INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Hypersensitivity pneumonitis (HP) is a worldwide lung disease caused by an immune response to a large variety of inhaled antigens. Responsible environments contain large quantities of airborne bacteria, fungi, or avian proteins. Typically, HP manifests itself as recurrent febrile episodes 3 to 8 h after antigenic contact, dyspnea, cough, and chest tightness. Weight loss eventually occurs also if the disease process is prolonged (1).

Clinical evaluation usually reveals inspiratory crackles, interstitial infiltrates on chest radiographs, restrictive alterations in lung function associated with a decrease in lung diffusion capacity, a lymphocytic alveolitis on lung lavage, and loosely formed granulomas on lung biopsies (2).

Exhaustive listings of environments and causative agents associated with HP have been published (1). The most frequent entities are farmer's lung, bird fancier's disease, and humidifier lung. Specific entities differ from one region of the world to another. A typical example of this is the importance of summer-type HP in Japan (3), a disease not found in America or Europe.

Peat moss, an organic substance derived from sphagnum moss, is harvested in many areas of the world. This product is most often used as fuel to heat homes, as a soil additive, or in the fabrication of organic filters. Although this material can be contaminated by molds (4), no report of HP associated with it has to our knowledge been reported. In a previous study of workers in a peat moss packaging plant, no evidence of HP was found (5).

Large deposits of peat moss are harvested in eastern Canada. Peat moss harvesters act like giant vacuum cleaners and suck the loose dry material as they sweep large open fields. This collected material is stacked in huge piles to further dry before being conveyed to a bagging plant. It is in these stacks that, depending on the humidity of the collected material, the growth of microorganisms can occur. In a previous study we have shown that Penicillium sp. and Monocillium sp. constitute the most abundant molds present in these stacks (4). Peat moss can also contain various mycobacteria, some of which are known pathogens (6). Sporotrix schenckii is also a common contaminant of peat moss (7).

Cutaneous sporotricosis is the most frequently reported disease associated with the handling of peat products (8), but other organs can also be infected (9). Considering that Penicillium molds are known to be responsible for some cases of HP (1), the lack of prior description of this disease in workers exposed to this material is surprising.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Case Study

A 50-yr-old ex-smoker, employed as a repair man in a peat moss processing plant, was referred to our institution with a history of progressive dyspnea and episodic bouts of afternoon fever over a 3-mo period. The patient noted that the febrile episodes always occurred at the end of a work shift and were not present on days off from work. These episodes culminated in a severe bout of dyspnea, which led to his admission to a rural hospital. On admission the patient was tachypneic and dyspneic at rest. He was cyanotic, his temperature was 40° C, and bilateral inspiratory crackles were heard over both lung fields. Chest radiographs revealed diffuse interstitial infiltrates. Arterial blood gas analysis while the patient breathed room air showed a Pa_O2 at 49 mm Hg and a Pa_CO2 at 37 mm Hg. HP was suspected, and the patient was administered supplemental O₂ and, on the following day, transferred to us for further evaluation; his fever had now abated and his dyspnea was slightly improved. Lung functions showed a severe restrictive pattern (TLC = 55%, RV = 66%, and VC = 58% predicted) with a reduction in lung diffusion capacity (DL_CO = 44% predicted). High resolution computed tomography (HRCT) showed patchy alveolar infiltrates. A bronchoscopy was performed to obtain a bronchoalveolar lavage (BAL) and transbronchial lung biopsies. Cellular counts were 1,225 × 10³ cells/ml; of these 57% were lymphocytes. Lung biopsies showed a lymphocytic infiltration with multiple multinucleated giant cells compatible with acute HP. Subsequent serology identified specific antibodies of molds found in the peat moss used in the plant (see below). The patients was treated with high dose corticosteroids (50 mg prednisolone/d) for 1 wk. His symptoms rapidly abated and this treatment was discontinued. Reevaluation 1 mo later showed that all sign and symptoms had resolved; chest radiographs and lung functions had returned to normal. Note that he had not resumed contact with the work environment.

We contacted the company where this man worked and obtained permission to: (1) take samples of the material used in the plant to analyze for possible microbial contamination, and (2) to examine the other workers for evidence of respiratory disorders and markers of an immune (antibodies) or inflammatory (cytokines) response to their work environment.

Microbiologic Analyses

Microbiologic analyses were performed on each of the six products used to prepare a commercial mixture of a peat-moss-based soil additive. Three 1-g samples of each of these products were placed in sterile plastic containers with 50 ml of sterile saline containing 0.05% Tween 80 and 1 g of glass beads (0.45 mm) (Braun, Melsungen, Germany) to increase the friction and make washing more efficient. The samples were vigorously agitated with a vortex intermittently for 15 min. Serial dilutions were made from the solution obtained (10⁰ to 10³) in sterile saline solution). One hundred microliters of these dilutions were plated on appropriate media for total bacterial counts and thermophilic actinomycetes, Tryptic Soy Agar (Difco Laboratories, Detroit, MI) supplemented with cycloheximide (500 mg/L) was used and incubated at 30° C for bacteria and 52° C for thermophilic actinomycetes counts. Molds were grown on Rose Bengal Agar (Difco) supplemented with chloramphenicol (50 mg/L). Microorganisms were counted and molds were identified by genus. For Penicillium, John I. Pitt's laboratory guide (10) was used for species identification.

Evaluation of Workers

Thirteen of the 14 coworkers agreed to participate in the initial phase of our study. The study was approved by our Ethics Committee, and all workers signed an informed consent form. This initial evaluation consisted of a clinical questionnaire on respiratory symptoms, past medical history, current medication, work history and smoking status, and a physical examination. Spirometry, using the standard ATS criteria (11), was performed, and blood samples were taken. One nonsmoker refused to give blood. The serum samples of the other 12 workers and that of the index case would be used to detect specific antibodies to any potential antigens identified in the peat moss and to measure proinflammatory cytokine levels.

Characteristics of the 13 workers are presented in Table 1. Thirteen control subjects not exposed to this work environment (all hospital workers) were selected to match the workers for age, sex, and smoking history. These control subjects also underwent forced expiratory flow testing and venous blood samplings. The blood sample taken from a control subject replaced that of the worker who refused to give blood, and it was used as a control for the serum of the index case for whom no matched control was selected for the lung function comparisons. This control subject was, as was the index case, a nonsmoker.

Measurement of Specific Antibodies

MaxiSorp Immuno plates (Nunc, Roskilde, Denmark) were coated with 100 µl of 250 µg/ml of total extracts of molds previously isolated from peat moss and incubated overnight at 4° C. The next morning, plates were washed and saturated with 1% bovine serum albumin to block residual protein binding sites of the plate. Specific antibodies in the serum were determined by enzyme linked immunosorbance assay (ELISA). Serum dilutions of 1:500 were used for these measurements. Duplicates of diluted samples were incubated in coated plates and antihuman IgG coupled to peroxydase with ortho-phenylenediamine as the substrate were used. Absorbance was read at 490 nm.

Subpopulation Study

Six of the 13 workers evaluated at the plant were selected and further evaluated at our hospital. These workers included the case suspected of having HP at the site visit, three asymptomatic nonsmokers of light smokers with positive antibodies to the molds tested, and two asymptomatic seronegative smokers. These six subjects had chest radiography, HRCT, BAL, and serum cytokine analysis.

Bronchoalveolar Lavage

Under local anesthesia, a fiberoptic bronchoscope was introduced through the mouth into the trachea and wedged into a segmental or subsegmental bronchus. Five 60-ml aliquots of warm 0.9% saline were instilled and the fluid was gently aspirated after each aliquot. BAL fluids were collected and kept on ice until centrifugation at 500 g for 10 min. Cell-free BAL fluid supernatants were aliquoted and frozen at 70° C for cytokine determination. Cells recovered from the BAL were resuspended in RPMI 1640 medium supplemented with 2 mM L-glutamine, 5% heat-inactivated fetal calf serum, and 1% penicillin-streptomycin (all from Canadian Life Technologies Inc., Burlington, ON, Canada). Cells were counted and viability was estimated by trypan blue exclusion. Cell differential counts were performed on Diff-Quik-stained glass-cover preparations (12). Lymphocyte subpopulations were studied by flow cytometry (Epics Elite ESP; Coulter Electronics, Burlington, ON, Canada). Alveolar macrophages (AM) were purified by adherence to plastic. Briefly, AM (1 × 10⁶ cells/ml) were plated into a 24-well plate, incubated for 2 h at 37° C/5% CO₂ and gently washed with Hanks' balanced salt solution. Adherent cells were more than 98% macrophages as verified by Diff-Quik coloration. AM were incubated for 24 h in complete RPMI, and the supernatants were collected and frozen at 70° C for further analysis of spontaneous cytokine release.

Cytokine Measurements

Levels of tumor necrosis factor alpha (TNF-) and interleukin-6 (IL-6) in unconcentrated BAL fluids and sera were assayed by high sensitivity immunoassay kits from R&D Systems Inc., (Minneapolis, MN). In vitro release of these cytokines by AM were assayed using conventional kits from the same company.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Results of the microbiologic analyses of different materials used in the plant are given in Table 2. Peat moss was the most heavily contaminated material. Monocillium sp. and Penicillium citreonigrum were by far the most prevalent molds found. High levels of bacteria were also found. These were of multiple colony morphologies. No thermophilic actinomycetes were found.

Beside the index case, one of the workers solicited had clinical signs and symptoms suggestive of HP. He complained of progressive dyspnea, had inspiratory crackles over both lung fields, and had decreased lung functions. His FEV₁ was 62% predicted, and his FVC was 70% predicted.

Lung functions showed that for the group as a whole, excluding the index case, workers had a significantly lower FVC (p = 0.02) with a normal FEV₁/FVC ratio, suggesting a restrictive pattern (Figure 1).

View larger version (26K): [in this window] [in a new window]   Figure 1.   Results (means ± SD) for the FEV1, the FVC, and the ratio of the two variables for the 13 workers and 13 control subjects. Results are given in percent predicted. For each variable, columns followed by different letters are significantly different. FVC was significantly lower in the workers than in the control subjects (p = 0.02, nonpaired t test).

Specific serum IgG antibody levels, reported in optical density, to molds found in peat moss are presented in Figure 2. Note that the index case and the subject with a clinical suspicion of HP both had high titers of specific antibodies to Monocillium sp. and Penicillium citreonigrum. Specific antibodies were also present in the other minimal smoker (two cigarettes per day) and in four of the five nonsmokers. None of the heavy smokers had serum antibodies to these two molds. Specific antibodies were present in only one nonsmoking control subject.

View larger version (12K): [in this window] [in a new window]   Figure 2.   Optical density of the ELISA testing for specific antibodies to Penicillium found in peat moss. Individual values are given for the 12 workers who accepted this part of the study, the index case, and 13 nonexposed control subjects. The second highest value was that of the index case. The two minimal smokers (see text) are included with the nonsmokers. Note the strikingly lower values in the exposed smokers compared with their nonsmoking coworkers (p = 0.003, Fisher's two-tailed exact test).

The diagnosis of HP was confirmed in the suspected case. This diagnosis was based on restrictive lung functions with decreased DL_CO, a lymphocytic alveolitis, patchy alveolar infiltrates on HRCT scan, and positive serology to molds contaminating the peat moss. The other five workers evaluated at our institution had normal lung functions, chest radiographs, and HRCT. Results of their cellular BAL findings, along with that of the index case and the other identified case of HP, are presented in Figure 3. Note that the two cases and the asymptomatic workers had an increased proportion of CD8+ lymphocytes (Table 3). One of the two seronegative smokers had too few BAL lymphocytes to measure these subpopulations.

View larger version (12K): [in this window] [in a new window]   Figure 3.   (A) Total number of cells per milliliter of fluid recovered. (B) Percent alveolar macrophages. (C ) Lymphocytes in the BAL of the seven workers who had this procedure: the index case and the other worker with HP, three asymptomatic seropositive nonsmokers, and two seronegative smokers.

Higher levels of TNF- and IL-6 were detected in the BAL fluid of the two patients with HP compared with the three asymptomatic subjects with a lymphocytic alveolitis and the two normal seronegative subjects (smokers) studied (Figure 4). Similarly, a fixed number (1 × 10⁶ cells) of AM from the two patients with HP spontaneously secreted higher amounts of TNF- and IL-6 than AM from the two seronegative subjects. AM from asymptomatic seropositive subjects released intermediary amounts of these inflammatory cytokines (Figure 4).

View larger version (16K): [in this window] [in a new window]   Figure 4.   BAL fluid levels of tumor necrosis factor alpha (TNF-) and interleukin-6 (IL-6) in picograms per milliliter and the spontaneous release of these cytokines by alveolar macrophages for the subjects with hypersensitivity pneumonitis (HP), and the asymptomatic seropositive and seronegative smoking workers.

TNF- and IL-6 were also measured in the sera of the 13 peat moss workers (the index case and the 12 who accepted this evaluation) and 13 normal unexposed subjects (Table 4). TNF- was significantly higher in the sera of the exposed workers than in that of the control subjects (p = 0.01). No difference in IL-6 levels were detected between the two groups. TNF- and IL-6 serum levels were higher in patients with HP than in the asymptomatic seropositive and the seronegative smoking workers, but because of the small number of subjects, no significant difference was found.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The significance of this report is not just yet another isolated cause of HP. The large proportion of workers with either the disease or sub-clinical manifestation of an immune response to this environment suggests that this workplace may constitute an important risk of HP for a large number of workers. In the province of Quebec alone, more than 700 workers are employed in similar processing plants as the one evaluated in the current study. In view of our prior results from another plant it is likely that all workers are not equally at risk (5). The risk may vary from 1 yr to another depending on dry weather conditions. Cross-sectional and follow-up studies of larger numbers of workers and different plants and in different countries are needed to establish the global prevalence and impact of HP in the peat moss industry.

The high prevalence of subjects with evidence of immune response to these two molds (two HP cases, all but one of the nonsmokers or light smokers), and the very high CD8+ proportion of the BAL lymphocytes attest to either the high level of exposure and/or the exquisite immunogenicity of the antigens. CD4/CD8 ratios are usually decreased in the lung lymphocytes of HP (13) and asymptomatic exposed workers who have lymphocytic alveolitis (14).

The cytokine results are consistent with an increased state of activation of AM during the acute or chronic phase of HP. Although no increase of TNF- and IL-6 was seen in BAL fluid of three asymptomatic workers, cultured AM from these subjects did secrete high amounts of these cytokines in vitro. The presence of fewer AM in BAL of asymptomatic subjects could explain why high levels of TNF- and IL-6 were not found in their BAL fluids. No TNF- and IL-6 could be measured in BAL fluid of the seronegative subjects or in the supernatant of cultured AM from the same subjects. These subjects were smokers and their results are consistent with the fact that AM from cigarette smokers are poor producers of inflammatory cytokines such as IL-1, IL-6, and TNF- (15). Overall, TNF- was higher in the sera of the exposed workers than in that of the unexposed normal control subjects, no significant difference was observed for IL-6.

The results of this study are added evidence for the previously described protective effect of cigarette smoking on the immune response to inhaled antigens responsible for HP (16). The interest of this observation is not that we should encourage exposed workers to smoke but to use this information to guide studies on the components of cigarette smoke and the mechanisms by which smoking confers such protection.

In conclusion, this study describes a new and potentially important environment where workers are at risk of developing hypersensitivity pneumonitis, identifies the causative antigens, and confirms a protective effect of cigarette smoking.