Is Isocyanate Exposure and Occupational Asthma Still a Major Occupational Health Concern? Systematic Literature Review

Isocyanate, whose disease-inducing mechanism is poorly understood, with poor prognosis, is widely used. Asthma is the most frequent manifestation of prolonged exposure. We assessed the evolution of the incidence of isocyanate-induced occupational asthma over time. PubMed and Cochrane databases were systematically searched for studies published since 1990 that assessed the relationship between occupational exposure to isocyanates and asthma. We identified 39 studies: five retrospective cohort studies, seven prospective cohort studies, three of which were inception cohorts), seven observational cross-sectional studies, five literature reviews, two case series, and 13 registry studies. The incidence of occupational asthma secondary to isocyanate exposure has decreased from more than 5% in the early 1990s to 0.9% in 2017 in the United States. Despite the wide use of optimal collective and individual protection measures, the risk of occupational asthma has stabilized. Occupational asthma risk can be assessed with good sensitivity using self-questionnaires and pulmonary function tests. Occupational avoidance should be implemented as soon as possible after the first symptoms appear because the prognosis becomes increasingly poor with the persistence of exposure. It is now necessary to study specifically cutaneous sensitization to isocyanates and to define what protective equipment is effective against this mode of exposure.


Introduction
Organic isocyanates are highly unstable chemical compounds, with at least one isocyanate function (-N=C=O), due to this function which is highly unsaturated. This explains their widespread industrial use, the complexity of the mechanisms of action of the induced pathologies as well as their poor prognosis, both on the medico-occupational and medicosocial levels [1]. Polyurethane, in the family of plastics, is formed when they react with polyol. They are used in many sectors of activity, such as the automotive industry, footwear, construction, foundries, tanneries, the electronics industry, the production of molded parts, printing, painting, maintenance [2].
The main route of contamination is respiratory, as these products are mainly used by spraying, which leads to the presence of isocyanates in the air in the form of aerosols. Asthma, the most common disease caused by prolonged exposure to isocyanates, is one of the most common occupational respiratory diseases in industrialized countries [3]. Asthma is a respiratory disease characterized by wheezing and coughing and chronic inflammation of the airways that can cause exacerbations and progressive lung function decline. The diagnosis is confirmed by an alteration in peak expiratory flow on spirometry. In France and other industrialized countries, in 2016, the average annual incidence of occupational asthma was estimated at 30.3 cases per million workers, i.e., an average annual decrease of 15.8% [3]. However, this incidence is known to be underestimated and should be considered as the minimal rate [4].
Although flour is the most frequent etiological agent for occupational asthma, isocyanates are the main chemical cause in France and are responsible for between 8 and 17% of cases [3,5]. Elsewhere the prevalence varies between 5 and 10% of cases, depending on the country [1].
The aim of this systematic literature review was to assess the evolution of the frequency of occupational asthma due to isocyanate exposure over time.

Information Sources and Search Strategies
PubMed and the Cochrane Library were searched from 1990 to 15 April 2021. The following reference keywords were identified using the terms 'occupational asthma' and 'isocyanates' in the Medical Subject Headings Terms (MeSH Terms) tool: asthma, occupational; work-related asthma; WRA; isocyanates.
The following search strategies were used:

Selection of Articles
Eligible publications were published between 1990 and 15 April 2021 in English or French and reported the diagnosis of de novo occupational asthma with work-related exposure to isocyanates. We excluded publications reporting results from animal studies, clinical trials, morbidity and mortality assessments, non-occupational asthma, studies of long-term evolution of occupational asthma or after occupational avoidance, studies of the effects of comorbidities on asthma severity, studies and recommendations on legal exposure values, studies of the correlation between dose and the severity of symptoms, studies and recommendations on personal protective equipment, comparative studies of diagnostic strategies, genetic studies, and analysis of biomarkers.
The records were initially screened on the titles and abstracts then the PDFs of selected publications were screened to identify those that corresponded to the inclusion criteria.
The search and analysis were carried out by EC with verification by the co-authors in compliance with PRISMA Guidelines [6].

Data Collection and Analysis
Data on the study design population characteristics, country, isocyanate studied, target sector, size of the study population, duration of follow-up, diagnostic criteria for asthma, the results, evidence of work-relatedness, and the main limitations of the study were extracted into tables and analyzed.

Results
Of 650 records identified from PubMed and the Cochrane Library, 189 were selected from the initial screen. The second screen resulted in the selection of 39 articles for inclusion in our systematic review: 12 cohort studies (five retrospective and seven prospective cohort studies, of which three were inception cohorts); seven observational cross-sectional studies; five literature reviews; two case series; and 13 registry studies.
The majority of studies were conducted in the United States, Commonwealth countries and Europe, demonstrating that it is an international issue in industrialized countries.
The main sectors of activity concerned were the automotive industry (seven studies), chemical industry (six studies), wood industry (two studies), foundries and molding plants (two studies), other industries (four studies).

Automotive Industry
We identified seven studies: four cross-sectional studies, two cohort studies and a case series published between 1996 and 2017 that reported the use of a variety of isocyanates in the automotive industry [7][8][9][10][11][12][13] (Table 1).
In 1996, Simpson et al. reported that 41% of workers had altered PFT after 10 weeks of exposure to BIC (butyl isocyanate) prepolymer in the absence of IPE. However, these disorders were resolved after the introduction of air-supplied masks [7].
Other studies reported a proportional relationship between the duration of exposure to isocyanates and the frequency and intensity of respiratory symptoms [8][9][10][11]. Small and confined workspaces were associated with an increased risk and the use of an air-supplied mask was associated with reduced risk [12]. Poor respiratory symptoms and the presence of specific IgG were reported to be prognostic markers for the development of occupational asthma [13]. The incidence of occupational asthma was estimated to be about 10% among car-body painters, and this increased with exposure to 23% in subjects who had been exposed for more than 20 years [8].

Chemical Industry
Six studies reported results for chemical industry employees: five cohort studies and one case series, covering the period from 1999 to 2017 (Table 2). TDI and HDI are the main isocyanates used in this industry [14][15][16][17][18][19].
A case series study by Kraw and Tarlo published in 1999 reported an incidence of occupational asthma of 2.5% over five years of regular exposure, without information on the use of IPE [14]. A cohort study, using data from employees' medical files, found 6.4% (19/313) suspected cases of occupational asthma (without a formal diagnosis) in employees who had been exposed to TDI for at least five years [15]. A higher incidence of occupational asthma in recently recruited employees who had previously been exposed to isocyanates was also reported.
In two studies conducted in companies where air measurements did not show a reading above the exposure limits, no occupational asthma cases were identified after 20 years of retrospective follow-up [16,17]. This finding was confirmed by a shorter inception cohort study (18). The proportion of employees with respiratory symptoms suggestive of occupational asthma (without a formal diagnosis) was 15% among exposed individuals, even if the exposure limits were not exceeded. In the absence of biomonitoring, the possibility that the employees were exposed to occasional peaks of isocyanates, without thus causing asthma cannot be excluded [17].
The cohort study by Collins et al. [19] estimated the incidence of occupational asthma at 0.009 cases/person-year in a population of workers exposed to TDI in different US companies. However, this figure is probably underestimated due to the non-inclusion of contract workers and the absence of formal diagnostic criteria.

Wood Industry
We identified two cohort studies in the wood industry which used MDI [20,21] ( Table 3). One study by Petsonk et al., published in 2000, reported no cases of occupational asthma, but 12% of the employees were symptomatic after two years of exposure to MDI [20]. The results from this study also suggested a link between the appearance of work-related breathing difficulties and the reported intensity of exposure, smoking status, poor use of respiratory IPE or exposure to skin splashes. The second study by Wang and Petsonk, published in 2004, aimed to identify early symptoms following daily exposure to MDI. The authors observed significant onset of wheezing (at rest), cough and chest tightness [21]. Daily productive cough was associated with smoking status, not occupational exposure. The appearance of new symptoms after two years of follow-up was reported for 13% of the exposed workers [21].

Foundries and Molding Plants
We identified two cohort studies in the foundry sector, which is also known for its use of isocyanates, particularly in the manufacture of sand molds [22,23] (Table 4). In 1993, Bernstein et al. published a study of employees exposed to MDI for three years and showed a prevalence of occupational asthma of about 1% (three cases in 243 employees) under standardized working conditions, with continuous monitoring that did not exceed the 0.05 ppb threshold, and work carried out in a closed environment with ventilation [22]. The prevalence was inversely proportional to the total exposure. The prevalence of mildly irritated respiratory symptoms was 11%, which is consistent with our current knowledge of occupational risk.
A retrospective cohort, published in 2002, reported a prevalence of chest tightness of 16% over five years, under similar exposure conditions, i.e., undetectable levels. There were no cases of occupational asthma [23].

Other Industries
Four other studies were identified in sectors of activity that traditionally use fewer isocyanates (textiles industry, bone glue factories and various industries) [24][25][26][27] (Table 5). These studies reported contradictory results. A retrospective cohort by Baur et al., published in 1994, analyzed data for 1780 employees in various industries and, with a rigorous diagnostic approach using an isocyanate-specific inhalation test, estimated that exposure to MDI or TDI had no effect on lung capacity in previously healthy subjects and a moderate impact in asthmatic subjects [25]. However, a cross-sectional study reported 13% of exposed subjects had positive specific inhalation test results and 26% had symptomatic disease [24]. Ten years later, Al-Batanony et al. reported a relative risk of four in subjects exposed to TDI with a concomitant decrease in FEV1 and FVC proportional to the duration of exposure [26]. Another study reported the prevalence of exposure among workers in Australia in 2014 [27]. The authors estimated that 3% of the Australian working population was exposed to isocyanates in the workplace. The three most exposed occupations were painters (34.7%), woodworkers (24.4%) and construction workers (17.7%).

Literature Reviews
We identified five literature reviews that assessed all types of isocyanates in a varied population of workers from different industries [28][29][30][31][32] (Table 6). In 2000, Van Kampen et al. estimated the prevalence of all respiratory diseases attributable to 5% isocyanates at 5% [28].
Two other reviews estimated that, despite a drastic increase in the production and use of isocyanates, there has been a clear decline in the incidence of occupational asthma over the past 50 years from more than 5% of exposed workers before 1980 to less than 1% since 2000 [29,30]. They attributed this decline to the widespread use of individual protective equipment in industry.
Another study by Jarvis et al. suggested that the occupational asthma hazard of isocyanate is proportional to the number of NCO groups that the molecule contains, but the authors were not able to exclude potential interaction between different substances [31].
A 2007 review confirmed the decrease in the incidence of isocyanate-related occupational asthma due to the sharp decrease in exposure by the respiratory route but warned of the potential role of skin exposure in the occurrence of respiratory signs [32].

Epidemiological Public Health Registers
Occupational asthma can be monitored in registry studies (Table 7). We identified 13 publications on national health monitoring registers in the United Kingdom, France, South Africa and Belgium published from 1992 to 2020 [33][34][35][36][37][38][39][40][41][42][43][44][45]. The data from these registers suggest an overall decrease in occupational asthma cases over time which is indexed on the industrialization rate of each region, as well as a decrease in the proportion of occupational asthmas related to isocyanates. The incidence of occupational asthma in the general working population was reported to range from 13 to 25 per million workers [35,37,40,42,43]. In highly industrialized regions such as the West Midlands in England, the incidence was more than 40 per million workers [33]. The incidence was much higher for jobs at risk of exposure, in particular car bodyworkers, with more than 1500 cases per million workers in the United Kingdom before 2000 [33,34], and 326 cases per million workers in France in 2002 [38].
Isocyanates were reported to be the leading cause of occupational asthma in the UK [33,36,39,41] and the second in France, after flour [38,44]. However, there has been an overall decrease in the percentage of occupational asthma attributable to isocyanates in all industrialized countries [34,45]. These observations were reported to be correlated with an increasingly systematic avoidance in the workplace after diagnosis [39].

Discussion
The objective of this literature review was to assess the risk of occupational asthma related to chronic exposure to isocyanates. Measuring the impact of various prevention and awareness strategies on the incidence of occupational asthma and its evolution over time remains an important issue in occupational medicine and health, as these compounds are still widely used in many industrial sectors.
Out of 650 records identified from PubMed and the Cochrane Library, 39 articles were selected following the PRISMA Guidelines, including studies from several industrial sectors, five literature reviews and 13 registry studies.
The incidence rate of isocyanate-induced occupational asthma in the US has fallen from 5% before 1980 to 0.9% in 2017 [19]. This decrease in risk has undergone two turning points: firstly, there was a clear and sharp decrease in the number of new cases reported at the start of the 2000s. The generalization of respiratory prevention measures in industry over the last 40 years is probably a major factor in this evolution [29]. In 1991 in France, 142 cases were compensated, corresponding to an incidence of one case per 100,000 employees. In 2016, the number of cases compensated fell to 23, i.e., an incidence divided by 6.5 [46]. Since 2007, the decrease in annual incidence appears to have slowed down considerably, with 31 cases recognized per year on average [46].
This observed decrease in incidence may be explained by the under-reporting of pathologies or by the delay in diagnoses and in declaring the case as an occupational disease. However, this decrease could also be explained by the reduced use of isocyanatecontaining products or the more effective implementation of preventive measures described in labor laws in many countries. Nevertheless, the current preventive strategies, which are now widely used and almost optimal, seem to have reached their limits.
The results of this literature review are consistent with our current knowledge. We observed that the risk of occupational asthma is still present when isocyanates are handled, despite the use of appropriate individual and collective respiratory protection measures. We also noted that the main respiratory symptoms (resting dyspnea, coughing, chest tightness and wheezing) usually appear without delay and are mostly resolved when exposure is stopped [21], although in the studies analyzed here, proven cases of occupational asthma after several years of isocyanate exposure were not consistently reported, probably because of the short follow-up time in the studies. Therefore, we assessed the incidence of respiratory symptoms suggestive of asthma, which seemed to be a reliable predictive sign of the risk of occupational asthma [21,22].
The occupations most at risk appear to be those in the manufacture of resin-bound flexible foams [29,47] and the use of spray paint [48]. In Australia in 2014, the occupations with the highest proportion of employees exposed to isocyanates were still painters (34.7%), woodworkers (24.4%) and construction workers (17.7%) [27].
The data concerning the most harmful type of isocyanate and exposure are contradictory: MDI would be more asthmatogenic in the inhalation test because of its rapid polymerization [25,27], but it would be less dangerous than TDI because it is less volatile and not heated. Similarly, the use of oligomers and prepolymers would seem to be preferable to isocyanate monomers, the latter being more easily inhaled [27]. The risk of occupational asthma would seem to be highest for di-isocyanates (two groups on the same molecule), while the risk is not clearly established for mono-isocyanates [48]. In 2012, a literature review reported that long-term, low-intensity exposures did not appear to be more of a risk for occupational asthma as short, high exposures [49], whereas other studies reported that it is the duration of daily exposure that is predictive for the risk of occupational asthma [17,50]. However, it appears that exposures at undetectable levels may result in pulmonary symptoms such as tightness of the chest [23].

Limitations of the Review
The quality of the occupational asthma epidemiological studies was often insufficient and well-conducted prospective cohort studies often lacked power, limiting the precision of their results. In addition, they mostly had short follow-up periods, which are not compatible with the latency of occupational asthma, and therefore tended to underestimate its real incidence. Another limitation frequently stated by the authors of these cohort studies is the large number of dropouts during the follow-up period, which leads to the suspicion of a 'healthy worker effect' bias, due to patients who are most likely to have respiratory symptoms either not accepting to participate at recruitment or by dropping out during the study. One study reported that those lost to follow-up, because they left their job after a short period, were often in poorer health than long-term employees [10].
In comparison, the retrospective studies, with potentially very long follow-up, are intrinsically subject to information and recall bias and therefore, the accuracy of the conclusions drawn is less robust. Finally, the literature reviews analyzed synthesized highly variable and not easily comparable data, as the included studies were generally disparate.
However, all these limitations are inherent to occupational health studies and do not call into question the health effects observed. It should be emphasized that the comparison between studies is often difficult because of their non-uniform methodology. We have analyzed data from different types of studies, which were conducted in dissimilar settings, and which included populations that were not always representative of the populations actually exposed to isocyanates. For example, subjects with a pneumological history were excluded from some studies, even though they had been assigned to high-risk occupations, and in some studies, participants had previously been exposed to isocyanates [7,10].
The type of isocyanate studied, which varied between studies, is also a potential confounding factor. The diagnostic criteria used for asthma varied between studies, ranging from a simple declaration of symptoms by questionnaire, without medical examination to more formal tests such as PFT or specific bronchial provocation tests. Finally, as environmental measurements were not systematically reported, it is impossible to estimate with precision the degree of exposure in each study.
This heterogeneity observed between the various published studies is the consequence of the persistently poor understanding of the pathophysiology of the disease, a non-consensual diagnostic strategy, and non-uniform prevention measures used in different countries. The causal link with the occupational environment should also be more rigorously defined in a standardized manner in studies.
Moreover, we cannot exclude the possibility that the deliberately restrictive inclusion criteria, used to target a very precise subject, led to the non-inclusion of studies that may have reported data that would have been useful for this review. In addition, we limited the review to publications in French and English published in the last 30 years, which may have resulted in an incomplete literature review.

Conclusions
Occupational asthma caused by isocyanates is still a current occupational health problem, and its reduction over time has stagnated for several years despite the implementation of individual and collective preventive measures. Several animal studies included in the literature reviews that we analyzed have shown that sensitization to isocyanates in animals is possible by the cutaneous route [18,20]. In humans, we currently have only indirect evidence of pulmonary sensitization by this route, i.e., urinary biomarkers and HDI-conjugated keratin found in employees whose workplace was subject to analytical measurements of respiratory exposure below the detection threshold [32,51]. It would be very useful to study this exposure route more closely, as it could explain why the incidence of isocyanate-related occupational asthma seems to have plateaued. This could then help to define which skin protection equipment would be most effective against these agents.
We also need to develop precise, rigorous and reproducible methods that should be used in future cohort studies. This common methodology would enable us to perform meta-analyses with acceptable power to compare the relative risk for isocyanate used between different occupations and different types of industry.