Cancer morbidity investigations: lessons from the Duluth study of possible effects of asbestos in drinking water

Evaluation of potential gastrointestinal carcinogenicity associated with the ingestion of asbestos

The inhalation of asbestos, depending on the fiber type and dose, may be associated with the development of mesothelioma and other asbestos-related diseases. However, little is known about the potential adverse effects associated with the ingestion of asbestos. Evidence of asbestos fibers released from asbestos-cement pipes used in water distribution systems has led to concerns of potentially contaminated drinking water. The purpose of this study is to determine whether ingestion of asbestos fibers may lead to cancerous effects on the gastrointestinal (GI) tract. Data from animal and human studies were analyzed using a weight-of-evidence approach to evaluate the potential risk of GI cancers associated with asbestos ingestion. Seventeen human and 23 animal studies were identified and evaluated in this study. Animal studies were conducted in multiple species with inconsistent dosing protocols. Overall, animal studies reported that the asbestos fibers, irrespective of fiber type and dose, failed to produce any definitive GI carcinogenic effect. The 17 identified human epidemiological studies reported the ingestion of asbestos-contaminated water with concentrations from 1 to 71,350 million fibers per liter (MFL). A majority of the epidemiology studies reported statistically significant increases in multiple GI-specific cancers. However, these findings are confounded due to several critical study limitations including flawed study design, small sample size, selection bias, lack of individual exposure history, lack of adequate latency, and the inability to account for confounders including occupational history, diet, and smoking history. Based on our weight-of-evidence assessment, there is insufficient evidence of causality between the ingestion of asbestos and an increased incidence of GI cancers.

Particle toxicology and health - where are we?

BACKGROUND

Particles and fibres affect human health as a function of their properties such as chemical composition, size and shape but also depending on complex interactions in an organism that occur at various levels between particle uptake and target organ responses. While particulate pollution is one of the leading contributors to the global burden of disease, particles are also increasingly used for medical purposes. Over the past decades we have gained considerable experience in how particle properties and particle-bio interactions are linked to human health. This insight is useful for improved risk management in the case of unwanted health effects but also for developing novel medical therapies. The concepts that help us better understand particles' and fibres' risks include the fate of particles in the body; exposure, dosimetry and dose-metrics and the 5 Bs: bioavailability, biopersistence, bioprocessing, biomodification and bioclearance of (nano)particles. This includes the role of the biomolecule corona, immunity and systemic responses, non-specific effects in the lungs and other body parts, particle effects and the developing body, and the link from the natural environment to human health. The importance of these different concepts for the human health risk depends not only on the properties of the particles and fibres, but is also strongly influenced by production, use and disposal scenarios.

CONCLUSIONS

Lessons learned from the past can prove helpful for the future of the field, notably for understanding novel particles and fibres and for defining appropriate risk management and governance approaches.

Risk of gastrointestinal cancers from inhalation and ingestion of asbestos

This paper summarizes the weight of epidemiological evidence to evaluate the hypothesis that asbestos exposure is causally associated with increased risk of gastrointestinal (GI) cancers as suggested by Selikoff in an early study of insulation workers. This review looks at populations that develop GI cancers, namely stomach, colorectal, colon and rectal. Guidelines for assessing causality are strength of association, biological gradient and consistency of the associations. Exposure-response (E-R) was evaluated using three methods to estimate exposure. Rate Ratios (RRs) for lung cancer and percent of mesothelioma are used as surrogate measures of asbestos exposure for all the cohorts of exposed workers. Quantitative or semi-quantitative estimates of cumulative exposure to asbestos were also used to assess E-R trends and were compared to E-R trends for lung cancer and mesothelioma in individual studies. Surrogate measures are important since there are few individual studies that have assessed E-R. None of the various methods to estimate asbestos exposure yielded consistent E-R trends and the strength of the associations were consistently weak or non-existent for the four types of GI cancers. The epidemiological evidence detracts from the hypothesis that occupational asbestos exposure increases the risk of stomach, colorectal, colon, and rectal cancer. Findings are briefly summarized below.

Does asbestos exposure cause gastrointestinal cancer?

The relationship between asbestos exposure and gastrointestinal malignancies is unlike the well-established correlation between occupational asbestos exposure and the subsequent development of pleuropulmonary neoplasms and mesotheliomas. Cohort studies on occupationally exposed workers suggest an association between asbestos and gastrointestinal cancer, but evaluation of dose-response, tissue analysis, animal experiment, and cell culture data yield inconsistent conclusions. No simplistic cause-effect relationship can be ascribed to asbestos at the present time, and the answer to the question, "Does asbestos exposure cause cancer?" must await the results of additional studies.

Bronchogenic carcinoma in Olmsted County, 1935-1979

The incidence rates of bronchogenic carcinoma based upon 414 cases that occurred in Olmsted County, Minnesota, over a period of 45 years are described. Histologic preparations were available for 97% of cases and were reviewed by one pathologist. The overall average annual age-adjusted rates per 100,000 person-years for the most recent 5 years were 45.6 and 14.0 for men and women, respectively. Squamous cell carcinoma was most common among men and adenocarcinoma was the most common cell type among women. Incidence rates for women have risen to approximately those of men 30 years ago, whereas the incidence for men in the last 5 years has ceased to increase. Age-specific rates show increases for women in all age groups 55 years and older, and for men 65 years and older, whereas the incidence in men younger than 65 years has declined. Survival for 5 years is best for women for all cell types except small cell carcinoma, and is essentially unchanged from the earlier 40-year report.

The use of asbestos-cement pipe for public water supply and the incidence of cancer in selected communities in Utah

We tested the hypothesis that there is an association between the use of asbestos-cement piping for drinking water supplies and the incidence of gastrointestinal and kidney cancer. Cancer incidence in 14 Utah communities that had used predominantly asbestos-cement piping for transporting their drinking water supplies for 20 years or more were compared to 27 Utah communities that had never used asbestos-cement piping. Cancer incidence was tabulated for 11 cancer sites for the years 1967-1976. Increased Standard Incidence Ratios (SIRs) were found for cancer of the kidney in men (SIR 192) and leukemia (a control site) in women (SIR 203). No increased SIRs were found for the opposite sex at these sites or for the other gastrointestinal sites singly or in combination. There was no increase in age-adjusted cancer incidence for the 11 sites in 4 of the 14 study communities that had used asbestos-cement piping for 30 years or more. Limitations of the study were that the water supplies were nonaggressive, and leaching from the pipes was minimal if at all. Furthermore, the latent period for observation was very short, suggesting that these results should be considered preliminary.