| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Medicine, The University of Texas Health Center at Tyler, Tyler, Texas
Correspondence and requests for reprints should be addressed to David Coultas, M.D., The University of Texas Health Center at Tyler, 11937 U.S. Highway 271, Tyler, TX 75708. E-mail: david.coultas{at}uthct.edu
ABSTRACT
Several sources of evidence, including investigations of pathogenesis and observational studies, support the hypothesis that environmental agents may have an etiologic role in idiopathic pulmonary fibrosis (IPF). Since 1990, six case-control studies have been conducted in three countries and have consistently demonstrated increased risk of IPF with exposures to a number of environmental and occupational agents. In a meta-analysis of these studies, six exposures were significantly associated with IPF (summary odds ratios [95% confidence intervals]), including ever smoking (1.58 [1.27–1.97]), agriculture/farming (1.65 [1.20–2.26]), livestock (2.17 [1.28–3.68]), wood dust (1.94 [1.34–2.81]), metal dust (2.44 [1.74–3.40]), and stone/sand (1.97 [1.09–3.55]). Although there are a number of limitations of the case-control design and these results alone do not establish a causal link, an assessment of all of the available evidence strongly suggests that IPF may be a heterogeneous disorder caused by a number of environmental and occupational exposures.
Key Words: interstitial lung diseases • occupation • risk • smoking
The idiopathic interstitial pneumonias are a heterogeneous group of clinicopathologic disorders, with idiopathic pulmonary fibrosis (IPF) being the most common distinct entity (1). Although the name suggests that there are no known causes and the diagnosis of IPF requires "exclusion of other known causes of interstitial lung diseases such as drug toxicities, environmental exposures, and connective tissue diseases" (2), numerous sources of evidence suggest that environmental agents may have an etiologic role in IPF. Moreover, although a wide range of agents may be considered environmental, including infections and drugs with limited evidence linking them to IPF, the focus of this review is on the role of cigarette smoking and exposure to other environmental and occupational agents.
FACTORS LIMITING RECOGNITION OF ENVIRONMENTAL AGENTS CAUSING IPF
In experimental or occupational settings, exposure to fibrogenic dusts, fibers, and fumes have long been recognized as causing fibrotic lung diseases (3). However, in the clinical setting or in the conduct of observational studies, a number of factors may limit recognition of an association between a patient with IPF and his or her environmental and occupational exposures. These factors may include diagnostic misclassification, infrequent occurrence of IPF, exposure misclassification, and variation in susceptibility to exposures. An awareness and understanding of these factors are necessary for evaluating the evidence on the etiologic role of environmental and occupational agents in IPF.
Diagnostic Misclassification
Although a surgical lung biopsy is the gold standard for diagnosis of IPF, few patients ever have a lung biopsy (4, 5). Moreover, an accurate diagnosis requires expertise in the assessment of the clinical and radiologic information, and interpretation of the lung pathology (6). Given the low utilization of the gold standard for diagnosing IPF and the complexity of diagnosis, there is high likelihood for misdiagnosis.
Until the early 1990s, with the advent of high-resolution computed tomography (HRCT), the potential for diagnostic misclassification was probably greater than today. Moreover, the etiologic studies of IPF discussed in this review were conducted before widespread availability of HRCT and the recent American Thoracic Society (ATS)/European Respiratory Society (ERS) classification system of the idiopathic interstitial pneumonias (1).
The ATS/ERS consensus classification system defines six distinct clinicopathologic entities, including IPF/usual interstitial pneumonia (UIP), nonspecific interstitial pneumonia, acute interstitial pneumonia, respiratory bronchiolitis–associated interstitial lung disease, desquamative interstitial pneumonia, and lymphocytic interstitial pneumonia (1). Before publication of this classification system, many of these pathologic entities were likely included in the clinical diagnosis of IPF. The consequence of diagnostic misclassification is underestimation of the magnitude of risk (7).
Occurrence of IPF
In various populations, the prevalence estimates for IPF have ranged from 6 per 100,000 to as high as 32 per 100,000 (8). Thus, the number of patients available for conducting etiologic studies is small, which limits statistical power to detect associations between environmental exposures and IPF. Furthermore, although the low prevalence makes the case-control design the most feasible approach for investigating risk factors, the validity of this study design may be compromised by a number of biases, such as recall bias, compared with the less feasible cohort design.
Exposure Misclassification
One of the greatest challenges in the clinical setting or when conducting etiologic studies is accurate measurement of past exposures, including dose and duration. In either setting, accurate assessment of exposures to environmental and occupational agents may be compromised not only because of faulty patient recall but also because of failure of the clinician to systematically inquire about past exposures (3).
Patient recall of past exposures is of particular concern for case-control studies of IPF. In case-control studies, exposure measurement is based on patient and control subject recall of past environmental and occupational exposures. The validity of recall about past exposures may vary between patients with IPF and unaffected control subjects; this is termed "recall bias." For example, patients may recall more about past exposures as they search for causal explanations for their disease compared with unaffected control subjects. This differential recall may spuriously increase the magnitude of an association (9). Moreover, the long latency between exposures and onset of disease, combined with the possibility of exposure to multiple potential causative agents, makes accurate assessment of exposure particularly challenging.
In addition to the basic assessment of exposure to multiple potential agents, obtaining an accurate estimate of dose and duration introduces an even greater level of complexity and potential for measurement error. This limitation for assessment of dose–response relationships of risk and judging causal inferences is discussed in greater detail below.
Variation in Susceptibility
Variation in individual susceptibility is well established for smoking-related nonmalignant and malignant respiratory diseases and for pneumoconiosis, and may be determined by a number of factors, including exposure-related (e.g., dose and duration) and genetic factors (10, 11). Thus, it is highly plausible for similar variation in susceptibility to environmental exposures among patients with IPF. This variation in susceptibility to multiple potential fibrogenic agents may make detection of associations in observational studies particularly challenging.
SOURCES OF EVIDENCE
Several sources of evidence, including studies of pathogenesis and observational studies, suggest that environmental agents may cause IPF. The biological plausibility for a link between various exposures and IPF is based on the following factors: (1) the hypothesized process for the development of IPF, which begins with alveolar epithelial injury followed by abnormal repair mechanisms (12–14), and (2) results from animal experiments, which demonstrate that a wide variety of mineral particles are taken up by and may injure epithelial cells (15). Moreover, genetic variations regulating the processes of epithelial particle uptake, epithelial cell injury, and the inflammatory and fibrotic responses (14) may alter susceptibility to a wide variety of agents that cause pulmonary fibrosis. The role of genetics and associated pathophysiologic processes in the development of pulmonary fibrosis is considered extensively in this issue and elsewhere (11, 16, 17). A brief review of evidence from studies of the pathogenesis of pulmonary fibrosis follows.
Studies of Pathogenesis
The pathogenesis of pulmonary fibrosis has been examined using several different experimental systems, including inhalation, radiation, and drug-induced (e.g., bleomycin) models, and in studies of biological markers in humans. For this review, results from models of inhalation-induced pulmonary fibrosis and biological markers in humans provide the most direct evidence on the biological plausibility of environmental agents in the pathogenesis of IPF. In addition, although UIP is the pathologic hallmark of IPF, we also consider in this section selected evidence derived from other clinicopathologic entities, which provide indirect evidence supporting the biological plausibility.
Conceptually, four overlapping biological mechanisms have been proposed for the development of pulmonary fibrosis (10) (Figure 1). These mechanisms include the following: (1) delivery and persistence of agent, (2) biochemical response (e.g., oxidant injury), (3) immunologic response, and (4) fibrotic response (10). Although particle dose and physical and chemical characteristics partly determine susceptibility for the development of pulmonary fibrosis, genetic variations determining the complex biological responses to these inhaled particles may be largely responsible for development of pulmonary fibrosis (11). These alleles may increase or decrease risk for disease, or modify the severity or progression of fibrosis.
|
Modeling studies have demonstrated that the pattern of particle or fiber penetration and deposition is partly determined by anatomic and physiologic characteristics. In various animal models, inhaled particle and fiber deposition is substantially increased in the presence of abnormal airways and/or lung parenchyma (19–21). Moreover, increases in minute ventilation (e.g., exercise) will increase the dose of agent delivered to the lung. The higher prevalence of asbestosis in smokers has been partly attributed to poorer fiber clearance and increased fiber retention (21).
An imbalance in levels of oxidative stress induced by inhaled agents (e.g., cigarette smoke, silica) and antioxidant protective mechanisms has a role in the development of pulmonary fibrosis (22). In cell culture, fibroblast proliferation is stimulated by hydrogen peroxide. A link between environmental agents causing oxidative stress in the lungs is provided by findings of abnormal levels of antioxidant glutathione-dependent enzymes from patients with coalworkers' pneumoconiosis and hypersensitivity pneumonitis. Low glutathione levels have been found in bronchoalveolar lavage fluid among patients with IPF (22). Together, these findings provide evidence for environmental agents inducing oxidative stress that may lead to development of pulmonary fibrosis.
Variations in the pattern of immunologic responses to a variety of agents have a major influence on the fibroblast and the fibrotic response to injury (23). Beryllium-induced ILD is one of the most extensively investigated environmental lung diseases, and available evidence suggests a major role of host susceptibility in determining development of disease. The genetic marker HLA-DPB1 glutamate 69 is associated with an increased risk for development of berylliosis (24), and the risk of disease is increased 8 to 10 times among heavily exposed workers with this allele compared with similarly exposed workers without this marker (25).
The cytokine response to environmental exposures has been studied in animal models and provides further evidence for the biological plausibility of environmental agents in the development of pulmonary fibrosis (26). In a rat model of titanium dioxide exposure, a fibroproliferative response was associated with increased expression of the cytokine osteopontin with chemoattractant and cell adhesive properties.
In summary, the available biological evidence suggests that the triggering event in IPF may be a series of multiple microscopic and continuous insults to the alveolar epithelial cells from a variety of inhaled environmental agents (Figure 1). To date, a number of cytokine genes, surfactant protein genes, and genes determining fibroblast-related pathways and the coagulation cascade have been studied in sporadic cases of IPF, but no single fibrosis susceptibility gene has been identified (14, 16). Recent evidence, however, suggests that IPF is characterized by expression of tissue remodeling and epithelial and myofibroblast genes (14).
Observational Studies
Until 1990, observational evidence linking IPF and environmental exposures, including cigarette smoking and occupational agents, came only from selected reports of single patients or series of patients with ILDs clinically similar to IPF. In a number of early case series of patients with IPF, the prevalence of cigarette smoking was high, suggesting that smoking may be a risk factor for the disease (27).
To date, smoking has consistently been associated with IPF in a number of case-control studies (Table 1) and in a recent study of familial pulmonary fibrosis (28). In a family-based case-control study of familial interstitial pneumonia, Steele and coworkers (28) identified 111 families, with 309 affected and 360 unaffected individuals. Adjusting for age and sex, ever smoking was strongly associated with familial interstitial pneumonia (odds ratio [OR], 3.6; 95% confidence interval [CI], 1.3–9.8). Although the case-control design has many potential limitations, including selection bias, recall bias, reduced response rate from smoking controls, and misclassification of smoking status, the consistent association between smoking and IPF suggests a potential etiologic role in IPF (Table 1).
|
A number of case reports link IPF to various occupations that involve dust or fume exposure, including diamond polishing (30, 31), industrial car cleaning (32), dairy work (33), welding (34, 35), gold extraction (36), and technical dental work (37). There is also a case report of a patient developing IPF after exposure to a malfunctioning domestic wood burner (38). However, without control groups the hypothesis that environmental agents are associated with IPF cannot be formally tested in these case reports and case series.
Lung mineralogic analysis of patients with IPF has revealed that these patients have an excess of silica (39, 40) and metals, including iron and nickel (41). In a group of 93 patients with desquamative interstitial pneumonitis, Abraham and Herzberg (42) found a specific particulate in 92% of patients using scanning electron microscopy and energy-dispersive X-ray analysis. Taken together, these case reports and mineralogic data suggest that occupational dust may be a risk factor for IPF.
Since 1990, a number of observational studies have been conducted testing the hypothesis that environmental and occupational exposures are associated with IPF, including six case-control studies (Table 1), an analysis of autopsy results (43), and a historical cohort study (44). Patients in these studies were excluded if they had been exposed to known fibrogenic agents or had a connective tissue disease (2).
The first case-control study was conducted by Scott and coworkers (45) in the United Kingdom. Lifetime occupational data were obtained from a mailed questionnaire from 40 patients with IPF and 106 age-, sex-, and community-matched control subjects. There was a small, nonsignificant association between exposure to any occupational dust (OR, 1.32; 95% CI, 0.84–2.04), a but substantially higher risk associated with exposure to livestock, metal dust, and wood dust (Table 1).
In a follow-up to the study by Scott and coworkers (45), Hubbard and coworkers (46) conducted a larger case-control study with 218 patients and 569 community-matched control subjects from the Trent region of the United Kingdom. Data on occupational dust exposure were collected using a mailed questionnaire and a telephone interview. The results confirmed a significant, independent, and dose–response increase in the risk of IPF associated with both metal (OR, 1.68; 95% CI, 1.07–2.65) and wood dust exposure (OR, 1.71; 95% CI, 1.01–2.92). Significant associations were also found with exposure to stone/sand and textile dust (Table 1).
Iwai and coworkers (43) used an alternative design in Japan by comparing occupational data included in the autopsy records of 1,311 patients with IPF with that in a systematic sample of 393,000 control subjects. Patients with IPF were significantly more likely to have worked as metal workers, woodworkers, painters, and barbers/beauticians (all p < 0.01). This investigation also included a case-control study of 86 patients matched to two control groups: healthy community control subjects and patients with other respiratory conditions. A small increase in the risk of IPF was found in association with jobs working with metals (OR, 1.34; 95% CI, 1.14–1.59). In addition, they found a significant association with farming and living in an agricultural area (OR, 3.01; 95% CI, 1.29–7.43).
A multicenter case-control study was conducted in the United States by Baumgartner and coworkers (47, 48) that included 248 patients with IPF and 491 age-, sex-, and geography-matched control subjects recruited using random-digit dialing (Table 1). As in the previous studies, metal dust was significantly associated with IPF (OR, 2.0; 95% CI, 1.0–4.0). Moreover, risk increased with duration of exposure to metal dust, suggesting a dose–response relationship. Other significant associations included exposures to stone/sand dust, livestock, farming/agricultural areas, hairdressing, and raising birds.
Because recall bias is an inherent potential limitation of the case-control design, alternative designs are needed to strengthen confidence in these results. Hubbard and coworkers (44) used a cohort design to examine the risk of metal dust exposure. They used pension fund archives from a major metal engineering company (Rolls-Royce PLC), which included a total of 20,526 deaths and 55 deaths from IPF. Using national mortality data, 39.5 deaths from IPF were expected, resulting in a proportional mortality ratio of 1.39 (95% CI, 1.07–1.82). In addition, lifetime occupational records were obtained from 22 patients with IPF, of whom 13 had worked with metal, and the overall age- and sex-adjusted OR associated with metal work was nonsignificant (OR, 1.08; 95% CI, 0.44–2.65). However, there was evidence of a dose–response increase in risk (OR per 10 yr of exposure, 1.71; 95% CI, 1.09–2.68), and there was no association between duration of employment and IPF for nonmetal workers.
Meta-Analysis
Results from the available observational studies demonstrate that the risk of IPF is consistently increased for a number of environmental and occupational exposures (Table 1). We calculated summary estimates of these risks using the Mantel-Haenszel method when data were available from two or more case-control studies (Table 2). From this meta-analysis, statistically significant increased risks for IPF were associated with cigarette smoking and exposures to agriculture and farming, livestock, wood and metal dust, and stone and silica (Table 2).
|
INTERPRETATION OF EVIDENCE
The associations observed from the available studies of pathogenesis and observational studies provide support for the hypothesis that IPF may be a heterogeneous disorder caused by a number of environmental and occupational exposures. Although causation of IPF may never be directly observable, it is possible to infer causal relationships from the available scientific evidence using criteria for evaluating causality described by Hill (49). These criteria include experimentation, plausibility, coherence, analogy, consistency of association, strength of the association, biological gradient, temporality, and specificity. Of the nine original criteria, specificity of association is no longer relevant because multiple causation of disease is a well-established concept today. Although these criteria have been criticized, they provide a useful framework for evaluating the evidence.
Of the eight relevant criteria, five provide strong support for a causal relationship between environmental exposures and IPF, including experimentation, plausibility, coherence, analogy, and consistency of association. Results from studies of pathogenesis previously described in this review provide evidence of the biological plausibility of environmental agents causing pulmonary fibrosis. The inhalation of environmental agents causing pulmonary fibrosis is coherent with our present knowledge of the pathogenesis of IPF based on results from experimental studies and clinical observations (12, 13). Asbestos exposure, a known cause of pulmonary fibrosis, suggests by analogy that other environmental exposures may cause IPF. Consistent associations between environmental and occupational exposures and IPF have been found in observational studies from three countries (Tables 1 and 2).
Of the remaining three criteria for judging causality, there is intermediate support for strength of the association and biological gradient, and limited evidence for temporality. Although the magnitude of risk has been relatively low for the environmental exposures that have been examined (Tables 1 and 2), the relevance of this criterion for making causal inferences is limited because the factors that determine the strength of association are the relative prevalence of component causes of the disease and bias rather than biological mechanisms (7). Furthermore, the population-attributable risk percentages for agricultural and farming exposures and cigarette smoking suggest that a substantial proportion of cases could be prevented (Table 2). A biological gradient or dose–response relationship between exposure and risk of disease has been found for cigarette smoking (47), and wood or metal dust exposures (44, 46). Finally, only one cohort study has been conducted where exposure to metal dust preceded the onset of IPF (44). The criterion of temporality, with exposure preceding onset of disease, cannot be established from the case-control design where exposure and disease are measured simultaneously.
CONCLUSIONS/FUTURE DIRECTIONS
Together, the available evidence strongly suggests that the term "IPF" may be a misnomer and multiple environmental agents may be causing pulmonary fibrosis in susceptible individuals. However, further evidence is needed to strengthen the causal link in patients with IPF. Although prospective cohort studies are needed to establish that exposures precede the onset of pulmonary fibrosis and to more precisely define dose–response relationships, the infrequent occurrence of IPF is a major obstacle for conducting this type of study. Despite multiple potential challenges for conducting etiologic studies in patients with IPF, the high morbidity and mortality without an effective therapy provide a strong rationale to pursue further evidence on environmental exposures in IPF that may be used to prevent this devastating disorder.
ACKNOWLEDGMENTS
The authors thank Dr. Jeffery Levin for his helpful review and comments.
FOOTNOTES
Conflict of Interest Statement: Neither of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript.
(Received in original form December 22, 2005; accepted in final form February 2, 2006)
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  A. K. Attili, E. A. Kazerooni, B. H. Gross, K. R. Flaherty, J. L. Myers, and F. J. Martinez Smoking-related Interstitial Lung Disease: Radiologic-Clinical-Pathologic Correlation RadioGraphics, September 1, 2008; 28(5): 1383 - 1396. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. M. Antoniou, D. M. Hansell, M. B. Rubens, K. Marten, S. R. Desai, N. M. Siafakas, A. G. Nicholson, R. M. du Bois, and A. U. Wells Idiopathic Pulmonary Fibrosis: Outcome in Relation to Smoking Status Am. J. Respir. Crit. Care Med., January 15, 2008; 177(2): 190 - 194. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Pottier, C. Chupin, V. Defamie, B. Cardinaud, R. Sutherland, G. Rios, F. Gauthier, P. J. Wolters, Y. Berthiaume, P. Barbry, et al. Relationships between Early Inflammatory Response to Bleomycin and Sensitivity to Lung Fibrosis: A Role for Dipeptidyl-Peptidase I and Tissue Inhibitor of Metalloproteinase-3? Am. J. Respir. Crit. Care Med., December 1, 2007; 176(11): 1098 - 1107. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Fireman, Y Lerman, M Ben Mahor, E Ganor, and M. Kramer Redefining idiopathic interstitial lung disease into occupational lung diseases by analysis of chemical composition of inhaled dust particles in induced sputum and/or lung biopsy specimens Toxicology and Industrial Health, November 1, 2007; 23(10): 607 - 615. [Abstract] [PDF]
|
|
|
|
|
|
A. L. Olson, J. J. Swigris, D. C. Lezotte, J. M. Norris, C. G. Wilson, and K. K. Brown Mortality from Pulmonary Fibrosis Increased in the United States from 1992 to 2003 Am. J. Respir. Crit. Care Med., August 1, 2007; 176(3): 277 - 284. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Arakawa, T. Johkoh, K. Honma, Y. Saito, Y. Fukushima, H. Shida, and N. Suganuma Chronic Interstitial Pneumonia in Silicosis and Mix-Dust Pneumoconiosis: Its Prevalence and Comparison of CT Findings With Idiopathic Pulmonary Fibrosis Chest, June 1, 2007; 131(6): 1870 - 1876. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Mueller-Mang, C. Grosse, K. Schmid, L. Stiebellehner, and A. A. Bankier What Every Radiologist Should Know about Idiopathic Interstitial Pneumonias RadioGraphics, May 1, 2007; 27(3): 595 - 615. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
