This Article Abstract Full Text (PDF) All Versions of this Article: 200311-1575SOv1 172/3/280    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Mapp, C. E. Articles by Fabbri, L. M. Search for Related Content PubMed PubMed Citation Articles by Mapp, C. E. Articles by Fabbri, L. M.

Occupational Asthma

Section of Hygiene and Occupational Medicine, Department of Clinical and Experimental Medicine, University of Ferrara; Department of Environmental Medicine and Public Health, University of Padua; and Section of Respiratory Diseases, Department of Oncology and Hematology, University of Modena, Italy

Correspondence and requests for reprints should be addressed to Cristina E. Mapp, M.D., Section of Hygiene and Occupational Medicine, Department of Clinical and Experimental Medicine, University of Ferrara, Via Fossato di Mortara 64/b, 44100 Ferrara, Italy. E-mail: mapp{at}unipd.it

	ABSTRACT

TOP ABSTRACT CONTENTS OCCUPATIONAL ASTHMA: DEFINITION... EPIDEMIOLOGY AND RISK FACTORS PATHOPHYSIOLOGY NATURAL HISTORY AND LONG-TERM... CONCLUSIONS AND FUTURE RESEARCH REFERENCES

 
Substantial epidemiologic and clinical evidence indicates that agents inhaled at work can induce asthma. In industrialized countries, occupational factors have been implicated in 9 to 15% of all cases of adult asthma. Work-related asthma includes (1) immunologic occupational asthma (OA), characterized by a latency period before the onset of symptoms; (2) nonimmunologic OA, which occurs after single or multiple exposures to high concentrations of irritant materials; (3) work-aggravated asthma, which is preexisting or concurrent asthma exacerbated by workplace exposures; and (4) variant syndromes. Assessment of the work environment has improved, making it possible to measure concentrations of several high- and low-molecular-weight agents in the workplace. The identification of host factors, polymorphisms, and candidate genes associated with OA is in progress and may improve our understanding of mechanisms involved in OA. A reliable diagnosis of OA should be confirmed by objective testing early after its onset. Removal of the worker from exposure to the causal agent and treatment with inhaled glucocorticoids lead to a better outcome. Finally, strategies for preventing OA should be implemented and their cost-effectiveness examined.

Key Words: asthma • management • risk • susceptibility • workplace

	CONTENTS

TOP ABSTRACT CONTENTS OCCUPATIONAL ASTHMA: DEFINITION... EPIDEMIOLOGY AND RISK FACTORS PATHOPHYSIOLOGY NATURAL HISTORY AND LONG-TERM... CONCLUSIONS AND FUTURE RESEARCH REFERENCES

 
Occupational Asthma: Definition and Causal Agents

What Is Occupational Asthma?

Variants

Causal Agents

Epidemiology and Risk Factors

Prevalence and Incidence of OA

Role of Exposure and Risk Factors

Genetics

Pathophysiology

Immunologic OA: IgE Dependent and IgE Independent

Irritant-induced Asthma

Controversial Issues

Natural History and Long-term Consequences

Diagnosis

Management

Prevention and Surveillance

Conclusions and Future Research

	OCCUPATIONAL ASTHMA: DEFINITION AND CAUSAL AGENTS

TOP ABSTRACT CONTENTS OCCUPATIONAL ASTHMA: DEFINITION... EPIDEMIOLOGY AND RISK FACTORS PATHOPHYSIOLOGY NATURAL HISTORY AND LONG-TERM... CONCLUSIONS AND FUTURE RESEARCH REFERENCES

 
Occupational asthma (OA) has become one of the most common forms of occupational lung disease in many industrialized countries (1, 2), having been implicated in 9 to 15% of adult asthma (3, 4). OA is important to recognize clinically, because it has serious medical and socioeconomic consequences (5, 6). Like asthma that develops in childhood, OA is probably the result of multiple genetic, environmental, and behavioral influences (7). Because in most cases OA can be accurately identified and exposures to the causal agent can be measured, and because OA does not differ in its clinical and pathologic features from nonoccupational asthma, it is likely that studies of OA have the potential to provide us with useful information about the effects of genetic, environmental, and behavioral interactions in adult-onset asthma (8).

A better understanding of the natural history of adult asthma will allow us to assess the effects of early diagnosis, environmental control, and therapy on the outcome of this disease. Of the criteria necessary to determine the natural history of any disease (9), the first, a precise definition of the disease, is met if the definition of OA is limited to those conditions in which asthma is caused by occupation (10); this definition allows subjects with the disease to be identified. The second criterion is the availability of longitudinal studies of the disease. OA can be studied throughout its course from onset to remission or persistence (11). An advantage of these studies is that if OA is diagnosed early, it can be cured in substantially higher proportion than adult-onset nonoccupational asthma. In addition, the ability to perform prospective studies of workers before exposure for the presence or absence of risk factors and to monitor the workers after onset of exposure for development of disease may help to determine the relevance of various predictors and their interaction with exposure. The third criterion is the evaluation of the effect of therapy on the course and outcome of the disease. In this regard, OA is unique, because it is possible to investigate not only the effects of treatment, but also the effects of ending exposure to the causal agent (12, 13).

What Is Occupational Asthma?
For a disease that was first described many centuries ago, attempts to define OA have met with surprising difficulty until more recently. Although the broader category of work-related asthma encompasses both OA and asthma aggravated by work or the work environment, what emerged in all the proposed definitions was the causal relationship between workplace exposure and asthma and/or the specificity of the causal agent to the workplace. In guidelines published by the American Thoracic Society, OA is defined as asthma caused by work exposure (14). Similarly, authors of a critical review of the definitions and types of OA (15) stated that the definition of OA should be limited to those conditions in which asthma is caused by occupation. Briefly, OA is a disease characterized by variable airflow limitation and/or airway hyperresponsiveness due to causes and conditions attributable to a particular occupational environment and not to stimuli encountered outside the workplace (10, 16). Two types of OA are distinguished by whether they appear after a latency period. (1) Immunologic: OA appears after a latency period of exposure necessary for the worker to acquire immunologically mediated sensitization to the causal agent. This type encompasses OA that is induced by an IgE mechanism (most high- and some low-molecular-weight agents), and OA in which an IgE mechanism has not been demonstrated consistently (low-molecular-weight agents, such as diisocyanates, western red cedar, and acrylates). (2) Nonimmunologic: OA is characterized by the absence of a latency period. It occurs after accidental exposure to high concentrations of a workplace irritant. This clinical entity has been defined as irritant-induced asthma (17, 18). The most definitive form of irritant-induced asthma is "reactive airway dysfunction syndrome" (RADS) occurring after a single exposure to high levels of an irritating vapor, fume, or smoke (19). In addition, work-related asthma encompasses variant syndromes, including eosinophilic bronchitis and asthma-like disorders caused by exposure to organic dusts.

There is general agreement that the key element in defining OA is evidence of a direct causal relationship between workplace exposure and the development of asthma; nevertheless, several issues and controversies persist.

One controversy involves the definition of OA. The definitions of OA have varied according to the focus of the authors, such as epidemiologic studies (20), workplace surveillance programs (16, 21, 22), clinical diagnosis (23), and medicolegal assessment (24). The clinical diagnosis of OA requires the highest level of evidence, because it has significant implications for both worker health and socioeconomic status.

Few authors have suggested the inclusion of work-aggravated asthma in the definition of OA (14, 25, 26), because this entity is common (27, 28) and may cause disability and socioeconomic consequences that are potentially preventable (29, 30). Work-aggravated asthma is defined as preexisting or concurrent asthma that is exacerbated by workplace exposure (15). Exertion and exposure to cold, dry air, dust, fumes, and sprays are common in the workplace and may aggravate asthma (27, 28), especially in those with moderate to severe disease and in those not receiving optimal treatment. When work-aggravated asthma occurs on a regular basis rather than from a single incident, it can be assessed by measuring peak expiratory flow rates (PEFRs) and by symptom/medication diaries. Work-aggravated asthma should be distinguished from OA, because the outcome, medical management, and preventive measures differ substantially. Reducing workplace exposure to respiratory irritants; limiting exposure to relevant environmental allergens and nonoccupational irritants such as tobacco smoke; and optimizing antiasthma therapy, educating the patient about how to use the drugs, and emphasizing the importance of compliance often allow workers with this type of asthma to continue working in the same job. These options could probably be managed at lower costs than the cost of completely removing a subject with OA from the workplace. However, these measures are not sufficient to prevent the relapse of true OA. Some individuals with work-aggravated asthma, as a consequence of new exposures to specific agents in the workplace, may develop true OA. Thus, OA can occur in workers with or without prior asthma. A label for this situation has often provoked controversy, because the individual has OA as well as nonoccupational asthma.

In the nonoccupational setting, the hypothesis that allergens are a direct cause of asthma is largely based on indirect evidence (31, 32). However, in the occupational setting, some studies, even if concerning few causal agents of OA, have shown that the environment can cause asthma de novo (33–35).

An area of uncertainty is the diagnosis of RADS, which requires that several criteria be met (16, 19), the most important being the strong temporal association between inhalation exposure and the rapid onset of asthmatic symptoms. It follows that a diagnosis of RADS should never be made in subjects with preexisting asthma. There is debate about how to define the worsening of preexisting asthma caused by inhalation of high levels of irritants (36). For some authors, strict application of criteria required to make a diagnosis of RADS causes difficulties in achieving an appropriate diagnosis and the recognition of an occupational accident when the patient has a history of cured allergic asthma or smoking-related chronic obstructive pulmonary disease (37). Another debate concerns whether RADS is a subset of irritant-induced asthma (38). We believe that RADS should be considered a subset of irritant-induced asthma, a broader term that characterizes workers who develop asthma after both single and multiple irritant exposures (39, 40). Vandenplas and Malo (15) suggested that, although widely used, the term "RADS" should be replaced in future nomenclature by "acute irritant-induced asthma" or "sudden-onset irritant-induced asthma" to avoid confusion with delayed or progressive forms of asthma associated with irritant exposures at work.

Another issue concerns whether asthma develops from multiple exposures, as suggested by two studies (18, 41), rather than from a single exposure to high levels of irritants. However, in the studies suggesting the requirement of multiple exposures, it is possible that the changes were progressive, because changes in asthma after each exposure were not measured. Data obtained in these studies confirmed that initiation of asthma required a single high-level exposure. Therefore, asthma resulting from multiple high-level exposures to irritants can be considered irritant-induced OA, provided that the onset of persistent asthmatic symptoms is temporally related to one documented, severe accidental inhalation exposure.

There is still controversy about whether intermittent high-level exposure and chronic low-level exposure to irritants can cause OA (39, 42). Both seem possible. It was shown that repeated peak exposure to irritant gases, for example, in the pulp industry, increased the risk for both adult-onset asthma and wheezing (43), and also that asthma symptoms developed in three patients after repetitive exposure to irritants that took place over a period of several days to months (44).

Prezant and coworkers (45) described a condition called "World Trade Center cough" in many firefighters exposed to inorganic dusts, products of pyrolysis, and other respirable materials at the site of the World Trade Center after the terrorist attacks on September 11, 2001. Airway hyperresponsiveness was present in about one-quarter of the firefighters who had high levels of exposure, whether or not they had World Trade Center cough. Other abnormalities included airway obstruction and a bronchodilator response, without evidence of parenchymal changes on chest radiography. These observations are consistent with irritant-induced asthma due to high levels of dust inhalation (46, 47) and suggest that, at variance with previous studies (48, 49), a severe inhalation injury requiring hospitalization is not needed for the development of irritant-induced asthma. A longitudinal study showed that airway hyperresponsiveness shortly after the collapse of the twin towers predicted reactive airway dysfunction at 6 months in highly exposed firefighters (50). It has been reported that the high alkalinity of the World Trade Center dust produced the airway hyperresponsiveness, persistent cough, and increased risk of asthma (51).

The main characteristics of the various forms of work-related asthma are shown in Table 1 (52).

This picture is further complicated by the description in clinical practice of other conditions that mimic asthma, for example, asthmalike disorders and eosinophilic bronchitis, and it is still debated whether these conditions should be considered OA.

Variants
Asthmalike disorders.
Asthmalike disorders have been especially related to exposure to vegetable dusts such as grain, cotton, and other textile fibers and to dust from animal confinement buildings. A population-based study performed on 1,906 subjects has shown that occupational exposure to organic dust is associated with an increased risk of asthma, particularly in atopic subjects (55). How can these disorders be differentiated from OA? One way is that, in asthmalike disorders, asthma symptoms may be associated with systemic symptoms, which are usually not present in OA. There is no latency and symptoms can occur in naive subjects on first exposure; these are not typical features of OA. Third, asthma symptoms in asthmalike disorders may be associated with transient increases in nonspecific airway responsiveness, increases that are usually important and persistent in OA. Finally, in asthmalike disorders, asthma symptoms may be associated with neutrophilic airway inflammation (56), which has been shown for some but not all causal agents of OA. Asthmalike disorders have been associated with exposure to endotoxin, and this observation makes their mechanism more intriguing (see PATHOPHYSIOLOGY).

Eosinophilic bronchitis.
Eosinophilic bronchitis has been described as a cause of chronic cough characterized by sputum eosinophilia in the absence of demonstrable variable airflow limitation or nonspecific airway hyperresponsiveness (57). Exposure to acrylates, natural rubber latex, mushroom spores, and an epoxy resin hardener in the workplace can result in increased sputum eosinophilia in the absence of airflow limitation and increased nonspecific airway responsiveness (58–60). Because increases in sputum eosinophils are remarkable and reproducible, the condition should be considered occupationally induced.

Although this condition clearly does not meet the definition of asthma, it is not known whether it can progress to typical OA (60). In children, it has been proposed that eosinophilic bronchitis may be a precursor of asthma (61), whereas in adults, one study suggests that eosinophilic bronchitis is unlikely to progress to typical OA, because it has been shown that eosinophilic bronchitis and nonoccupational asthma have different clinical, physiological, and pathologic features (62). However, Park and coworkers (63), by performing a prospective follow-up study on patients with eosinophilic bronchitis, found that repeated episodes of eosinophilic bronchitis are associated with the development of chronic airflow obstruction, including asthma.

Whether the immunopathologies of eosinophilic bronchitis and asthma are similar is uncertain. One study suggests that eosinophilic airway inflammation and increased basement membrane thickening are regulated independently of airway hyperresponsiveness because airway hyperresponsiveness is absent in eosinophilic bronchitis (64). In another study, results obtained in both sputum samples and bronchial biopsies have shown that both sputum concentration and expression of interleukin (IL)-13 were higher in patients with mild asthma than in patients with eosinophilic bronchitis (65).

Causal Agents
Agents that cause OA with latency encompass more than 300 natural and synthetic chemicals (Tables 2 and 3). They are listed in textbooks (17, 66), review articles (67–70), and websites (e.g., www.asmanet.com, www.asthme.csst.qc.ca) (71, 72). These agents are categorized into high-molecular-weight and low-molecular-weight agents. High-molecular-weight agents are proteins of animal or vegetable origin acting through an IgE-mediated mechanism. Low-molecular-weight agents include organic and inorganic compounds that, with a few exceptions, are not associated with an IgE mechanism (Table 4). Common identified agents of OA with latency are diisocyanates, flour, allergens from laboratory animals and insects, enzymes, colophony fluxes, solders, wood dusts, natural rubber latex, acrylates, and glutaraldehyde. After sensitization, workers with OA may develop an asthmatic attack on low exposures to the sensitizer. The extent of airway responsiveness may diminish away from exposure but usually increases with reexposure to the sensitizer.

A few comments are necessary after consulting the long list of causal agents of OA. First, several case reports of OA due to unusual allergens have been published, but from a practical point of view, many of them are allergens one would never see in one's lifetime. Second, in addition to emerging low-molecular-weight sensitizers, such as acrylates, there are still new causes of OA due to high-molecular-weight agents. Third, some workers, for example, spray painters, health professionals, bakers, food processors, farmers, hairdressers, and plywood mill workers (73), are exposed to more than one sensitizer, raising the possibility that sensitization may occur through the interaction of different agents. Furthermore, some jobs, such as that of radiographer, expose workers to both sensitizers and irritants (74). Fourth, information about the causal agents of OA should be made available to workers in the occupations at risk. Such information has been prepared by Henriette Dhivert-Donnadieu and coworkers (71) in Montpellier, France, and is available at the site www.asmanet.com. Finally, OA can be caused by agents that are not found on existing lists and databases.

	EPIDEMIOLOGY AND RISK FACTORS

TOP ABSTRACT CONTENTS OCCUPATIONAL ASTHMA: DEFINITION... EPIDEMIOLOGY AND RISK FACTORS PATHOPHYSIOLOGY NATURAL HISTORY AND LONG-TERM... CONCLUSIONS AND FUTURE RESEARCH REFERENCES

 
Prevalence and Incidence of OA
Epidemiology is the study of the distribution, determinants, and outcome of a disease (75). In the 1980s, some studies of industrialized countries reported an increase in the incidence of OA (76, 77). Blanc and Toren (3) reviewed most of the epidemiologic studies on OA published from 1996 to 1999, and gave a median overall estimate of the attributable risk (i.e., the fraction of cases in a population that arise because of occupational exposures) of OA of 9% (range, 5–25%). More recently, after a review of the published literature regarding the magnitude of the attributable risk of asthma due to occupation, a median value of 15% (range, 4–58%) among all asthma cases was proposed as a reasonable estimate by Balmes and coworkers (4). In the general population, the occupations that contribute to asthma are in particular industries such as construction, metal work, rubber work, plastic work, printing, and industrial cleaning (78, 79).

Epidemiologic study designs include the cross-sectional study, the population-based study, the randomized control trial, the prospective cohort, the retrospective cohort, the case-referent study, and the case series. These designs vary regarding advantages, limitations, and information obtained. A community-based study with special reference to single and multiple exposures showed that application of job exposure matrices can be a useful tool with which to estimate asthma risk attributable to specific occupational exposures (i.e., exposures to high-molecular-weight agents) in the general population (80). The authors underlined the importance of interpreting specific exposure risks in connection with all concomitant exposures present in the work environment.

Most epidemiologic studies of OA have been cross-sectional. This type of study suffers from survivor bias (i.e., it underestimates the prevalence because of the workers with OA who left work), whereas prospective cohort studies, even if more expensive, are less affected by selection or survivor biases.

Both prospective and retrospective studies (81) have been successful in determining the exposure–response relationship. Some case-referent studies have been undertaken in workers exposed to causal agents of OA, such as acid anhydride (82), rat allergens (32), and flour (83). These studies are useful in identifying and quantifying risk factors. In addition, as reviewed by Newman Taylor (84), they show a clear exposure–response relationship, enabling the development of strategies to prevent sensitization and disease. Evidence of this relationship has also been obtained in a case-referent study on workers exposed to isocyanates (85). Available information indicates that the exposure–response relationships are more evident in those workers who develop immunologic sensitization and work-related symptoms soon after the onset of exposure (33, 86).

Table 2 reports an update of the estimated prevalence of work-related asthma obtained from 15 cross-sectional studies (87–101). The estimated incidence of work-related asthma and estimates of the attributable risk for work-related asthma are shown in Table 3 (1, 25, 102–123). It should be emphasized that, in epidemiologic studies, the definition of OA varies according to the study population, the study design, and the aim of the study (20). Estimates of incidence have been made using national registers, medicolegal statistics, registers based on voluntary or mandatory physician reporting, and self-reporting by employees (1, 33, 116, 124–127). In addition, estimates are reported in a variety of units, and some of the data cannot be converted into percentages, making the studies difficult to compare.

Information on the population-attributable risk of work-related asthma and work-related wheezing has been obtained from population-based studies (77, 78, 119–123, 128). Data from the Third National Health and Nutrition Survey, 1988–1994, in the United States indicated a prevalence of work-related asthma of 3.70% (95% confidence interval [95% CI], 2.88–4.52) and a prevalence of work-related wheezing of 11.46% (95% CI, 9.87–13.05). The population-attributable risk for work-related asthma was 36.5% and for work-related wheezing was 28.5%, suggesting that OA should be a priority on the public health agenda (128). In a population-based incident case-control study (1997–2000) of 521 cases and 932 control subjects in south Finland, Jaakkola and coworkers (129) showed that, in men, the strongest risk factors for asthma were metal work (odds ratio [OR], 4.52; 95% CI, 2.35–8.70) and forestry work (OR, 6.00; 95% CI, 0.96–37.5), whereas for women, asthma risk was highest for waitresses (OR, 3.03; 95% CI, 1.10–8.31), cleaners (OR, 1.42; 95% CI, 0.81–2.48), and dental workers (OR, 4.74; 95% CI, 0.48–46.5).

Large discrepancies in the prevalence of work-related asthma have been reported. One reason might be that many prevalence studies have relied on symptom and job history questionnaires. When objective testing was included in the studies (20), prevalence rates of about 5% for high-molecular-weight agents and greater than 5% for low-molecular-weight agents were found. Other confounders included the definition of OA adopted in the study, the type of agent, the work practice, and the fact that the studies were performed in different countries.

Surveillance systems enable us to estimate the incidence of work-related asthma, to describe the characteristics of affected workers, and to implement and facilitate public health interventions. These programs have been undertaken in many countries including Europe, the United States, South Africa, Australia, and New Zealand (Table 3). Incidence estimates obtained with surveillance programs are usually lower than those obtained from population studies. Gautrin and coworkers (75) gave an explanation for this observed difference. Population-based studies are likely to include all individuals ever exposed in at-risk workplaces, rather than those currently exposed, with the consequences of a higher incidence of OA and a reduction in survivor effects.

To develop a health surveillance strategy, investigators in the United Kingdom used data from 351 laboratory animal workers participating in an ongoing cohort study to develop diagnostic and prognostic models that detect and predict occupational allergic diseases (130). The study showed that, in these workers, both the risk of future sensitization and the severity of allergy can be predicted accurately with diagnostic and prognostic prediction models based on questionnaire items. An interesting observation of this study was that workers with an increased risk of future sensitization also showed severe allergic symptoms at follow-up over a 2- to 3-year period.

Medicolegal data may be useful in estimating the incidence of work-related asthma, provided that the diagnosis is certain, even though the medicolegal handling of accepted cases of OA is unsatisfactory in many countries, developed and nondeveloped. Tarlo and coworkers (27) reported that only 5% of all cases of adult asthma (310 individuals) referred to a tertiary care hospital, and who were still employed at the time of doctor's visit, had probable OA.

Studies of the prevalence of irritant-induced asthma have been rare and often have relied on historical data or have modified certain diagnostic criteria, such as expanding the definition to include the onset of symptoms from 24 hours to several days after exposure. Although irritant-induced asthma is not as widespread as immunologic OA (131), the low prevalence (52) could be due to poor recognition of this condition as OA. Estimates of 15 and 11% of all work-related asthma cases were reported as irritant-induced asthma in two sentinel projects, the Surveillance of Work-related and Occupational Respiratory Disease (SWORD) in the United Kingdom (124) and the Sentinel Event Notification System for Occupational Risks in four states in the United States (21). Similarly, estimates of 15% of all work-related asthma cases in Ontario, Canada, were reported as irritant-induced asthma (39). In Finland, cases of irritant-induced asthma doubled between 1981 and 1991 (132). Cases also appeared to increase in the United States. Data obtained from surveillance systems of work-related asthma undertaken in some states of United States were used to identify cases of irritant-induced asthma associated with exposure to cleaning products at work (133). Among new-onset work-related asthma, 22% were consistent with irritant-induced asthma, and the most likely exposure was in medical settings, schools, and hotels. An increased risk of asthma from exposure to cleaning agents was confirmed in a study of the general population in France, where ORs increased from 1.55 (95% CI, 1.08–2.23) for "ever asthma" to 2.17 (95% CI, 1.41–3.34) for asthma onset after age 14 years, to 2.35 (95% CI, 1.38–4.00) for asthma onset after beginning current job, and to 2.51 (95% CI, 1.33–4.75) for asthma with airflow limitation (79).

There is insufficient evidence of new-onset work-related asthma due to chronic low-level exposures to irritants (75). Similarly, estimates of irritant-induced work-related asthma in developing countries are unavailable. Studies are needed to determine the airway effects of intermittent high-level and chronic low-level exposures, because a population-based study has shown a high relative risk of asthma in jobs with expected low to moderate exposure to irritants (121). Suitable industries for studying the effects of low chronic exposure to irritants include metal refining (vanadium), fertilizers (ammonia), and mining (oxides of nitrogen). A review of the effects of low-level exposure to respiratory irritants has been published (134).

A descriptive study examined not only the risk of work-related new-onset asthma but also the risk of work-aggravated asthma, by using cases reported to the National Institute for Occupational Safety and Health from four state Sentinel Event Notification System for Occupational Risks surveillance programs for 1993–1995 (135). Two hundred and ten work-aggravated asthma cases and 891 new-onset asthma cases were reported. The risk of work-aggravated asthma was highest in the public administration, and the risk of new-onset asthma was highest in both manufacturing and public administration. The authors concluded that work-aggravated asthma cases reported many of the same adverse consequences as new-onset asthma cases.

Role of Exposure and Risk Factors
In 1995, Chan-Yeung and Malo (136) stated that exposure is the most important determinant in the development of OA. In a review of studies on OA with latency, it was observed that the higher was the degree of exposure to an occupational agent, the higher was the prevalence of asthma (137). This concept was proposed again by Frew (13), who stated that, in general, the higher the level of exposure, the more likely the sensitized person is to develop asthma. Once a subject is sensitized, the main factor that influences the onset of symptoms is the degree of exposure (138, 139). However, there is still a lack of information regarding the risk of sensitization at low concentrations and the existence of a "no-effect level" (140). For isocyanate-induced asthma, one study has suggested that peak exposures could be more relevant than the cumulative dose of exposure (141); another stated that continuous exposure rather than intermittent acute exposure to high concentrations of isocyanates increased the risk of developing symptoms of asthma (92). Moreover, an experimental animal model of isocyanate-induced asthma has shown that sensitization can occur through subchronic inhalation of vapor-phase diisocyanate levels as low as 20 ppb (142). Other authors have suggested that in the development of sensitization to low levels of occupational sensitizers, genetic susceptibility markers seem to be important determinants (143), whereas the effect of atopy is independent of exposure level (33, 144).

Studies have shown that the intensity of exposure is an important determinant of sensitization and asthma caused by respiratory sensitizers. A review of exposure–response relationships for occupational inhaled allergens (145) suggests that there are enough data for assessment of exposure–response relationships for several occupational agents (146). Becklake and coworkers (20) reported a dose–response relationship for cedar, colophony, and flour. Other studies have shown a dose–response relationship between the levels of exposure to occupational sensitizers such as flour, fungal -amylase, laboratory animal proteins, detergent enzymes, platinum salts, and acid anhydrides and the development of IgE-mediated sensitization and/or work-related symptoms (81, 86, 147–153).

The finding of a concentration of an occupational agent below which sensitization is uncommon is relevant for prevention. An important concept to keep in mind is that the concentration of an allergen that sensitizes is quite different from one that provokes symptoms in workers already sensitized (154). Thus, the minimum concentration that induces sensitization is at least one order, and probably two orders, of magnitude greater than the minimum concentration that elicits symptoms. So the permissible exposure limit (155) for eliminating sensitization is easier for industry to achieve than the permissible exposure limit for eliminating asthmatic symptoms. Furthermore, it is much more cost-effective to reduce exposure to prevent sensitization, and reduction of exposure should be done as early as possible. An animal model of rats sensitized to trimellitic anhydride showed that trimellitic anhydride challenge of sensitized rats caused concentration-related allergic airway inflammation, changes in breathing pattern, and an increase in nonspecific airway responsiveness, and that the lowest no-observed-effect level based on the most sensitive end point investigated was 0.2 mg/m³, a value that is well below the concentration that causes irritation (156). The authors of this study concluded that the assessment of safe human exposure levels is feasible.

Even though aeroallergen exposure is more complex than exposure to toxic materials, new immunologic methodologies and the use of personal sampling have made possible the measurement of high-molecular-weight allergens in the workplace. Detailed information on this topic has been published (154, 157). In addition, data obtained by quantified environmental challenge with allergens may provide objective data for recommended permissible exposure limits and support of environmental assessment guidelines. For example, data obtained by performing quantified environmental challenge with powdered natural rubber latex gloves suggest that a concentration of 0.6 ng/m³ (natural rubber latex aeroallergen) as a time-weighted average threshold, which is critical for allergic symptom onset or clinical features of natural rubber latex sensitivity, may be too conservative (158).

A strategy similar to that adopted for high-molecular-weight agents should be considered for low-molecular-weight agents (159). An overview of assessment of exposure to low-molecular-weight agents by chemical assays has been published (160). Despite some progress, it is still not known whether peak or mean exposures to low-molecular-weight agents are more important in causing sensitization and OA.

What is known is that, even if the level of exposure is a critical factor for the development of OA (161), given the same level of exposure, only a small proportion of workers will develop sensitization and/or OA, suggesting that host susceptibility may be a factor. This view is supported by the observation that, among workers exposed to rat urinary allergens, the sensitization rate increased with increasing aeroallergen exposure, but there was a difference between atopic and nonatopic subjects. Atopic subjects had a clearly elevated sensitization risk at low exposure levels, but the risk increased little with increasing exposure. For nonatopic subjects, a steadily increasing risk was observed with increasing exposure (149).

Various risk factors for OA have been established. Atopy (skin reactivity to common inhalants) is a predisposing factor in workers exposed to high-molecular-weight agents, but it is a weak predictor of sensitization and development of OA (162). Atopy is not a risk factor for asthma induced by low-molecular-weight agents such as western red cedar (163) or diisocyanates (164). An important observation was made by Gautrin and coworkers (34), who found that skin reactivity to pets (relative risk, 4.11; 95% CI, 1.6–10.8) was a significant predictor of laboratory workers who develop OA after exposure to laboratory animals, whereas atopy was not (relative risk, 2.09; 95% CI, 0.8–5.6). The same investigators conducted a study of about 800 apprentices in the fields of animal health technology, pastry making, and dental hygiene, and found that 32% of the incident cases of sensitization and 27% of the incident cases of OA were among nonatopic subjects (34, 102). A study conducted on a large cohort of bakers in Belgium showed that atopy and sensitization to bakers' allergens were independent of each other (144).

Other factors besides atopy, such as rhinoconjunctivitis symptoms and having a measurable PC₂₀ (provocative concentration of histamine producing a 20% decrease in FEV₁), could be important in the development of respiratory symptoms and in the development and/or worsening of asthma. Of these potential factors, having a measurable PC₂₀ was the most significant (165).

An important question concerns whether sensitization to common aeroallergens precedes or follows sensitization to occupational allergens. For exposure to high-molecular-weight work-related allergens, subjects with new occupational sensitization are at greater risk of developing sensitization to common aeroallergens than are subjects without sensitization (166). In addition, new sensitization to common aeroallergens often occurs at about the same time as sensitization to work-related agents (166). However, after removal or diminution of exposure to both low- and high-molecular-weight agents causing OA, subjects are not at increased risk for developing IgE-mediated sensitization to common allergens, indicating that atopic status does not increase even years after the diagnosis of OA (167).

Cigarette smoking has been reported to be associated with the development of OA in workers exposed to platinum salts and anhydride compounds, which are chemicals that cause asthma through an IgE mechanism (168, 169). In workers exposed to platinum salts, cigarette smoking interacts with exposure to both high and low levels in inducing sensitization (170). Cigarette smoking also seems to affect the underlying mechanisms involved in OA, as the cellular composition of airway mucosa appears different in smokers with asthma and nonsmokers with potroom asthma (171). In addition, cigarette smoking increases the risk of sensitization to high-molecular-weight agents that cause OA through an IgE mechanism. In contrast, as reviewed by Mapp and Newman Taylor (172), cigarette smoking does not increase the risk of asthma caused by low-molecular-weight agents, such as diisocyanates and red cedar, for which a specific IgE is not usually considered the main mechanism of the development of the disease. Moreover, at the time of diagnosis of immunologic OA, the majority of cases are mild, but cigarette smoking is associated with a greater severity of disease (173).

Sex plays a role in the distribution of occupational lung diseases, because there are sex differences in specific jobs and therefore differences in exposure to agents causing these diseases (174). Women report significantly more exposure to cleaning products, biological agents, and textile fibers than men. In addition, one study reported that the risk of OA in the service sector was higher for women (175). A case-control study performed in Göteborg by Toren and coworkers (176) showed that among women, the risk of adult asthma was increased after exposure to paper dust and textile dust, whereas among men, the risk of adult asthma was increased after exposure to flour dust, welding fumes, man-made mineral fibers, and solvents. Because a strong association has been found between welding metal fume fever and welding-related respiratory symptoms suggestive of OA, it has been proposed that welding metal fume fever be viewed as a premarker of welding-induced OA (177).

Genetics
Genetic markers of occupational asthma have begun to be explored (178–182). As Apter (183) stated, genetic predisposition might be both a confounder and an effect modifier. Understanding of the interaction between genes and the environment is facilitated by the ability to precisely characterize the phenotype of an individual with OA. Also facilitated is the ability to assess and estimate exposure to the causal occupational agent, which in turn makes it possible to compare sensitized and nonsensitized individuals working in the same workplace with similar exposure.

Of interest is the pool of major histocompatibility complex genes on chromosome 6p, which encode the HLA class II molecules required for presentation of an antigen to a T-cell receptor to initiate the cascade of events that lead to an antibody response. Data obtained in occupational studies indicate that major histocompatibility complex class II proteins are important factors for the specificity of the response to occupational agents such as acid anhydrides, diisocyanates, western red cedar, complex platinum salts, natural rubber latex, and animal proteins (184–189). The largest study of the association of HLA molecules, which assessed specific symptoms and sensitization to rat lipocalin allergens, showed that about 40% of OA in the population examined could be attributed to an HLA-DRß1*07 phenotype; the attributable proportions for atopy and daily work in an animal housing facility were 58 and 74%, respectively (189). The HLA associations identified in that study might be found for other major animal allergens, in light of the similarities between lipocalin animal allergens. The HLA phenotype is also a significant determinant of sensitization to complex platinum salts, because the strength of the HLA association varied with the intensity of exposure to the sensitizing agent (190). Evidence of HLA associations has not been confirmed for diisocyanates by some investigations (191, 192), but in these studies, accurate phenotypes of the subjects were not provided. Even though HLA class II molecules are involved in immune recognition of occupational agents, HLA associations are not strong enough to be used for prevention. A significant association could be due to a causal relationship, but it could also occur by chance (multiple comparisons without a prior hypothesis) or by linkage disequilibrium. Despite these limitations, especially for asthma induced by low-molecular-weight agents, for which the absence of a specific IgE might challenge the immunologic mechanism, HLA associations indicate that HLA class II molecules contribute to individual susceptibility to low-molecular-weight agents, and provide evidence of a specific immunologic response in asthma induced by these agents. For other occupational agents, such as natural rubber latex, HLA associations confirm the importance of T cells in the regulation of IgE responses (193).

A second pool of genes that could be involved in OA is the superfamily of glutathione S-transferase (GST), a family that is critical for protecting cells from oxidative stress products, including lipid peroxides. Evidence of the involvement of GST is that among subjects who were exposed to toluene diisocyanate (TDI) for 10 years or more, the frequency of the GSTP1 Val/Val genotype was lower in those who had asthma and in those with moderate to severe airway hyperresponsiveness to methacholine (194). Conversely, because the protective effect of homozygosity for the GSTP1* Val allele increases in proportion to the duration of exposure to TDI, one might argue that the role of GST and its allelic variants could lie in determining which subjects will have persistent asthma (180). Moreover, individuals lacking this genotype may, over time, exhibit ongoing TDI-induced proinflammatory processes and consequent airway remodeling, leading to irreversible asthma symptoms. However, a study of GST genotypes in workers exposed to various diisocyanates showed no significant association between the GSTP1 genotype and the risk of asthma (195). A possible explanation for the differences between these two studies is that a protective effect might be observed only in subjects exposed to diisocyanates for at least 10 years.

More recently, the same investigators showed that, in addition to GST, N-acetyltransferase genotypes may play a role in diisocyanate-induced asthma, especially TDI-induced asthma, in which the N-acetyltransferase slow acetylator genotypes posed a 7.77-fold risk of asthma (196). In this common type of OA, by contrast, HLA class I antigens and the tumor necrosis factor A-308G are not associated with susceptibility or resistance to the development of TDI-induced asthma (197). Similarly, in potroom asthma, no associations have been found between this type of OA and genotyping for the ß₂-adrenoreceptor, high-affinity IgE receptor, and tumor necrosis factor, confirming that the mechanisms involved in potroom asthma remain poorly understood (198).

Because genes affect virtually all human characteristics and diseases, what lesson can be learned by identifying a genetic predisposition to OA? It seems clear that the patient needs to receive information about the health risks. However, because asthma is considered a disease caused by complex interactions over time between genes and environment (7, 199), efforts should be made to use the genetic information appropriately. Moreover, any reported association between a genetic marker and risk for disease cannot be considered definitive until the findings of the study have been replicated (200, 201). A complete review of the ethical, legal, and social implications of genomic medicine has been published (202). Wisely, the author writes that the "DNA sequence is not the Book of Life."

	PATHOPHYSIOLOGY

TOP ABSTRACT CONTENTS OCCUPATIONAL ASTHMA: DEFINITION... EPIDEMIOLOGY AND RISK FACTORS PATHOPHYSIOLOGY NATURAL HISTORY AND LONG-TERM... CONCLUSIONS AND FUTURE RESEARCH REFERENCES

 
Immunologic OA: IgE Dependent and IgE Independent
The pathophysiology of immunologic OA usually involves an IgE-dependent mechanism. OA induced by IgE-dependent agents is similar to allergic asthma that is unrelated to work (203–210). Most high-molecular-weight agents (e.g., flour and animal proteins) induce asthma by producing specific IgE antibodies. Certain low-molecular-weight agents (e.g., platinum salts, trimellitic anhydride, and other acid anhydrides) also induce specific IgE antibodies, probably acting as haptens and combining with a body protein to form functional antigens (211, 212). Cross-linking of allergens with a specific IgE antibody on the surface of mast cells, basophils, and possibly macrophages, dendritic cells, eosinophils, and platelets, gives rise to a cascade of events that result in the influx and activation of inflammatory cells and in the release of preformed and newly formed inflammatory mediators that orchestrate the inflammatory process.

Other low-molecular-weight agents, such as diisocyanates and plicatic acid, cause OA that has the clinical and pathologic features of immunologic asthma, but do not consistently induce specific IgE antibodies (206, 213–215). It has been suggested that, when specific antibodies to plicatic acid are present, they may be markers of exposure and not causes of disease (216). However, even though an IgE antibody to diisocyanates, which has strong positive response levels (an RAST score of 3 or greater) is present and exhibits high specificity (217), it has no sensitivity in detecting OA. The sensitivity increases when a blood sample is taken less than 30 days from the last exposure, which is consistent with the observed approximate 6 months of half-life of IgE.

Specific inhalation challenge with these low-molecular-weight agents in sensitized subjects induces various patterns of asthmatic reactions, including isolated early or late asthmatic reactions, a biphasic reaction, a progressive reaction, or atypical reactions (218). Atypical reactions have been reported principally after exposure to isocyanates. These reactions are uncommon after exposure to high-molecular-weight agents (219, 220).

The airway inflammation process is similar in IgE-dependent and IgE-independent asthma (221–223) and is characterized by the presence of eosinophils, lymphocytes, mast cells, and thickening of the reticular basement membrane (Figure 1) (222, 224). In the bronchial airways, inflammatory cells are not only increased in number but are also activated (224), resulting in the secretion of a wide range of proinflammatory mediators and proteins; these mediators and proteins have a variety of harmful effects, such as toxic damage to epithelial cells. In the airway inflammatory process of OA, eosinophilia is associated with an increased number of T cells, especially CD4⁺ cells, which exhibit signs of activation (224). Increased expression of lymphocyte markers, such as IL-2 receptor and very late activation antigen-1, has also been found (225). Along with the increased expression of lymphocyte activation markers, in asthma induced by low-molecular-weight agents (e.g., diisocyanates) an increased number of cells producing proinflammatory cytokines has been reported (225, 226). These proinflammatory cytokines, produced primarily by mononuclear phagocytes, may contribute to airway inflammation by several mechanisms, including increased expression of adhesion molecules, chemotaxis, and stimulation of inflammatory leukocytes. Together with these findings, antigen stimulation of monocyte chemoattractant protein 1 and tumor necrosis factor has been shown in isocyanate-induced asthma, suggesting that isocyanate-specific cellular immune reactions result in activation of macrophages, which may be important in the pathogenesis of this type of OA (227).

View larger version (43K): [in this window] [in a new window]   Figure 1. Schematic summary of possible mechanisms in occupational asthma (OA). Causal agents of OA are categorized into high-molecular-weight (HMW) and low-molecular-weight (LMW) agents. Exposure to high levels of respiratory irritants can induce irritant-induced asthma. HMW agents are recognized by antigen-presenting cells (APCs) and mount a CD4 type 2 immunologic response leading to production of specific IgE antibodies by interleukin (IL)-4/IL-13–stimulated B cells. Certain LMW agents also induce specific IgE antibodies, probably acting as haptens and combining with a body protein to form functional antigens. However, most LMW agents do not consistently induce specific IgE antibodies. In this type of OA, a mixed CD4/CD8 type 2/type 1 immunologic response or induction of /-specific CD8 may play a role. Inhalation of high levels of irritants may damage airway epithelium. In subjects who develop irritant-induced asthma, alarm signals from damaged epithelial cells might in turn activate immunocompetent cells. Binding of IgE to their receptors, Th2 (IL-5) and Th1 (IFN-) cytokines, and other proinflammatory chemokines (monocyte chemoattractant protein 1 [MCP-1]; tumor necrosis factor [TNF-]) induce recruitment and activation of inflammatory cells. These cells (mast cells, eosinophils, macrophages, and, in some instances, neutrophils) characterize airway inflammation, which contributes to the functional alterations of OA, that is, airway hyperresponsiveness and airflow obstruction. Subepithelial fibrosis due to thickening of the reticular basement membrane is considered a histopathologic feature of OA. However, the role of this remodeling of the airways in lung function is obscure.

Irritant-induced Asthma
The mechanism of asthma induced by irritants is unknown (233). Many reports indicate that unintentional high-level respiratory irritant exposures can induce the new onset of asthma (39). Because this type of OA occurs after inhalation of high levels of irritants, the main target for the initial injury could be the bronchial epithelium, which becomes denuded and loses its protective properties. Consequences of the damage in the bronchial epithelium are the loss of relaxing factors derived from epithelium, the exposure of nerve endings leading to neurogenic inflammation, and the release of inflammatory mediators and cytokines after the nonspecific activation of mast cells (234). A further consequence of the disruption of the epithelium accompanied by the secretion of growth factors for epithelial cells, smooth muscle, and fibroblasts, together with matrix degradation, is a tissue-regenerative and remodeling response (234). Pathologic changes consist of marked fibrosis of the bronchial wall and denudation of the mucosa with fibrinohemorrhagic exudates in the submucosa (235). In an animal model, chlorine exposure causes functional and pathologic changes in the airways resulting from oxidative stress (236).

Controversial Issues
Specific antibodies and OA.
Controversial issues exist regarding the pathophysiology of OA. One of these controversies is the role of specific antibodies in asthma induced by low-molecular-weight agents, such as diisocyanates and western red cedar. In clinical practice, a test for the diagnosis of this type of asthma would be appreciated. Many laboratories have investigated the usefulness of specific antibodies, for example, in diisocyanate-induced asthma, with conflicting results. Specific IgE antibodies to TDI in a range of 0 to 50% of workers have been reported (237–240). If these antibodies have an RAST score below 3, they are not highly specific (217). However, if these antibodies have an RAST score of 3 or greater, they exhibit high specificity and therefore are diagnostic of isocyanate-induced asthma. An explanation for the variable results in prevalence studies of serum-specific IgE antibody to TDI–human serum albumin conjugate in TDI-induced asthma has been proposed by Park and coworkers (241–243). These investigators found that the specific IgE binding to a new antigenic determinant of TDI–human serum albumin conjugate can be heterogeneous and can differ between one individual and another.

It is still not clear whether IgE-mediated responses contribute to the development of asthmatic symptoms in workers exposed to TDI. In fact, specific IgE responses to isocyanates are detected in a minority of isocyanate-induced asthma. The half-life of specific IgE differs in published studies on diisocyanate-induced asthma, ranging from a few months to several years (239, 240, 244). Moreover, it has been reported that the presence of high levels of serum-specific IgE at initial diagnosis may indicate a better prognosis (244, 245). Because the more convincing evidence for a role for specific IgE comes from studies conducted on subjects living in Korea, a country where TDI is the most common cause of asthma, it would be reasonable to investigate the possible differences that exist between this population and other populations regarding antibody-mediated immunity in this type of asthma. To complicate the picture, the results of specific IgE responses to other diisocyanates, such as methylene diisocyanate and HDI, may be different from the responses to TDI (246).

If the role of specific IgE responses in asthma induced by low-molecular-weight agents looks uncertain, the role of specific IgG responses to occupational agents seems even more complex. Serum-specific IgG may persist for many years after the last exposure to TDI (244). Because the sensitivity of specific IgG in the diagnosis of TDI-induced asthma based on the results of specific inhalation challenge is higher than that of specific IgE, but the specificity is poor, it has been suggested that IgG could be used to monitor the effect of exposure to diisocyanates before clinical disease appears (247). For other occupational agents, conflicting results have been reported, showing that the prevalence of specific IgG is significantly higher in symptomatic workers, with no correlation with the level and duration of exposure (248), or that the presence of specific IgG may be a response to high levels of exposure but is unrelated to the development of respiratory symptoms (249, 250).

Neutrophils and OA.
Another controversial issue is the role of neutrophils in OA. In the model of TDI-induced asthma, which has been widely investigated because diisocyanates are an important cause of OA in many countries, the role of neutrophils is unclear. Neutrophilia was found in the bronchoalveolar lavage fluid of subjects with TDI-induced asthma, especially in those exhibiting a late asthmatic response after specific inhalation challenge (251); however, the levels of TDI used in the challenge were higher than those used in the studies that followed. In nonoccupational asthma, the presence of neutrophils is considered a marker of the severity of the disease (252). What is unclear in OA is whether neutrophilia can also be considered a marker of severe disease due, perhaps, to higher levels of exposure, or whether, for unknown reasons, the same level of exposure may produce an inflammatory airway infiltrate in which eosinophils or neutrophils are the predominant cells (253, 254). Park and coworkers, by examining bronchial biopsies, serum, and induced sputum, found that, in TDI-induced asthma, neutrophils play a role (255, 256). They proposed a role for IL-8 released after exposure to TDI in the activation of neutrophils. Mast cells or basophils could be the origin of the neutrophil chemotactic activity found after TDI challenge (257).

Similarly, in grain dust–induced asthmalike disorder, the number of neutrophils in the bronchial mucosa is higher than that in allergic asthma, and the levels of IL-8 in induced sputum are significantly higher after inhalation challenge with grain dust extract than at baseline (258). The actual cause for these higher levels is currently unknown. Some of the clinical, functional, and pathologic features of grain dust asthmalike disorder may be reproduced by endotoxin inhalation.

Endotoxin levels tend to be highest in environments where there are farm animals, because the fecal flora of larger mammals is a major source of endotoxin. Chronic exposure to significant levels of endotoxin is associated with the development and/or progression of many diseases, including asthma and chronic bronchitis (259, 260). Because work-related asthma symptoms in workers exposed to laboratory animals occur even in those not sensitized to occupational antigens, other factors in the workplace may explain these symptoms, including endotoxin. The few investigations on the role of endotoxins in respiratory symptoms observed in technicians handling small animals gave controversial results (97, 261, 262).

Potroom asthma.
Potroom asthma has been recognized for many years in workers employed in the production of aluminum smelting, where the alumina is partially dissolved in an electrolyte of molten cryolite at about 960°C (263–265). Only one study has documented a dual asthmatic reaction with an associated increase in nonspecific airway responsiveness after exposure to the potroom workplace (266). Bronchial biopsies obtained in workers with potroom asthma have shown that pathologic alterations are similar to those described in other types of asthma (171). Often, this occupational environment has not been included among causal agents of OA because different mechanisms could be involved including an immunologically mediated reaction against trace amounts of metals, or an irritant effect resulting from exposure to hydrogen fluoride and sulfur dioxide. A dose–response relationship has been described for fluoride exposure and airway hyperresponsiveness in potroom workers (267), but whether fluorides are the causative agents, coagents, or simply markers for the causative agent(s) of potroom asthma remains to be determined.

	NATURAL HISTORY AND LONG-TERM CONSEQUENCES

TOP ABSTRACT CONTENTS OCCUPATIONAL ASTHMA: DEFINITION... EPIDEMIOLOGY AND RISK FACTORS PATHOPHYSIOLOGY NATURAL HISTORY AND LONG-TERM... CONCLUSIONS AND FUTURE RESEARCH REFERENCES

 
Chan-Yeung and Malo (6, 268) have provided a "schema" of the natural history of OA with a latency period. For each stage (the onset of exposure; the development of sensitization and asthmatic symptoms; the occurrence, in some cases, of rhinoconjunctivitis symptoms; the onset of airway inflammation; the development of OA; the cessation of exposure; and the cure or persistence of asthma), one might investigate the role of modulating factors, including host markers, characteristics of both the occupational agent and exposure, and the effect of therapy.

To summarize the natural history of OA and the role of specific agents, the risk of OA is highest soon after the first exposure, because most subjects develop asthma within 1 to 2 years of exposure. Nevertheless, the latency period can vary from months to years (6). The rate of acquiring both sensitization and asthmatic symptoms may differ according to the nature of the agent (268), and the intensity of exposure. For example, workers exposed to proteins of laboratory animals usually develop sensitization during the first 2 years after the first exposure, but bakers exposed to flour do not, suggesting that animal allergens are more potent sensitizers than flour (102). In workers exposed to laboratory animals, rhinoconjunctivitis symptoms usually occur in years 1 and 2 after the first exposure, but the incidence of respiratory symptoms is greater in years 2 and 3 (269). However, Rodier and coworkers (270) reported on a high incidence of occupational rhinitis among new recruits to animal laboratories. Rhinoconjunctivitis symptoms often precede the onset of asthma symptoms in workers exposed to high-molecular-weight agents but less frequently in those exposed to low-molecular-weight agents (271). It should be underlined that patients with occupationally induced rhinitis have a high risk of asthma. In fact, in a study of 3,637 patients with work-induced rhinitis, the risk of asthma was highest in the first year after notification of occupational rhinitis, and a roughly threefold risk persisted for several years thereafter (272).

OA is a potentially fatal condition: death from asthma has been reported for a subject exposed to diisocyanates, for a baker, and for a subject exposed to shark cartilage dust (273–275).

Most subjects who develop OA after a latency period do not recover, even several years after cessation of exposure. Symptoms and nonspecific airway hyperresponsiveness persist in about 70% of affected workers. The percentages of those with persistent nonspecific airway hyperresponsiveness after removal from exposure are lower for electronic (276) and aluminum potroom workers (277). Moreover, removal from exposure is associated with worse socioeconomic outcome (278).

Important determinants of recovery are the total duration of exposure, the duration of symptoms, the severity of asthma, the lung function, the degree of airway hyperresponsiveness at the time of diagnosis, and the duration of follow-up. Workers who, after diagnosis of OA, continue to be exposed to the same causal agent worsen with time. The most effective treatment of immunologic OA is complete avoidance of exposure, which is associated with improvement in asthma symptoms (wheezing, shortness of breath, cough) and functional variables (airway hyperresponsiveness). However, Vandenplas and coworkers (279) showed that workers allergic to natural rubber latex were able to function when exposure was reduced, improving in symptoms and keeping their jobs.

Once a worker is sensitized to an occupational agent, bronchial reactivity usually persists. Most workers with OA retain specific bronchial reactivity to the occupational causal agent even two or more years after removal from exposure (280). Some individuals show an asthmatic reaction to occupational agents despite the normalization of nonspecific airway responsiveness. This finding has practical implications, because some subjects who are considered cured and return to their previous jobs are at risk for an asthmatic reaction when reexposed to the causal agent. Following up this observation, Lemière and coworkers (281, 282) showed that, despite treatment, the absence of asthmatic symptoms, and normal nonspecific airway responsiveness, subjects with OA induced by exposure to high-molecular-weight agents (e.g., flour, psyllium, and guar gum) and with high levels of specific IgE reacted to the causal agent within a few minutes of exposure. Thus, persistent immunologic sensitization appears to be a key factor in the persistence of specific bronchial reactivity to occupational agents. This finding emphasizes the importance for subjects with OA caused by high-molecular-weight agents to avoid exposure, even if they are asymptomatic and have normal airway responsiveness. Similarly, in workers exposed to isocyanates, despite removal from exposure to the offending agent, persistence of specific bronchial reactivity to these low-molecular-weight agents has been reported (283, 284). Is persistent immunologic sensitization the unique factor in causing persistence of sensitization to occupational agents? Does this factor also cause the persistence of asthmatic symptoms in subjects with OA? And how must we consider subjects who have airway hyperresponsiveness but are asymptomatic: as "sensitized" or as "asthmatics"? One might argue that these asymptomatic subjects are, of course, still sensitized but that they also have "latent disease," because it is likely that if they return to work and are exposed to the same causal agent, their asthma will again become clinically detectable.

Consistent with persistent symptoms and nonspecific airway hyperresponsiveness in about 70% of workers with immunologic OA, studies of the outcome of OA indicate that resolution is a slow process that continues for years (285). In general, cessation of exposure is associated with an improvement in symptoms, and when glucocorticoids are used the asthma is further improved but not cured (286). Only a few longitudinal studies of airway pathology in subjects with OA after cessation of exposure have been conducted (287–291). In a follow-up study of diisocyanate-induced asthma, although subepithelial collagen deposition decreased, airway eosinophilia persisted (289). It is still not known whether the persistence of OA is related to genetic susceptibility or to persistence of an inflammatory process in the airways. These two possibilities are not mutually exclusive, because genetic susceptibility may affect the capability of an individual to deal with airway inflammation (194).

Airway inflammation is considered to be the cause of asthma, whereas the role of airway remodeling in the pathogenesis and severity of the disease remains controversial (292). Some investigators think that thickening of the airway wall causes airway hyperresponsiveness whereas others think it protects against airway narrowing and attenuates airway reactivity in individuals with asthma (293). However, airway remodeling provides an explanation for many conditions observed in subjects with asthma, such as corticosteroid-resistant airway hyperresponsiveness (294) and the accelerated decline in lung function that occurs over time in adult asthma (295). The traditional view that airway remodeling is caused by longstanding inflammation needs to be changed. Studies of children with asthma showed that the remodeling begins early in the development of asthma and might occur in parallel with inflammation or even be required for the establishment of persistent inflammation (296). Thus, in adult-onset asthma, where both airway inflammation and airway remodeling are present, one might argue with the traditional view that airway inflammation occurs first and remodeling later.

What can we learn about the role of airway remodeling from studying the pathology of OA? A longitudinal study of airway pathology performed in subjects with OA caused by TDI showed that the histopathologic characteristics of asthma and the classic indicators of remodeling, such as a thickened reticular layer of the basement membrane, were present at diagnosis (289) and therefore may be considered part of the inflammatory response to the offending agent (297). However, because a thickened reticular layer of the basement membrane, even if reduced, was present 6 to 21 months after the cessation of exposure, it may also be considered a marker of long-lasting structural changes of the airway wall (298).

At the present time, it is difficult to establish the relative contributions of airway inflammation and remodeling to the chronicity of OA; probably, both are associated with the persistence of asthmatic symptoms and of nonspecific airway hyperresponsiveness. Epithelial cells, smooth muscle cells, bronchial vessels, eosinophils, myofibroblasts, macrophages, and sensory neuropeptides (299–301) may contribute to both processes, and the secretion of cytokines, chemokines (302–304), growth factors, and matrix metalloproteinases (305) may help to establish a particular tissue microenvironment that sustains the chronicity of the disease.

Diagnosis
Diagnosis of OA should be confirmed by objective testing for asthma and then by establishing the relation between asthma and work (306, 307). Physicians should consider the possibility of OA in all adults with asthma; therefore, an occupational history should be the first step in the initial evaluation of the patient. The diagnosis should be confirmed as soon as possible to prevent worsening of symptoms and should be investigated when workers are at the workplace, because a prolonged avoidance of exposure may influence the reliability of diagnostic procedures (308). More accurate criteria are required for medical purposes than for case identification in field surveys.

The diagnosis of asthma is based on a compatible history and the presence of variable airflow limitation or, if lung volumes are normal, of nonspecific airway hyperresponsiveness, which is generally assessed by means of a histamine or methacholine challenge test. The two protocols, recommended by the American Thoracic Society (309), should be expected to give different results. If greater sensitivity of the test is sought, deep breaths should be avoided (i.e., by using specific airway conductance), whereas if greater specificity is sought, a measurement requiring deep breaths (e.g., FEV₁) is recommended (310). Because measurements of nonspecific airway responsiveness are log-normally distributed in the population (311), an arbitrary cutoff (e.g., a PC₂₀ FEV₁ of 8–16 mg/ml) has been selected in defining an abnormal test. Cut points are highly sensitive, but have low specificity. The test may be positive in cases of allergic rhinitis, cystic fibrosis, bronchiectasis, chronic obstructive pulmonary disease, sarcoidosis, and congestive heart failure (312). The strength of the methacholine inhalation challenge is its negative predictive value, because a negative response is unlikely to occur in subjects with asthma (313). However, in the occupational setting, normal methacholine test results have been reported in subjects with asthma induced by diisocyanates (314). These normal results could have been due to the avoidance of exposure for several days or more. In this common type of asthma, and, in general, in each type of sensitizer-induced OA, in the absence of changes in FEV₁ in response to specific inhalation challenge, nonspecific airway responsiveness should be assessed before and after the specific inhalation challenge (314–316).

The other diagnostic information, occupational history, should be accurate (317). The physician should ask the patient about the frequency and intensity of exposures and the frequency of exposures to peak concentrations of the potential causal agent, although many patients are not aware of the agents to which they are exposed at work. The physician should be aware of the potential exposures experienced in different occupations (Table 5). Additional information can be obtained from a visit to the workplace by experts in occupational hygiene, from material safety data sheets for workplace chemicals, and from the manufacturers of the workplace substances. Lists and databases of the etiologic agents are available (71). The physician should also be aware of the following: that the presentation of OA is variable, some subjects developing asthma symptoms immediately on exposure and others developing symptoms after several hours; that the absence of airway hyperresponsiveness does not rule out sensitization to an occupational sensitizer; and that not all symptomatic workers have "true" OA attributable to a specific agent (318).

A history of improvement of symptoms during weekends and holidays and a worsening on return to work suggests OA. Rhinoconjunctivitis usually precedes the onset of asthma symptoms in the case of high-molecular-weight agents (271