mechanics behind this, why chrysotile in particular dissolves so much faster, in the geology article.
It is precisely the amphiboles that are decisive for environmental burden, since tremolite and actinolite occur naturally in serpentinitic rocks. That an environmental exposure to tremolite, that is, outside the workplace, can trigger mesotheliomas is well documented: in Metsovo in Greece, where a predominantly asbestiform, tremolite-bearing whitewash was used (Langer et al. 1987), and in New Caledonia, where a population-based case-control study found, for the use of a tremolite-bearing whitewash, a mesothelioma risk with an odds ratio of 40.9 (95 % confidence interval 5.15 to 325) (Luce et al. 2000). The wide confidence interval reflects the small number of cases; the direction of the finding is, however, unambiguous. A serpentinite that carries, alongside chrysotile, also tremolite or actinolite is therefore more than "just white asbestos".
For experts (or the expert-curious): the evidence behind the amphibole hypothesis
The lung-burden analyses are unambiguous. Roggli et al. (2002) examined the fibre content of the lung in 312 mesothelioma cases: tremolite was found in 166 cases (53 %) and lay above background level in 81 cases (26 %), chrysotile by contrast in only 32 cases (10 %); Roggli et al. trace this tremolite partly to accompanying chrysotile, partly to talc. In the Quebec chrysotile mining districts, in turn, mesothelioma mortality was higher in the central area, where the tremolite contamination is strongest (McDonald and McDonald 1997). Both support that a substantial part of the mesotheliomas observed under "chrysotile exposure" goes back to amphibole asbestos, above all tremolite.
The conclusion from this is, however, not that chrysotile would be harmless. Stayner et al. (1996) record that the mechanistic and lung-burden evidence does not convincingly support the amphibole hypothesis, that chrysotile is associated with both lung cancer and mesothelioma, and that for a lower lung-cancer risk of chrysotile there is little evidence. Lower mesothelioma potency and harmlessness are two different things.
Taking Burgenland as an example: the southern-Burgenland material carries, by our own findings as well as others', not only chrysotile but also the amphiboles tremolite and actinolite (see the [Burgenland page](/en/asbestos-burgenland/)). It is thus exactly the combination in which the mesothelioma-relevant potency lies not at the lower but at the upper end of the scale. The concrete samples and findings are placed in context by the second part.
For experts (or the expert-curious): biopersistence, why chrysotile dissolves and amphibole remains
Biopersistence is one of the key parameters of fibre toxicity: the longer a fibre withstands the physiological clearance and dissolution processes, the longer it can act in the tissue (Gualtieri et al. 2018). In the same dissolution experiment the amphibole fibres and the fibrous erionite showed, even after 9 to 12 months, only slight signs of dissolution, whereas chrysotile rapidly disintegrates into a silica-rich, amorphous pseudomorph (Gualtieri et al. 2018). The structural cause is that in chrysotile the outer, brucite-like magnesium layer is leached out first, whereas in the amphiboles the cations are built into the resistant silicate framework; we cover this in detail in the [geology article](/en/blog/serpentinite-asbestos-geology/).
Taking Burgenland as an example: the amphibole asbestos detected in the Burgenland material is, by this logic, the more durable and thus the health-wise more critical share; the weighting of this difference and the question of detectability are covered in the second part and on the [standards page](/en/asbestos-standards/).
Sources
- Baas, P., Scherpereel, A., Nowak, A. K. et al. (2021): First-line nivolumab plus ipilimumab in unresectable malignant pleural mesothelioma (CheckMate 743): a multicentre, randomised, open-label, phase 3 trial. The Lancet 397(10272):375–386. doi.org/10.1016/S0140-6736(20)32714-8
- DGUV (2013): BK-Report 1/2013 „Faserjahre". Deutsche Gesetzliche Unfallversicherung, Institut für Arbeitsschutz (IFA), 5. Auflage. ISBN 978-3-86423-042-4.
- Frost, G. (2013): The latency period of mesothelioma among a cohort of British asbestos workers (1978–2005). British Journal of Cancer 109(7):1965–1973. doi.org/10.1038/bjc.2013.514
- Gualtieri, A. F., Pollastri, S., Bursi Gandolfi, N. & Gualtieri, M. L. (2018): In vitro acellular dissolution of mineral fibres: a comparative study. Scientific Reports 8:7071. doi.org/10.1038/s41598-018-25531-4
- Hodgson, J. T. & Darnton, A. (2000): The quantitative risks of mesothelioma and lung cancer in relation to asbestos exposure. Annals of Occupational Hygiene 44(8):565–601. doi.org/10.1093/annhyg/44.8.565
- IARC (2012): Asbestos (Chrysotile, Amosite, Crocidolite, Tremolite, Actinolite and Anthophyllite). In: Arsenic, Metals, Fibres and Dusts. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Vol. 100C. International Agency for Research on Cancer, Lyon. ncbi.nlm.nih.gov/books/NBK304374
- Langer, A. M., Constantopoulos, S. H., Nolan, R. P. & Moutsopoulos, H. M. (1987): Association of Metsovo lung and pleural mesothelioma with exposure to tremolite-containing whitewash. The Lancet 329(8539):965–967. doi.org/10.1016/S0140-6736(87)90305-9
- Lee, P. N. (2001): Relation between exposure to asbestos and smoking jointly and the risk of lung cancer. Occupational and Environmental Medicine 58(3):145–153. doi.org/10.1136/oem.58.3.145
- Luce, D., Bugel, I., Goldberg, P. et al. (2000): Environmental exposure to tremolite and respiratory cancer in New Caledonia: a case-control study. American Journal of Epidemiology 151(3):259–265. doi.org/10.1093/oxfordjournals.aje.a010201
- McCormack, V., Peto, J., Byrnes, G., Straif, K. & Boffetta, P. (2012): Estimating the asbestos-related lung cancer burden from mesothelioma mortality. British Journal of Cancer 106(3):575–584. doi.org/10.1038/bjc.2011.563
- McDonald, J. C. & McDonald, A. D. (1997): Chrysotile, tremolite and carcinogenicity. Annals of Occupational Hygiene 41(6):699–705. doi.org/10.1093/annhyg/41.6.699
- Ngamwong, Y., Tangamornsuksan, W., Lohitnavy, O. et al. (2015): Additive synergism between asbestos and smoking in lung cancer risk: a systematic review and meta-analysis. PLOS ONE 10(8):e0135798. doi.org/10.1371/journal.pone.0135798
- Pohlabeln, H., Wild, P., Schill, W. et al. (2002): Asbestos fibreyears and lung cancer: a two phase case-control study with expert exposure assessment. Occupational and Environmental Medicine 59(6):410–414. doi.org/10.1136/oem.59.6.410
- Robinson, B. W. S., Musk, A. W. & Lake, R. A. (2005): Malignant mesothelioma. The Lancet 366(9483):397–408. doi.org/10.1016/S0140-6736(05)67025-0
- Roggli, V. L., Vollmer, R. T., Butnor, K. J. & Sporn, T. A. (2002): Tremolite and mesothelioma. Annals of Occupational Hygiene 46(5):447–453. doi.org/10.1093/annhyg/mef056
- Stayner, L. T., Dankovic, D. A. & Lemen, R. A. (1996): Occupational exposure to chrysotile asbestos and cancer risk: a review of the amphibole hypothesis. American Journal of Public Health 86(2):179–186. doi.org/10.2105/AJPH.86.2.179
- Stayner, L., Smith, R., Bailer, J. et al. (1997): Exposure-response analysis of risk of respiratory disease associated with occupational exposure to chrysotile asbestos. Occupational and Environmental Medicine 54(9):646–652. doi.org/10.1136/oem.54.9.646
- TRGS 910 (2014): Technische Regeln für Gefahrstoffe 910 „Risikobezogenes Maßnahmenkonzept für Tätigkeiten mit krebserzeugenden Gefahrstoffen". Bundesanstalt für Arbeitsschutz und Arbeitsmedizin (BAuA), GMBl 2014, mit Aktualisierungen. baua.de
- Vogelzang, N. J., Rusthoven, J. J., Symanowski, J. et al. (2003): Phase III study of pemetrexed in combination with cisplatin versus cisplatin alone in patients with malignant pleural mesothelioma. Journal of Clinical Oncology 21(14):2636–2644. doi.org/10.1200/JCO.2003.11.136
- Wagner, J. C., Sleggs, C. A. & Marchand, P. (1960): Diffuse pleural mesothelioma and asbestos exposure in the North Western Cape Province. British Journal of Industrial Medicine 17(4):260–271. doi.org/10.1136/oem.17.4.260
- Wolff, H., Vehmas, T., Oksa, P., Rantanen, J. & Vainio, H. (2015): Asbestos, asbestosis, and cancer, the Helsinki criteria for diagnosis and attribution 2014: recommendations. Scandinavian Journal of Work, Environment & Health 41(1):5–15. doi.org/10.5271/sjweh.3462
Questions on the topic? Free Initial Consultation.