Published ahead of print on May 31, 2007, doi:10.1164/rccm.200611-1620OC
© 2007 American Thoracic Society doi: 10.1164/rccm.200611-1620OC
Bronchiolitis Obliterans Syndrome in Chemical Workers Producing Diacetyl for Food Flavorings1 Netherlands Expertise Centre for Occupational Respiratory Disorders, Utrecht, The Netherlands; 2 Division of Environmental Epidemiology, Institute for Risk Assessment Sciences, Utrecht University, Utrecht, The Netherlands; and Departments of 3 Pulmonology and 4 Radiology, University Medical Centre Utrecht, Utrecht, The Netherlands Correspondence and requests for reprints should be addressed to Frits G.B.G.J. van Rooy, M.D., Division of Environmental Epidemiology, Institute for Risk Assessment Sciences (IRAS), P.O. Box 80.176, NL-3508 TD Utrecht, The Netherlands. E-mail: g.b.g.j.vanrooy{at}iras.uu.nl
Rationale: Workers in microwave popcorn plants are at risk of developing bronchiolitis obliterans associated with exposure to butter flavoring volatiles, including diacetyl. Objectives: To investigate the risk of bronchiolitis obliterans for chemical workers producing diacetyl, with exposure to less complex mixtures of chemicals. Methods: We interviewed and conducted spirometry on 175 of 196 workers from a chemical production plant that produced diacetyl between 1960 and 2003. We used all available historical exposure data to classify all workers into three exposure groups with varying exposure profiles to diacetyl, based on frequency and level of exposure. Measurements and Main Results: Workers with fixed airway obstruction underwent further pulmonary function testing (including diffusing capacity and lung volumes) and paired inspiratory and expiratory high-resolution computed tomography studies. We identified three cases consistent with bronchiolitis obliterans syndrome with air trapping on high-resolution computed tomography of the lungs, in the highest exposure group of 102 process operators. Two of these cases were lifelong nonsmokers. Potential exposures included acetoin, diacetyl, acetaldehyde, and acetic acid, with diacetyl exposures in the range previously reported to be associated with fixed airway obstruction in the microwave popcorn industry. Conclusions: Exposure to an agent during diacetyl production appears to be responsible for causing bronchiolitis obliterans syndrome in chemical process operators, consistent with the suspected role of diacetyl in downstream food production.
Key Words: popcorn • bronchiolitis • exposure • occupation • diacetyl
Several recent publications indicate a new, potentially severe occupational lung disease in workers exposed to flavorings in North American food processing industries ("popcorn worker's lung"). These studies include clinical case studies in former employees, cross-sectional epidemiologic studies among current employees, explorative exposure studies, and a few animal exposure studies (1–9). Carefully performed clinical case series of workers occupationally exposed to inhalable flavoring vapors documented a rare, severe lung disease, consistent with the diagnosis of bronchiolitis obliterans syndrome (BOS). Within the spectrum of butter flavoring vapors, diacetyl plays a prominent role (2, 10). Until now, no cases of BOS have been reported in chemical production industries related to flavorings or outside of North America. The novelty of the present study is that it was conducted in the chemical industry in a European diacetyl production plant. A cohort of workers of the plant who were potentially exposed to diacetyl in the period 1960–2003 was identified by employment records. Among these workers, cases of BOS were not previously detected or suspected by the occupational health service or otherwise. The aim of this study was to investigate in a population-based approach whether there were cases of BOS in the cohort of workers exposed to diacetyl, to examine the degree of risk, and to assess exposures. Some of the results of this study have been previously reported in the form of an abstract (11).
Study Design and Population The chemical plant producing diacetyl was based in The Netherlands. The Human Resources department identified 206 workers who had potentially been exposed to diacetyl in the period 1960–2003, of whom 10 had died. Their cause of death was not investigated. We traced the remaining 196 and obtained written, informed consent from 175 (89%). Participants completed a self-administered standardized questionnaire supplemented with questions about respiratory, mucous membrane, and atopic symptoms and work history. The questionnaire is based on one with items described in National Institute for Occupational Safety and Health Health Hazard Evaluation Reports (12, 13), and some of the questionnaire items were taken from the European Community Respiratory Health Survey.
Exposure Assessment
The plant produced diacetyl by oxidation of 2,3-butylene glycol into acetyl-methyl-carbinol (AMC; acetoin), which was further partly oxidized into diacetyl (2,3-butanedione). Acetaldehyde and acetic acid were side products of the two primary reaction products, AMC and diacetyl. Diacetyl production took place in a completely closed system in a reactor vessel at an elevated process temperature ( Limited routine exposure monitoring was done by company representatives using cartridges containing silica gel coated with dinitrophenylhydrazine, and each sample was analyzed by an external laboratory for both diacetyl and acetaldehyde using gas chromatography. A total of 26 ambient samples (sampling duration, 82–219 min) and four personal task-based samples (33–90 min) taken between 1995 and 2003 could be traced.
Spirometry
Clinical Evaluation A diagnosis of BOS was made in cases with fixed airway obstruction and an HRCT scan with features of air trapping with hypoattenuation in segmental or lobular areas and a mosaic pattern of perfusion (16, 17).
Epidemiologic and Statistical Analysis
Characteristics of the Study Population Table 1 shows the characteristics of the study population that completed a questionnaire and underwent spirometric lung function tests in 2005. Regression analysis showed that operators had significantly lower FEV1 values (–292 ml) than workers in the internal reference group (Table 2). Regression coefficients for age and standing height were comparable to those from reference regression equations from the European Respiratory Society (14). Current smokers generally had lower lung function compared with nonsmokers, but ex-smokers did not show the expected pattern of lower lung function compared with nonsmokers.
Exposure Assessment The air concentrations, as determined by area sampling, ranged from 1.8 to 351 mg/m3 for diacetyl and from 0.4 to 29 mg/m3 for acetaldehyde. Control measures taken in 2001, with the aim to enclose the process, led to a reduced exposure for both diacetyl (geometric mean change from 10.0 to 5.8 mg/m3) and acetaldehyde (geometric mean change from 7.6 to 0.7 mg/m3). During production, process operators were potentially exposed to diacetyl during several tasks but discharge of diacetyl in containers had the highest exposure potential for process operators. Personal task-based sampling results during this activity ranged from 3 to 396 mg/m3 diacetyl and 0.2 to 14 mg/m3 acetaldehyde. These exposure data were collected for compliance testing of acetaldehyde exposure. Although exposure to diacetyl was mainly relevant for process operators, workers with several other occupational titles were also potentially exposed to diacetyl. Maintenance workers were likely to be exposed to diacetyl but exposure was highly variable. Laboratory workers were potentially exposed to diacetyl but no further qualitative or quantitative information was available. All other occupational titles (transport, health and safety, and management and research and development personnel) had low exposure potential for diacetyl and, if they had exposure, it was always of short duration. The diacetyl production plant was one of several plants in operation at the production site. Most production workers, including all operators, also worked in other chemical plants on the production site. As a result, all workers were also potentially exposed to other chemical agents, including irritants such as ammonia and chlorine.
Clinical Investigation
All three subjects had been process operators. Two of them were lifelong nonsmokers, and the third was a 59-year-old smoker with a cumulative smoking history of 14 pack-years. Their symptoms had started 10 to 20 years earlier, on average after 5 years of employment at the diacetyl plant (Table 3). None had ever been hospitalized for a pneumonia or an infection, showed signs of a connective tissue disorder or colitis, used medication known to cause bronchiolitis obliterans, or undergone radiotherapy in the past.
All three subjects had previously been diagnosed by chest physicians as having chronic obstructive pulmonary disease (COPD) or asthma, for which they were treated by inhaled Lung function tests were available for cases from the Occupational Health Service through 1994 and thereafter from hospital records (Figure 1). The average FEV1 decline during employment was 175 ml/year for case 1 (1995 to 2003) and 55 ml/year for case 3 (1986 to 2001). No lung function data were available for case 2 during employment. All cases reported exertional dyspnea as the first symptom and none had an acute onset or reported coughing or phlegm.
All three workers had confounding exposure from other production facilities within the plant during the same time period. Cases 1 and 2 reported no inhalation accidents. Case 3 reported a 1997 inhalation accident 12 years after his first reported chest symptoms and after commencement of accelerated FEV1 decline. He was exposed to acetaldehyde in the diacetyl production process during a few minutes without respiratory protection after which he immediately became dyspneic. In 1998, he consulted a chest physician because symptoms worsened.
Lung Function
Radiology
Thoracic Lung Biopsy Case 1 underwent video-assisted thoracoscopic lung biopsy, with wedge biopsies taken from the right upper and lower lobe. Histology showed no signs of constrictive bronchiolitis. In some sections of the biopsies, emphysema and chronic bronchiolitis were present, reflecting nonspecific small airway disease.
Our study found a cluster of BOS cases in a diacetyl production plant. This supports the conclusion that an agent in the diacetyl production process has caused BOS in process operators. The clinical characteristics and physiologic and imaging features of the cases are consistent with the cases identified in a popcorn plant in the United States (4). The third case, a modest smoker, had an unusual pattern of air trapping combined with a nearly complete collapse of the central bronchial tree during expiration. He did not have an active smoking history or physiology sufficient to explain COPD. A case with similar features in a diacetyl-exposed group was reported by Kreiss and coworkers (2). Lung biopsies obtained in one case showed nonspecific small airway disease. The main histologic finding in bronchiolitis obliterans is a constrictive bronchiolitis as seen in patients with connective tissue disease and in bone marrow and (heart–)lung transplants. Because lesions are often patchy or may present as centrilobular emphysema, the diagnosis may be missed (18). When pathology is lacking, the term "obliterative bronchiolitis" may more accurately describe the clinical syndrome—that is, airway obstruction due to bronchiolitis. In the study by Akpinar-Elci and colleagues (4), three cases underwent thoracoscopic lung biopsies. In only one case, histology supported a diagnosis of constrictive bronchiolitis. The other two cases showed severe airway obstruction and mosaic patterns on HRCT, but bronchiolitis was not found in the lung biopsies, which was attributed to a sampling error. Although bronchiolitis obliterans is a histologic diagnosis, the typical features on HRCT obviate the need for biopsy in those with occupational exposure to a recognized agent. Moreover, the risk of surgical intervention and sampling error should be weighed against the need for a histologic diagnosis. Cases of bronchiolitis obliterans are easily missed (4). HRCT may be normal during inspiration as BOS requires expiratory HRCT to visualize air trapping. In smokers under middle age and especially in nonsmokers with chronic airway obstruction, a diagnosis other than COPD should be considered. Clusters of such cases in work environments provide opportunities to identify new etiologies of bronchiolitis obliterans.
Population at Risk The decline of FEV1 value of cases was higher than expected in periods in which they worked at the diacetyl plant. Typically, FEV1 shows average declines of 24 ml/year in nonsmokers and 28 ml/year in smokers in longitudinal studies (19). The excess declines may have been affected by spirometry measured in different settings with different devices and technicians, but each of the cases had quite consistent declines despite different data sources.
Exposure
Causal Relationship between BOS and Diacetyl Exposure In this study, as in the study by Kreiss and colleagues (2), diacetyl may either be a cause of respiratory disease or a marker of another causative exposure. The spectrum of exposures is much smaller in this production plant compared with the popcorn processing plants where a wide range of chemicals was identified. Consistent with animal inhalation studies showing airway injury with diacetyl vapor (10), our study suggests a causal role of diacetyl. However, we cannot rule out a possible contribution of acetoin or even acetaldehyde (24), either as causative or contributing agents.
Study Limitations Moreover, our analyses of exposure have not addressed the potential importance of short-term or peak exposures among workers because such information was lacking. By not pursuing decedents and not obtaining clinical evaluation of one case of fixed airway obstruction, we may have underestimated the number of BOS cases. Indeed, an additional case was found after the study among the 10 nonparticipants among the former workers at the plant. This post-study case was a 63-year-old nonsmoking process operator who worked from 1968 to 1970 and stopped working in 1970 when he became symptomatic and was diagnosed with severe COPD (FEV1, 31% predicted; FEV1/FVC, 37%). The 2006 HRCT scan of the lungs was compatible with a diagnosis of bronchiolitis obliterans. Thus, at least four cases of clinical bronchiolitis obliterans arose in 206 workers employed in this diacetyl manufacturing plant. This population-based survey establishes the presence of BOS, or popcorn worker's lung, in chemical workers manufacturing a flavoring ingredient with exposures to diacetyl, acetoin, and acetaldehyde. Any or all of these exposures may contribute to the risk of this emerging occupational lung disease.
The authors thank Dr. Kay Kreiss, M.D., who reviewed the draft manuscript and provided many helpful suggestions for improvements, and Dr. Jeffrey Myers, M.D., for revising the pathology slides.
Originally Published in Press as DOI: 10.1164/rccm.200611-1620OC on May 31, 2007 Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form November 10, 2006; accepted in final form May 31, 2007
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
AT A GLANCE COMMENTARY
360°C). Exposure for process operators only existed at the end of the production process and they did not have exposures to heated product. 
-agonists and corticosteroids. Case 1 used a long-acting anticholinergic agent; none received maintenance oral corticosteroid therapy. 
