Asbestos exposure and laryngeal cancer: is there an association?

Medicine is replete with assumptions and myths based on faulty reasoning. It is important for all of us to be aware of this problem and diligent about assessing evidence to draw the best possible conclusions. One of the most frequent errors results from post hoc ergo propter hoc reasoning. This classic error in logic assumes that because Event B happens after Event A, Event B is caused by Event A. Proof of causation requires considerably more rigorous evidence. The suggestion that asbestos exposure can cause laryngeal cancer appears to be an example of this flaw in logic.

It was estimated that in 2008, laryngeal carcinoma would affect 12,250 adults, of whom 3,670 would die. The median age of diagnosis of laryngeal carcinoma is 65 years, and the mean age of death caused by this tumor is 69 years. (1) Known risk factors for laryngeal carcinoma include cigarette smoking and ethanol use. These factors have been shown to have a synergistic effect in the development of carcinoma. In addition, a possible association between asbestos exposure and laryngeal carcinoma has been evaluated throughout the years. However, there are numerous conflicting reports, and a link between asbestos and laryngeal cancer has not been proven.

In 1975, Stell and McGill were the first authors to propose a link between asbestos exposure and laryngeal carcinoma. (2) One hundred nineteen patients with squamous cell carcinoma of the larynx were identified. Compared to agematched controls, a greater proportion of patients had significant asbestos exposure (27.7 vs. 2.5%; p < 0.001), but this study failed to account for risk factors such as smoking. Stell and McGill do mention that the patient group had a greater incidence of smoking than the control group, and they suggest that smoking is a cofactor in the development of laryngeal carcinoma, but the failure to account for known risk factors in their analysis undermines the credibility of the study.

Griffiths and Molony reviewed 60 articles published between 1973 and 2001. (3) They cited postmortem studies by Roggli et al, (4) which noted that asbestos bodies but no dysplasia was found in the larynges of patients who had been exposed to asbestos. Others performing similar studies reviewed by Griffiths and Molony (3) demonstrated hyperplasia but no dysplasia in exposed larynges. Based on these studies, Griffiths and Molony conclude that asbestos is an irritant but that it cannot be deemed a causative agent in laryngeal carcinoma.

Griffiths and Molony also reviewed case-control studies) Of the 14 studies reviewed, 3 (including that of Stell and McGill (2)) were found to have a statistically significant correlation between asbestos exposure and laryngeal carcinoma. However, none of these studies adjusted for the smoking risk associated with laryngeal carcinoma. The remaining studies failed to demonstrate a statistically significant link after adjusting for smoking and alcohol exposure.

A study by Gustavsson et al, not included in the Griffiths study, evaluated 545 patients with laryngeal cancer who admitted to occupational exposure to asbestos. (5) A dose-dependent relationship between exposure and risk of laryngeal carcinoma was demonstrated, with a relative risk of 1.69 (95% CI: 1.05 to 2.74) in the highest exposure group. Griffiths and Molony also reviewed a number of cohort studies with mixed results. Most of the studies that supported a correlation between asbestos exposure and laryngeal carcinoma were criticized for failing to adjust for smoking and alcohol use. Based on their review of all studies, Griffiths and Molony concluded that an association between asbestos exposure and laryngeal carcinoma could not be proven.

Goodman et al performed a meta-analysis incorporating 69 cohort reports from 1967 to 1997. (6) Of these, 4 analyzed laryngeal carcinoma with a latency of at least 10 years after asbestos exposure, and 27 analyzed laryngeal carcinoma without a latency period. The pooled analysis of the studies yielded a meta-SMR (standardized morbidity rate) of 157 (95% CI: 95 to 245; p = 0.23) for studies with a latency period and a meta-SMR of 133 (95% CI: 114 to 155; p= 0.99) for those without latency. Goodman et al concluded that an association between asbestos exposure and laryngeal carcinoma is suggested but not proven.

Browne and Gee performed a review of the literature to analyze 24 prospective studies. (7) The SMR from these studies ranged from 0.4 to 5.5. They concluded that if the studies had been adjusted for smoking and alcohol use, any increase in laryngeal carcinoma seen in patients with asbestos exposure would disappear and that no association exists between asbestos exposure and laryngeal cancer.

Excluding the study by Gustavsson et al, which demonstrated with statistical significance a correlation between high asbestos exposure and laryngeal carcinoma but which remains unconfirmed, no studies have been able to demonstrate a convincing correlation. Many have suggested an increased risk; however, none has been able to withstand criticism regarding study design flaws, particularly lack of adjustment for confounding factors such as cigarette smoking and alcohol consumption. Asbestos bodies have been identified in the larynges of those exposed and may be associated with laryngeal irritation and hyperplasia, but no association between histologically confirmed asbestos bodies and laryngeal dysplasia or carcinoma has been seen.

At present, a causal relationship between asbestos exposure and laryngeal cancer has not been established. None of the studies suggesting a relationship between asbestos and laryngeal carcinoma published to date has controlled adequately for tobacco and alcohol consumption, and the best-designed studies and histologic studies do not show even a convincing association, let alone a causal relationship.

Otolaryngologists should recognize that patients who have been exposed to asbestos (or who have asbestosis) may develop laryngeal carcinoma; but in the absence of evidence, we cannot conclude that their asbestos exposure caused their cancers. As always, we must remain vigilant in distinguishing between the coexistence of two events and a causal link between them. The evidence must be studied carefully whenever a causal link is suggested.

References

(1.) Ries LAG, Melbert D, Krapcho M, et al (eds). SEER Cancer Statistics Review, 1975-2005, National Cancer Institute. Bethesda, MD. http://seer.cancer.gov/csr/1975_2005/, based on November 2007 SEER data submission, posted to the SEER Web site, 2008. Last accessed Sept. 8, 2009.

(2.) Stell PM, McGill T. Exposure to asbestos and laryngeal carcinoma. J Laryngol Otol 1975;89(5):513-17.

(3.) Griffiths H, Molony NC. Does asbestos cause laryngeal cancer? Clin Otolaryngol Allied Sci 2003;28(3):177-82.

(4.) Roggli VL, Greenberg SD, McLarty JL, et al. Asbestos body content of the larynx in asbestos workers. A study of five cases. Arch Otolaryngol 1980;106(9):533-5.

(5.) Gustavsson P, Jakobsson R, Johansson H, et al. Occupational exposures and squamous cell carcinoma of the oral cavity, pharynx, larynx, and oesophagus: A case-control study in Sweden. Occup Environ Med 1998;55(6):393-400.

(6.) Goodman M, Morgan RW, Ray R, et al. Cancer in asbestos-exposed occupational cohorts: A meta-analysis. Cancer Causes Control 1999;10(5):453-65.

(7.) Browne K, Gee JB. Asbestos exposure and laryngeal cancer. Ann Occup Hyg 2000;44(4):239-50.

Sidrah M. Ahmad, BS

Drexel University College of Medicine

Philadelphia

Robert T. Sataloff, MD, DMA, FACS

Editor-in-Chief

EAR, NOSE & THROAT JOURNAL

Professor and Chairman

Department of Otolaryngology-Head and Neck Surgery

Drexel University College of Medicine

Philadelphia