Bronchoalveolar Mast Cells in Normal Farmers and Subjects with Farmer's Lung Diagnostic, Prognostic, and Physiologic Significance1. 2
MICHEL LAVIOLETTE, YVON CORMIER, ALAIN LOISEAU, PAUL SOLER, PIERRE LEBLANC, and ALLAN J. HANCE
Introduction
Hypersensitivity pneumonitis is an acute interstitial lung disease resulting from an immune response directedagainst inhaled organic antigens (1-3). These patients are sensitized against the inhaled antigen as manifested by the presence of precipitins and antigen-specific T lymphocytes in the lung and peripheral blood (4-6). Furthermore, in animal models, susceptibility to disease can be transferred with lymphoblasts (7). In sensitized individuals, antigen exposure results in the rapid recruitment of neutrophils and eosinophils into the alveoli; these inflammatory cells likely play an important role in producing the acute alveolar damage and symptoms seen in such patients (8). The mechanisms responsible for the recruitment of inflammatory cells are incompletely understood and could be mediated by antigen-antibody complexes and/or the release of cytokines in the course of the immune response. Although this scenario for the pathogenesis of the disease is attractive, a major paradox exists. Many asymptomatic individuals exposed to antigens causing hypersensitivity pneumonitis also are sensitized to these antigens (9). Furthermore, the levels of precipitins (9), the subclasses of antibodies present (10), the intensity of the lymphocytic alveolitis (11-13), and the phenotype of T lymphocytes present in asymptomatic individuals (614) can overlap those seen in patients with acute hypersensitivity pneumonitis. The lymphocytic alveolitis can persist for 2 yr in normal farmers or asymptomatic patients with previous episodes of hypersensitivity pneumonitis without clinical evidence of progressive disease (15-17). Thus it is unclear why only some individuals develop symptoms. Recently, special attention has been given to the role of mast cells in this dis-
To evaluate the specificity and significance of increased lavage mast cells in farmer's lung, we examined the lavage cell differentials of 89 farmers and 19 normal nonfarming control subiects. The farmers were divided Into four groups: acute farmer's lung (n = 17), farmers with one or more prior episodes of farmer's lung who remained In dally contact with hay (n = 26) or quit farming (n = 14), and normal farmers (n = 36). A total of 14 of the subjects with prior episodes of farmer's lung and still farming and 15 normal farmers were evaluated twice at a 2-yr Interval. The lavage mast cell numbers were significantly higher in acute farmer's lung (7.5 ± 7.3 x 10'/ml, mean ± SO)and ex-farmer's lung who were still farming (1.2 ± 1.3 x 10'/ml) than in normal farmers (0.1 ± 0.1 x 10'/ml) (p < 0.01). A total of 8 of 14 exfarmer's lung patients who had quit farming and 18 of 36 normal farmers had an Increased number of mast cells In lavage, but mast cell count never exceeded 0.5% of total recovered cells. In the acute farmer's lung and ex-farmer's lung-still farming groups, the mast cell count correlated with the lymphocyte count: r 0.83 and r 0.69 (p < 0.001), respectively. In the two groups evaluated twice, mast cell numbers at the first study did not correlate with changes seen at the second study In chest roentgenogram and pUlmonary functions. Weconclude that an Increase in lavage mast cells occurs commonly as a part of the immune response against thermophilic bacteria. However, this increase In the number of lavage mast cells Is found in farmers without clinical signs of farmer's lung, parallels that of lymphocytes, and does not predict the outcome. Consequently the role and clinical significance of an Increased number of lavage mast cells in farmers with or without farmer's lung remains undefined. SUMMARY
=
=
AM REV RESPIR DIS 1991; 144:855-860
ease (18-22). Patients with hypersensitivity pneumonitis have markedly increased numbers of mast cellson the alveolar surface, and the number of mast cells recovered by lavage from patients with hypersensitivity pneumonitis is in general higher than that observed in other interstitiallung diseases. Mast cellscould play an important role in the pathogenesis of the disorder, for example by recruiting inflammatory cells to the lung. Furthermore, it has been suggested that the presence or absence of mast cells could explain why only some sensitized individuals become symptomatic (18). To further explore the importance of mast cells in hypersensitivity pneumonitis, we reviewed lavage and clinical data from a large group of farmers whom we have studied since 1981 (23,24), including both asymptomatic individuals and subjects with recent or prior farmer's lung disease. The goals of the study were to determine if mast cells are always increased in farmer's lung, to evaluate the
possibility that the presence or absence of mast cells correlates with the likelihood of developing symptoms, and to identify factors that contribute to the appearance of mast cellson the alveolar surface of antigen-exposed individuals. Method Study Populations A total of 89 farmers from rural communities surrounding Quebec City and 19normal control subjects were included in the study. Most of these subjects had participated in prior studies, and their clinical characteristics (Received in original form February 13, 1990 and in revised form March 1, 1991) 1 From the Unite de Recherche, Centre de Pneurnologie, Hopital et Universite Laval, Ste-Foy,Quebec, Canada, and INSERM 0.82, Faculte de Medecine Xavier Bichat, Paris, France. 1 Correspondence and requests for reprints should be addressed to Michel Laviolette, M.D., Centre de Pneumologie, H6pital Laval, 2725, Chemin Ste-Foy, Ste-Foy, Quebec GlY 405, Canada.
855
856
LAVIOLETTE, CORMIER, LOISEAU, SOLER, LEBLANC, AND HANCE
have been published elsewhere (15-17,23, 24). Briefly, the patients were divided into the following groups. Acute farmer's lung. Seventeen subjects (14 men and 3 women; 16 nonsmokers and 1 former smoker; age 40 ± 11 yr, mean ± standard deviation [SD]) were diagnosed as having acute farmer's lung by previously described criteria (24). The interval between the onset of acute symptoms and lavage ranged from 1 to 30 days; the individuals were studied at a median of 4 days after the last contact with the farm environment. Prior farmer's lung. Forty subjects diagnosed as having had one or more episodes of acute farmer's lung were lavaged more than 1 yr after the last acute episode. In this group, 4 subjects also were studied during their acute episode and are included in the group of acute farmer's lung. A total of 26 subjects (21 men and 5 women; age 47 ± 11 yr; 21 nonsmokers, 3 former smokers, and 2 smokers) remained in daily contact with hay in the interval between the last acute episode and the lavage. Fourteen individuals in this group underwent a second lavage 2 yr after the original study. Among these 14 patients, 3 individuals were identified who developed subsequent evidence of farmer's lung. A single individual had evidence of an episode of acute farmer's lung between the two studies and subsequently developed other acute episodes; another had clinical and physiologic evidence of progressive disease at the time of the second study, which subsequently improved; and a third developed acute farmer's lung after the second study. All 3 individuals gave a history of important antigen exposure. The other subjects did not demonstrate changes in physical examination, chest roentgenogram, or pulmonary function. In all current farmers lavage was performed during the winter months after the farmers had finished their morning farm chores. Fourteen of the sub-
jects with prior farmer's lung episodes (13 men and 1 woman; age 41 ± 9 yr; 6 nonsmokers, 7 former smokers, and 1 smoker) subsequently quit farming and had no known contact with hay for at least 2 yr. Normalfarmers. A total of 36 subjects were farmers with no past history of farmer's lung (33 men and 3 women; age 41 ± 14yr; 4 former smokers, 31 nonsmokers, and 1 smoker). A total of 18 subjects were precipitin positive for thermophilic Actinomycetes antigen, and 18 subjects were precipitin negative. A total of 15 individuals (6 precipitin positive and 9 precipitin negative) in this group were restudied 2 yr after the initial evaluation. None developed symptoms of farmer's lung during the interim, and chest radiographs remained normal. All normal farmers were lavaged during the winter months after finishing their morning farm chores. Control subjects. A total of 19 nonsmoking normal volunteers (11 men and 8 women; age 27 ± 6 yr) served as controls. None had signs or symptoms of present or past lung disease, and none was known to have exposure to antigens known to cause hypersensitivity lung diseases.
Protocol At the time of evaluation, all subjects underwent clinical examination, pulmonary function testing, chest radiography (except volunteers), and bronchoalveolar lavage, which were performed using previously described techniques (23, 25). All lavages were performed in the right middle lobe using a total volume of 300 ml normal saline solution. The mean volumes of fluid recovered were: acute farmer's lung, 60.8 ± 11.40/0; exfarmer's lung-still farming, 63 ± 9.5%; exfarmer's lung-quit farming, 67.2 ± 7.4%; normal farmers, 67.1± 7.6%, and nonfarming control subjects, 66.1 ± 8.3%. Lavage fluids were centrifuged, and cells
were resuspended in Hanks' balanced salt solution. Cytocentrifuge preparations were made, fixed in methanol, and stained with Wright-Giemsa stain as previously described (11). Cytocentrifuge preparations were coded, and a I,OOO-cell differential count including metachromatic cells was performed by one investigator (P.S.) without knowledge of the clinical characteristics of the subjects. For comparison, slides of farmers' groups also were counted blindly by another observer (M.L.). Because few metachromatic cells with the morphologic characteristics of basophiis were observed, all metachromatic cells were counted as "mast cells" (26). Although some authors have suggested that toluidine blue staining is more sensitive for the detection of mast cells, especially at low mast cell counts (26), this could not be evaluated in our study because unstained cytocentrifuge preparations were not available.
Statistical Analysis Results are expressed as mean ± SD. The distribution of residual error was not normal, as demonstrated by the Koimogorov D statistic using a SAS statistical package (SAS Insititute, Cary, NC). Consequently, for multiple comparisons between groups we used the Kruskal-Wallis test followed by a nonparametric analog Student-Newman-Keuls test. Comparison between two groups was made with Mann-Whitney U test for unpaired data and Wilcoxon's test for paired data. Comparisons of variables between groups were made using Fisher's exact test. For correlations we used the hypothesis test for the Spearman coefficient (27). The significance level was set at 0.05. In all cases the upper limits of normal for lymphocyte, neutrophil, eosinophil, and mast cell lavage numbers were determined by the one-tailed 95% confidence interval (mean ± 1.65 SD).
TABLE 1 NUMBERS AND TYPES OF CELLS RECOVERED BY LAVAGE Group Acute farmer's lung Cells x 103/m!' Percentage
Total Cell Count
Macrophages
Lymphocytes
Neutrophils
Eosinophils
Mast Cells
558.1 ± 256.0 A
126.4 ± 68.5 A 24.1 ± 10.3
349.4 ± 232.2 A 59.0 ± 15.0
66.1 ± 75.0 A 13.9 ± 15.2
8.6 ± 9.7 A 1.6 ± 1.6
7.5 ± 7.3 A 1.3 ± 0.8
312.2 ± 185.4 8
131.5 ± 91.aA 45.8 ± 20.3
168.0 ± 131.3 8 50.3 ± 19.7
10.8 ± 14.78 3.3 ± 3.5
2.0 ± 2.gB 0.6 ± 0.7
1.2 ± 1.3 8 0.4 ± 0.4
186.9 ± 149.5 c
133.7 ± 111.5 A 74.0 ± 16.2
45.9 ± 66.3 c 23.3 ± 16.4
4.5 ± 8.3c 1.8 ± 1.8
2.2 ± 5.9 8c 0.7 ± 1.3
0.2 ± 0.4 c 0.1 ± 0.2
129.8 ± 88.8 c
96.0 ± 67.gA 75.9 ± 16.2
27.9 ± 37.2 c 19.7 ± 14.8
4.7 ± 8.~ 3.3 ± 3.4
1.1 ± 3.1 8c 1.0 ± 3.8
0.1 ± O.l c 0.1 ± 0.1
55.9 ± 16.6 0
49.0 ± 14.3 8 88.0 ± 6.8
5.6 ± 3.7 0 9.7 ± 5.6
1.0 ± 1.3 0 1.7 ± 1.8
0.4 ± 0.1C) 0.6 ± 1.0
0.01 ± 0.03(D} 0.02 ± 0.05
2 Exfarmer's lungstill farming Cells x 103/ml Percentage
3 Exfarmer's lungquit farming Cells x 103/ml Percentage
4 Normal farmers Cells x 103/ml Percentage
5 Nonfarming controls Cells x 10'/ml Percentage
• Values expressed as cellslml were compared using the Kruskal-Wallis test followed by an analog Student-Newman-Keuls test. For each parameter, the values identified by a different lelter are different, p < 0.01 or p < 0.05 (Ielters in brackets).
BRONCHOALVEOLAR MAST CELLS IN FARMER'S WNG
857 A. LYMPHOCYTES
1000 100
......:' :3; a-
•
...•
.,. .,1. a ••• ••
10
A. ACUTE FARMER'S LUNG 40
• •••• • •
•
•
30 0 00
;;
Cbg8>
::1 ·1"" •• •
20
r" 0.1I3 p < 0.001
10
•
0
•
0
o
00
o
0
0
0
.1 Fig. 1. The number of lymphocytes (A), neutrophils (B), and mast cells (C) recovered by lavage from farmers with acute farmer's lung (Group 1),those with prior episodes of farmer's lung who either continued farming (Group 2) or quit farming (Group 3), normal farmers (Group 4), and normal nonfarming control subjects (Group 5). For normal farmers, individuals who were precipitin positive are designated by closed circles (Group 4A) and those who were precipitin negative are shown by open circles (Group 48). The solid line represents the upper limit of the normal range. Note that the ordinate is a logarithmic scale. Dots with number at the bottom of certain columns represent individuals in whom specific cells were not detectable.
0
100 10
•••
o
1000
5
••• ••• .:.
4
•
..-
3
..ir.
,::
•
--
1I00
6
.1
l"l
600
400
B. EX-FARMER'S LUNG, STILL FARMI:-; 200 x 1Q3 Iymphocyteslml lavage fluid on cells recovered by lavage were considerathe first lavage are indicated by closed circles. Note that bly lower than those of patients with the ordinate is a logarithmic scale. acute farmer's lung. In many cases, neutrophils (11 of 14) and mast cells (6 of the number of these cells recovered from 14) returned to the normal range, but in patients with prior farmer's lung but who only 2 of 14 cases did lymphocytes recontinued farming was less than that turn to normal. The patients who quit from patients with acute farmer's lung, the farm also had fewer lymphocytes, it is noteworthy that considerable over- neutrophils, and mast cells than the palap was observed comparing the two tients who continued farming (table 1). groups (figure 1). As was observed for When plotted as a function oftime since patients with acute farmer's lung, a di- the last episode of acute farmer's lung, rect correlation was observed between the the numbers of lymphocytes and mast number of lymphocytes and mast cells cells (data not shown) recovered by larecovered by lavage from patients with vage did not change significantly. India prior episode of farmer's lung who were viduals who quit the farm were never still farming (figure 2B; r = 0.69, p < evaluated before 32 months after the last 0.001). The variability in the number of episode, as was the case for 6 individuals LAVAGE
who continued farming. Nevertheless, even when the comparison was restricted to individuals evaluated more than 32 months after the last episode of acute farmer's lung, those who quite farming still had significantly fewer lavage lymphocytes and mast cellsthan their counterparts who continued farming. In contrast to the other groups, no significant correlation was observed between the number of lymphocytes and mast cells recovered by lavage.
Normal Farmers Increased numbers of lymphocytes (23 of 36), neutrophils (12 of 36), and mast cells (18 of 36) were recovered by lavage from some normal farmers. Mast cells, however, never exceeded 0.5% of total cells (figure 1). A weak correlation was observed between the number of lymphocytes and mast cells recovered by lavage (figure 2C; r = 0.33, p < 0.05). The two individuals with the greatest numbers of lymphocytes had clearly increased numbers of mast cells;if these two individuals were excluded, no correlation remained. Farmers with increased lavage lymphocytes also were more likely to have increased lavage neutrophils than those with normal lavage lymphocytes (11 of 23 and 1 of 13 individuals, respectively; p < 0.05, Fisher's exact test). Although normal farmers with positive precipitins were more likely to have increased lymphocytes than precipitin-negative farmers, the number of mast cells recovered from the two groups was not significantly different. Six precipitin-positive and nine precipitin-negative normal farmers were studied 2 yr after initial evaluation. No significant differences were observed in the numbers or types of cells recovered at the two studies, and no correlation was observed between the number of mast cells recovered at the first lavage and the changes in physiologic parameters occurring between the two study periods (data not shown). Mast cellcounts obtained from the second observer (M.L.) were acute farmer's lung, 0.9 ± 0.6070; exfarmer's lung-still farming, 0.3 ± 0.3%; exfarmer's lungquit farming, 0.1 ± 0.2%; and normal farmers, 0.1 ± 0.1%. These counts are slightly lower in comparison to the first observer's counts (table 1)in acute farmer's lung and exfarmer's lung-still farming, but similar in exfarmer's lung-quit farming, and normal farmers. Therefore, eventhough there was some interobserver variability, the conclusions of this study would have been similar using mast
BRONCHOALVEOLAR MAST CELLS IN FARMER'S WNG
cell counts obtained from the second observer. Discussion
All patients with acute farmer's lung had large numbers of mast cells on the alveolar surface. This finding fully supports previous studies showing that increased numbers of mast cells can be recovered by lavage from patients with hypersensitivity pneumonitis (18, 19)and is consistent with the idea that the release of mediators by mast cells may playa role in the pathogenesis of this disease. This study does not support the hypothesis that the presence of increased numbers of mast cells on the alveolar surface leads ineluctably to clinical disease,however,nor that the development of alveolitis containing mast cells can be used to identify individuals who are likely to go on to develop hypersensitivity pneumonitis. Similar conclusions have been previously suggested for alveolar lymphocyte count (24). In this regard, many normal farmers have an alveolitis with characteristics similar to those of patients with acute farmer's lung (e.g., increased numbers of lymphocytes, neutrophils, and mast cells). What serves to distinguish the normal farmers from patients with farmer's lung is the intensity of the alveolitis, not its general characteristics. Since very few normal farmers go on to develop farmer's lung (28), the presence of mast cells in the lavage from these individuals is not a poor prognostic sign. Similarly, some individuals with a previous episode of farmer's lung who continued farming had numbers of mast cellson the alveolar surface that werewell within the range observed in acute farmer's lung. Most of these individuals did not develop clinical symptoms, and when considered as a group, pulmonary functions changed little over a 2-yr follow-up period (17).Furthermore, no correlation was observed between changes in physiologic tests and the numbers of mast cells recovered by lavage. Moreover, for individuals with acute farmer's lung and those with a prior episode of acute farmer's lung who continued farming, a direct correlation was observed between the numbers of mast cells and lymphocytes recovered by lavage. Normal farmers with substantial lymphocytic alveolitisalso tended to have increased mast cells. Thus, in the presence of continuing antigen exposure, the intensity of the lymphocytic alveolitiswas correlated with the numbers of mast cells present. It is noteworthy that when indi-
859
viduals were removed from contact with the antigen (the exfarmer's lung subjects who quit farming), a moderately elevated mast cell and lymphocyte counts persisted in some individuals, but no correlation was observed between these two cell types. Some findings in this study suggest that the presence of increased numbers of mast cells on the alveolar surface need not precede, but rather may result from an antigen-driven immune response. First, consistent with the hypothesis that an early inflammatory response preceded the egress of mast cells in the alveolar space, an inversecorrelation wasobserved between the number of mast cells and neutrophils recovered by lavage from patients with farmer's lung. Second, individuals with acute farmer's lung who were lavaged within the first fewdays after the last antigen exposure tended to have fewer mast cells than those lavaged after a longer interval, as previously described (18). These results do not resolve the question of why only some farmers who have evidence of an immune response against thermophilic bacteria go on to develop overt disease but suggest that several factors may be important. First, the followup of patients with a prior episode of farmer's lung was consistent with the possibility that antigen exposure is an important determinant of whether further symptoms develop. If further antigen contact is entirely avoided (those who quit farming), the evidence of lung inflammation decreased significantly, although a modest alveolitis may persist for years. In contrast, the intensity of the alveolitis in patients with a prior episode of farmer's lung who continued farming was more variable. In some individuals the intensity of the alveolitis was similar to that of patients with acute farmer's lung, whereas in other cases the numbers of lymphocytes, neutrophils, and mast cells decreased appreciably. Differences in the degree to which these patients reduced their antigen exposure could explain these findings. In this context, the three individuals who subsequently developed further symptoms continued to have intense alveolitis, and all gave a history of important antigen exposure. Second, it also is possible that differences in the type of immune response that develops are also important in determining the likelihood of developing symptoms. No differences in the types of cells, including mast cells, could be identified in this study, which clearly distinguished symptomatic and asymptomatic individ-
uals. Other factors not evaluated in this study could playa role, however. For example, lavage lymphocytes from symptomatic bird fanciers have been shown to have a significantly increased blastogenic response to phytohemagglutin and pigeon serum compared with those of asymptomatic breeders (5). In conclusion, the development of alveolitis characterized by lymphocytes, mast cells, and neutrophils occurs commonly in farmers but is not associated with symptoms in most individuals. Although this alveolar inflammation probably sets the stage for farmer's lung, additional factors, such as exposure to high levels of antigen or as yet unidentified differences in the nature of the inflammatory response, are required for the development of symptoms. Acknowledgment The writers thank SylvieThibault and Francoise Maher for preparation of the manuscript. References 1. Pepys J. Hypersensitivity diseases of the lung due to fungi and organic dusts. Monogr Allergy 1969; 4:1-145. 2. Schuyler M, Salvaggio J. Hypersensitivity pneumonitis. Semin Respir Med 1984; 5:246-54. 3. Fink IN. Hypersensitivity pneumonitis. J Allergy Clin Immunol 1984; 74:1-9. 4. Pepys J, Jenkins PA. Precipitin test in farmer's lung. Thorax 1965; 20:21-35. 5. Keller RH, Swartz S, Schlueter DP, Bar-Sela S, Fink IN. Immunoregulation in hypersensitivity pneumonitis: phenotypic and functional studies of bronchoalveolar lavage lymphocytes. Am Rev Respir Dis 1984; 130:766-71. 6. Semenzato G, Agostini C, Zambello R, et al. Lung T cells in hypersensitivity pneumonitis: phenotypic and functional analyses. J Immunol 1986; 137:1164-72. 7. Schuyler M, Cook C, Listrom M, FenoglioPreiser C. Blast cells transfer experimental hypersensitivity pneumonitis. Am Rev Respir Dis 1988; 137:1449-55. 8. Fournier E, Tonnel AB, Gosset P, Wallaert B, Ameisen JC, Voisin C. Early neutrophil alveolitis after antigen inhalation in hypersensitivity pneumonitis. Chest 1985; 88:563-6. 9. Cormier Y, Belanger J, Durand P. Factors influencing the development of serum precipitins to farmer's lung antigen in Quebec dairy farmers. Thorax 1985; 40:138-42. 10. Stokes rc, Thrton CWG, Turner-Warwick M. A study of immunoglobulin G subclasses in patients with farmer's lung. Clin Allergy 1981; 11:201-7. 11. Solal-Celigny P, Laviolette M, Hebert J, Cormier Y. Immune reactions in the lungs of asymptomatic dairy farmers. Am Rev Respir Dis 1982; 126:964-7. 12. Cormier Y, Belanger J, Beaudoin J, Laviolette M, Beaudoin R, Hebert J. Abnormal bronchoalveolar lavage in asymptomatic dairy farmers. Study of lymphocytes. Am Rev Respir Dis 1984; 130: 1046-9. 13. Moore VL, Pedersen GM, Hauser WC, Fink IN. A study of lung lavage materials in patients
860 with hypersensitivitypneumonitis: in vitro response to mitogen and antigen in pigeon breeders' disease. J Allergy Clin Immunol 1980; 65:365-70. 14. Leatherman JW, Michael AF, Schwartz BA, Hoidal JR. Lung T cells in hypersensitivity pneumonitis. Ann Intern Med 1984; 100:390-2. 15. Leblanc P, Belanger J, Laviolette M, Cormier Y. Relationship among antigen contact, alveolitis and clinical status in farmer's lung disease. Arch Intern Med 1986; 146:153-7. 16. Cormier Y, Belanger J, Laviolette M. Persistent bronchoalveolar lymphocytosis in asymptomatic farmers. Am Rev Respir Dis 1986; 133:843-7. 17. Cormier Y, Belanger J, Laviolette M. Prognostic significanceof bronchoalveolar lymphocytes in farmer's lung. Am Rev Respir Dis 1987; 135:692-5. 18. Soler P, Nioche S, Valeyre D, et al. Role of mast cells in the pathogenesis of hypersensitivity pneumonitis. Thorax 1987; 42:565-72.
LAVIOLETTE, CORMIER, LOISEAU, SOLER, LEBLANC, AND HANCE
19. Haslam PL, DewarA, Butchers P, Primett ZS, Newman-TaylorA, Turner-Warwick M. Mast cells, atypical lymphocytes,and neutrophils in bronchoalveolar lavage in extrinsic allergic alveolitis. Am Rev Respir Dis 1987; 135:35-47. 20. Kawanami 0, Basset F, Barrios R, Lacronique JG, Ferrans VJ, Crystal RG. Hypersensitivity pneumonitis in man. Light- and electron-microscopic studies of 18 lung biopsies. Am J Pathol 1983; 1l0:277-91. 21. Takizawa H, Ohta K, Hirai K, Misaki Y, Shiga J, Miyamoto T. Importance of mast cells in hypersensitivity pneumonitis (abstract). J Allergy CIin Immunol 1989; 83:283. 22. Bjerrner L, Engstrom-Laurent A, Lundgren R, Rosenhall L, Hallgren R. Bronchoalveolar mastocytosis in farmer's lung is related to the disease activity. Arch Intern Med 1988; 148:1362-5. 23. Cormier Y, Belanger J, Tardif A, Leblanc P, Laviolette M. Relationships between radiographic
change, pulmonary function, and bronchoalveolar lavage fluid lymphocytes in farmer's lung disease. Thorax 1986; 41:28-33. 24. Cormier Y, Belanger J, Leblanc P, Laviolette M. Bronchoalveolar lavage in farmer's lung disease: diagnostic and physiologic significance. Br J Ind Med 1986; 43:401-5. 25. Laviolette M. Lymphocytefluctuation in bronchoalveolar lavage of normal volunteers. Thorax 1985; 40:651-6. 26. Agius RM, Godfrey RC, Holgate ST. Mast cell histamine content of human bronchoalveolar lavage fluid. Thorax 1985; 40:760-7. 27. Zar JH. Biostatistical analysis. 2nd ed. Englewood Cliffs, NJ: Prentice Hall, 1984. 28. Gariepy L, Cormier Y, Laviolette M, Tardif A. Predictive value of bronchoalveolar lavage cells and serum precipitins in asymptomatic dairy farmers. Am Rev Respir Dis 1989; 14:1386-9.
MICHEL LAVIOLETTE, YVON CORMIER, ALAIN LOISEAU, PAUL SOLER, PIERRE LEBLANC, and ALLAN J. HANCE
Introduction
Hypersensitivity pneumonitis is an acute interstitial lung disease resulting from an immune response directedagainst inhaled organic antigens (1-3). These patients are sensitized against the inhaled antigen as manifested by the presence of precipitins and antigen-specific T lymphocytes in the lung and peripheral blood (4-6). Furthermore, in animal models, susceptibility to disease can be transferred with lymphoblasts (7). In sensitized individuals, antigen exposure results in the rapid recruitment of neutrophils and eosinophils into the alveoli; these inflammatory cells likely play an important role in producing the acute alveolar damage and symptoms seen in such patients (8). The mechanisms responsible for the recruitment of inflammatory cells are incompletely understood and could be mediated by antigen-antibody complexes and/or the release of cytokines in the course of the immune response. Although this scenario for the pathogenesis of the disease is attractive, a major paradox exists. Many asymptomatic individuals exposed to antigens causing hypersensitivity pneumonitis also are sensitized to these antigens (9). Furthermore, the levels of precipitins (9), the subclasses of antibodies present (10), the intensity of the lymphocytic alveolitis (11-13), and the phenotype of T lymphocytes present in asymptomatic individuals (614) can overlap those seen in patients with acute hypersensitivity pneumonitis. The lymphocytic alveolitis can persist for 2 yr in normal farmers or asymptomatic patients with previous episodes of hypersensitivity pneumonitis without clinical evidence of progressive disease (15-17). Thus it is unclear why only some individuals develop symptoms. Recently, special attention has been given to the role of mast cells in this dis-
To evaluate the specificity and significance of increased lavage mast cells in farmer's lung, we examined the lavage cell differentials of 89 farmers and 19 normal nonfarming control subiects. The farmers were divided Into four groups: acute farmer's lung (n = 17), farmers with one or more prior episodes of farmer's lung who remained In dally contact with hay (n = 26) or quit farming (n = 14), and normal farmers (n = 36). A total of 14 of the subjects with prior episodes of farmer's lung and still farming and 15 normal farmers were evaluated twice at a 2-yr Interval. The lavage mast cell numbers were significantly higher in acute farmer's lung (7.5 ± 7.3 x 10'/ml, mean ± SO)and ex-farmer's lung who were still farming (1.2 ± 1.3 x 10'/ml) than in normal farmers (0.1 ± 0.1 x 10'/ml) (p < 0.01). A total of 8 of 14 exfarmer's lung patients who had quit farming and 18 of 36 normal farmers had an Increased number of mast cells In lavage, but mast cell count never exceeded 0.5% of total recovered cells. In the acute farmer's lung and ex-farmer's lung-still farming groups, the mast cell count correlated with the lymphocyte count: r 0.83 and r 0.69 (p < 0.001), respectively. In the two groups evaluated twice, mast cell numbers at the first study did not correlate with changes seen at the second study In chest roentgenogram and pUlmonary functions. Weconclude that an Increase in lavage mast cells occurs commonly as a part of the immune response against thermophilic bacteria. However, this increase In the number of lavage mast cells Is found in farmers without clinical signs of farmer's lung, parallels that of lymphocytes, and does not predict the outcome. Consequently the role and clinical significance of an Increased number of lavage mast cells in farmers with or without farmer's lung remains undefined. SUMMARY
=
=
AM REV RESPIR DIS 1991; 144:855-860
ease (18-22). Patients with hypersensitivity pneumonitis have markedly increased numbers of mast cellson the alveolar surface, and the number of mast cells recovered by lavage from patients with hypersensitivity pneumonitis is in general higher than that observed in other interstitiallung diseases. Mast cellscould play an important role in the pathogenesis of the disorder, for example by recruiting inflammatory cells to the lung. Furthermore, it has been suggested that the presence or absence of mast cells could explain why only some sensitized individuals become symptomatic (18). To further explore the importance of mast cells in hypersensitivity pneumonitis, we reviewed lavage and clinical data from a large group of farmers whom we have studied since 1981 (23,24), including both asymptomatic individuals and subjects with recent or prior farmer's lung disease. The goals of the study were to determine if mast cells are always increased in farmer's lung, to evaluate the
possibility that the presence or absence of mast cells correlates with the likelihood of developing symptoms, and to identify factors that contribute to the appearance of mast cellson the alveolar surface of antigen-exposed individuals. Method Study Populations A total of 89 farmers from rural communities surrounding Quebec City and 19normal control subjects were included in the study. Most of these subjects had participated in prior studies, and their clinical characteristics (Received in original form February 13, 1990 and in revised form March 1, 1991) 1 From the Unite de Recherche, Centre de Pneurnologie, Hopital et Universite Laval, Ste-Foy,Quebec, Canada, and INSERM 0.82, Faculte de Medecine Xavier Bichat, Paris, France. 1 Correspondence and requests for reprints should be addressed to Michel Laviolette, M.D., Centre de Pneumologie, H6pital Laval, 2725, Chemin Ste-Foy, Ste-Foy, Quebec GlY 405, Canada.
855
856
LAVIOLETTE, CORMIER, LOISEAU, SOLER, LEBLANC, AND HANCE
have been published elsewhere (15-17,23, 24). Briefly, the patients were divided into the following groups. Acute farmer's lung. Seventeen subjects (14 men and 3 women; 16 nonsmokers and 1 former smoker; age 40 ± 11 yr, mean ± standard deviation [SD]) were diagnosed as having acute farmer's lung by previously described criteria (24). The interval between the onset of acute symptoms and lavage ranged from 1 to 30 days; the individuals were studied at a median of 4 days after the last contact with the farm environment. Prior farmer's lung. Forty subjects diagnosed as having had one or more episodes of acute farmer's lung were lavaged more than 1 yr after the last acute episode. In this group, 4 subjects also were studied during their acute episode and are included in the group of acute farmer's lung. A total of 26 subjects (21 men and 5 women; age 47 ± 11 yr; 21 nonsmokers, 3 former smokers, and 2 smokers) remained in daily contact with hay in the interval between the last acute episode and the lavage. Fourteen individuals in this group underwent a second lavage 2 yr after the original study. Among these 14 patients, 3 individuals were identified who developed subsequent evidence of farmer's lung. A single individual had evidence of an episode of acute farmer's lung between the two studies and subsequently developed other acute episodes; another had clinical and physiologic evidence of progressive disease at the time of the second study, which subsequently improved; and a third developed acute farmer's lung after the second study. All 3 individuals gave a history of important antigen exposure. The other subjects did not demonstrate changes in physical examination, chest roentgenogram, or pulmonary function. In all current farmers lavage was performed during the winter months after the farmers had finished their morning farm chores. Fourteen of the sub-
jects with prior farmer's lung episodes (13 men and 1 woman; age 41 ± 9 yr; 6 nonsmokers, 7 former smokers, and 1 smoker) subsequently quit farming and had no known contact with hay for at least 2 yr. Normalfarmers. A total of 36 subjects were farmers with no past history of farmer's lung (33 men and 3 women; age 41 ± 14yr; 4 former smokers, 31 nonsmokers, and 1 smoker). A total of 18 subjects were precipitin positive for thermophilic Actinomycetes antigen, and 18 subjects were precipitin negative. A total of 15 individuals (6 precipitin positive and 9 precipitin negative) in this group were restudied 2 yr after the initial evaluation. None developed symptoms of farmer's lung during the interim, and chest radiographs remained normal. All normal farmers were lavaged during the winter months after finishing their morning farm chores. Control subjects. A total of 19 nonsmoking normal volunteers (11 men and 8 women; age 27 ± 6 yr) served as controls. None had signs or symptoms of present or past lung disease, and none was known to have exposure to antigens known to cause hypersensitivity lung diseases.
Protocol At the time of evaluation, all subjects underwent clinical examination, pulmonary function testing, chest radiography (except volunteers), and bronchoalveolar lavage, which were performed using previously described techniques (23, 25). All lavages were performed in the right middle lobe using a total volume of 300 ml normal saline solution. The mean volumes of fluid recovered were: acute farmer's lung, 60.8 ± 11.40/0; exfarmer's lung-still farming, 63 ± 9.5%; exfarmer's lung-quit farming, 67.2 ± 7.4%; normal farmers, 67.1± 7.6%, and nonfarming control subjects, 66.1 ± 8.3%. Lavage fluids were centrifuged, and cells
were resuspended in Hanks' balanced salt solution. Cytocentrifuge preparations were made, fixed in methanol, and stained with Wright-Giemsa stain as previously described (11). Cytocentrifuge preparations were coded, and a I,OOO-cell differential count including metachromatic cells was performed by one investigator (P.S.) without knowledge of the clinical characteristics of the subjects. For comparison, slides of farmers' groups also were counted blindly by another observer (M.L.). Because few metachromatic cells with the morphologic characteristics of basophiis were observed, all metachromatic cells were counted as "mast cells" (26). Although some authors have suggested that toluidine blue staining is more sensitive for the detection of mast cells, especially at low mast cell counts (26), this could not be evaluated in our study because unstained cytocentrifuge preparations were not available.
Statistical Analysis Results are expressed as mean ± SD. The distribution of residual error was not normal, as demonstrated by the Koimogorov D statistic using a SAS statistical package (SAS Insititute, Cary, NC). Consequently, for multiple comparisons between groups we used the Kruskal-Wallis test followed by a nonparametric analog Student-Newman-Keuls test. Comparison between two groups was made with Mann-Whitney U test for unpaired data and Wilcoxon's test for paired data. Comparisons of variables between groups were made using Fisher's exact test. For correlations we used the hypothesis test for the Spearman coefficient (27). The significance level was set at 0.05. In all cases the upper limits of normal for lymphocyte, neutrophil, eosinophil, and mast cell lavage numbers were determined by the one-tailed 95% confidence interval (mean ± 1.65 SD).
TABLE 1 NUMBERS AND TYPES OF CELLS RECOVERED BY LAVAGE Group Acute farmer's lung Cells x 103/m!' Percentage
Total Cell Count
Macrophages
Lymphocytes
Neutrophils
Eosinophils
Mast Cells
558.1 ± 256.0 A
126.4 ± 68.5 A 24.1 ± 10.3
349.4 ± 232.2 A 59.0 ± 15.0
66.1 ± 75.0 A 13.9 ± 15.2
8.6 ± 9.7 A 1.6 ± 1.6
7.5 ± 7.3 A 1.3 ± 0.8
312.2 ± 185.4 8
131.5 ± 91.aA 45.8 ± 20.3
168.0 ± 131.3 8 50.3 ± 19.7
10.8 ± 14.78 3.3 ± 3.5
2.0 ± 2.gB 0.6 ± 0.7
1.2 ± 1.3 8 0.4 ± 0.4
186.9 ± 149.5 c
133.7 ± 111.5 A 74.0 ± 16.2
45.9 ± 66.3 c 23.3 ± 16.4
4.5 ± 8.3c 1.8 ± 1.8
2.2 ± 5.9 8c 0.7 ± 1.3
0.2 ± 0.4 c 0.1 ± 0.2
129.8 ± 88.8 c
96.0 ± 67.gA 75.9 ± 16.2
27.9 ± 37.2 c 19.7 ± 14.8
4.7 ± 8.~ 3.3 ± 3.4
1.1 ± 3.1 8c 1.0 ± 3.8
0.1 ± O.l c 0.1 ± 0.1
55.9 ± 16.6 0
49.0 ± 14.3 8 88.0 ± 6.8
5.6 ± 3.7 0 9.7 ± 5.6
1.0 ± 1.3 0 1.7 ± 1.8
0.4 ± 0.1C) 0.6 ± 1.0
0.01 ± 0.03(D} 0.02 ± 0.05
2 Exfarmer's lungstill farming Cells x 103/ml Percentage
3 Exfarmer's lungquit farming Cells x 103/ml Percentage
4 Normal farmers Cells x 103/ml Percentage
5 Nonfarming controls Cells x 10'/ml Percentage
• Values expressed as cellslml were compared using the Kruskal-Wallis test followed by an analog Student-Newman-Keuls test. For each parameter, the values identified by a different lelter are different, p < 0.01 or p < 0.05 (Ielters in brackets).
BRONCHOALVEOLAR MAST CELLS IN FARMER'S WNG
857 A. LYMPHOCYTES
1000 100
......:' :3; a-
•
...•
.,. .,1. a ••• ••
10
A. ACUTE FARMER'S LUNG 40
• •••• • •
•
•
30 0 00
;;
Cbg8>
::1 ·1"" •• •
20
r" 0.1I3 p < 0.001
10
•
0
•
0
o
00
o
0
0
0
.1 Fig. 1. The number of lymphocytes (A), neutrophils (B), and mast cells (C) recovered by lavage from farmers with acute farmer's lung (Group 1),those with prior episodes of farmer's lung who either continued farming (Group 2) or quit farming (Group 3), normal farmers (Group 4), and normal nonfarming control subjects (Group 5). For normal farmers, individuals who were precipitin positive are designated by closed circles (Group 4A) and those who were precipitin negative are shown by open circles (Group 48). The solid line represents the upper limit of the normal range. Note that the ordinate is a logarithmic scale. Dots with number at the bottom of certain columns represent individuals in whom specific cells were not detectable.
0
100 10
•••
o
1000
5
••• ••• .:.
4
•
..-
3
..ir.
,::
•
--
1I00
6
.1
l"l
600
400
B. EX-FARMER'S LUNG, STILL FARMI:-; 200 x 1Q3 Iymphocyteslml lavage fluid on cells recovered by lavage were considerathe first lavage are indicated by closed circles. Note that bly lower than those of patients with the ordinate is a logarithmic scale. acute farmer's lung. In many cases, neutrophils (11 of 14) and mast cells (6 of the number of these cells recovered from 14) returned to the normal range, but in patients with prior farmer's lung but who only 2 of 14 cases did lymphocytes recontinued farming was less than that turn to normal. The patients who quit from patients with acute farmer's lung, the farm also had fewer lymphocytes, it is noteworthy that considerable over- neutrophils, and mast cells than the palap was observed comparing the two tients who continued farming (table 1). groups (figure 1). As was observed for When plotted as a function oftime since patients with acute farmer's lung, a di- the last episode of acute farmer's lung, rect correlation was observed between the the numbers of lymphocytes and mast number of lymphocytes and mast cells cells (data not shown) recovered by larecovered by lavage from patients with vage did not change significantly. India prior episode of farmer's lung who were viduals who quit the farm were never still farming (figure 2B; r = 0.69, p < evaluated before 32 months after the last 0.001). The variability in the number of episode, as was the case for 6 individuals LAVAGE
who continued farming. Nevertheless, even when the comparison was restricted to individuals evaluated more than 32 months after the last episode of acute farmer's lung, those who quite farming still had significantly fewer lavage lymphocytes and mast cellsthan their counterparts who continued farming. In contrast to the other groups, no significant correlation was observed between the number of lymphocytes and mast cells recovered by lavage.
Normal Farmers Increased numbers of lymphocytes (23 of 36), neutrophils (12 of 36), and mast cells (18 of 36) were recovered by lavage from some normal farmers. Mast cells, however, never exceeded 0.5% of total cells (figure 1). A weak correlation was observed between the number of lymphocytes and mast cells recovered by lavage (figure 2C; r = 0.33, p < 0.05). The two individuals with the greatest numbers of lymphocytes had clearly increased numbers of mast cells;if these two individuals were excluded, no correlation remained. Farmers with increased lavage lymphocytes also were more likely to have increased lavage neutrophils than those with normal lavage lymphocytes (11 of 23 and 1 of 13 individuals, respectively; p < 0.05, Fisher's exact test). Although normal farmers with positive precipitins were more likely to have increased lymphocytes than precipitin-negative farmers, the number of mast cells recovered from the two groups was not significantly different. Six precipitin-positive and nine precipitin-negative normal farmers were studied 2 yr after initial evaluation. No significant differences were observed in the numbers or types of cells recovered at the two studies, and no correlation was observed between the number of mast cells recovered at the first lavage and the changes in physiologic parameters occurring between the two study periods (data not shown). Mast cellcounts obtained from the second observer (M.L.) were acute farmer's lung, 0.9 ± 0.6070; exfarmer's lung-still farming, 0.3 ± 0.3%; exfarmer's lungquit farming, 0.1 ± 0.2%; and normal farmers, 0.1 ± 0.1%. These counts are slightly lower in comparison to the first observer's counts (table 1)in acute farmer's lung and exfarmer's lung-still farming, but similar in exfarmer's lung-quit farming, and normal farmers. Therefore, eventhough there was some interobserver variability, the conclusions of this study would have been similar using mast
BRONCHOALVEOLAR MAST CELLS IN FARMER'S WNG
cell counts obtained from the second observer. Discussion
All patients with acute farmer's lung had large numbers of mast cells on the alveolar surface. This finding fully supports previous studies showing that increased numbers of mast cells can be recovered by lavage from patients with hypersensitivity pneumonitis (18, 19)and is consistent with the idea that the release of mediators by mast cells may playa role in the pathogenesis of this disease. This study does not support the hypothesis that the presence of increased numbers of mast cells on the alveolar surface leads ineluctably to clinical disease,however,nor that the development of alveolitis containing mast cells can be used to identify individuals who are likely to go on to develop hypersensitivity pneumonitis. Similar conclusions have been previously suggested for alveolar lymphocyte count (24). In this regard, many normal farmers have an alveolitis with characteristics similar to those of patients with acute farmer's lung (e.g., increased numbers of lymphocytes, neutrophils, and mast cells). What serves to distinguish the normal farmers from patients with farmer's lung is the intensity of the alveolitis, not its general characteristics. Since very few normal farmers go on to develop farmer's lung (28), the presence of mast cells in the lavage from these individuals is not a poor prognostic sign. Similarly, some individuals with a previous episode of farmer's lung who continued farming had numbers of mast cellson the alveolar surface that werewell within the range observed in acute farmer's lung. Most of these individuals did not develop clinical symptoms, and when considered as a group, pulmonary functions changed little over a 2-yr follow-up period (17).Furthermore, no correlation was observed between changes in physiologic tests and the numbers of mast cells recovered by lavage. Moreover, for individuals with acute farmer's lung and those with a prior episode of acute farmer's lung who continued farming, a direct correlation was observed between the numbers of mast cells and lymphocytes recovered by lavage. Normal farmers with substantial lymphocytic alveolitisalso tended to have increased mast cells. Thus, in the presence of continuing antigen exposure, the intensity of the lymphocytic alveolitiswas correlated with the numbers of mast cells present. It is noteworthy that when indi-
859
viduals were removed from contact with the antigen (the exfarmer's lung subjects who quit farming), a moderately elevated mast cell and lymphocyte counts persisted in some individuals, but no correlation was observed between these two cell types. Some findings in this study suggest that the presence of increased numbers of mast cells on the alveolar surface need not precede, but rather may result from an antigen-driven immune response. First, consistent with the hypothesis that an early inflammatory response preceded the egress of mast cells in the alveolar space, an inversecorrelation wasobserved between the number of mast cells and neutrophils recovered by lavage from patients with farmer's lung. Second, individuals with acute farmer's lung who were lavaged within the first fewdays after the last antigen exposure tended to have fewer mast cells than those lavaged after a longer interval, as previously described (18). These results do not resolve the question of why only some farmers who have evidence of an immune response against thermophilic bacteria go on to develop overt disease but suggest that several factors may be important. First, the followup of patients with a prior episode of farmer's lung was consistent with the possibility that antigen exposure is an important determinant of whether further symptoms develop. If further antigen contact is entirely avoided (those who quit farming), the evidence of lung inflammation decreased significantly, although a modest alveolitis may persist for years. In contrast, the intensity of the alveolitis in patients with a prior episode of farmer's lung who continued farming was more variable. In some individuals the intensity of the alveolitis was similar to that of patients with acute farmer's lung, whereas in other cases the numbers of lymphocytes, neutrophils, and mast cells decreased appreciably. Differences in the degree to which these patients reduced their antigen exposure could explain these findings. In this context, the three individuals who subsequently developed further symptoms continued to have intense alveolitis, and all gave a history of important antigen exposure. Second, it also is possible that differences in the type of immune response that develops are also important in determining the likelihood of developing symptoms. No differences in the types of cells, including mast cells, could be identified in this study, which clearly distinguished symptomatic and asymptomatic individ-
uals. Other factors not evaluated in this study could playa role, however. For example, lavage lymphocytes from symptomatic bird fanciers have been shown to have a significantly increased blastogenic response to phytohemagglutin and pigeon serum compared with those of asymptomatic breeders (5). In conclusion, the development of alveolitis characterized by lymphocytes, mast cells, and neutrophils occurs commonly in farmers but is not associated with symptoms in most individuals. Although this alveolar inflammation probably sets the stage for farmer's lung, additional factors, such as exposure to high levels of antigen or as yet unidentified differences in the nature of the inflammatory response, are required for the development of symptoms. Acknowledgment The writers thank SylvieThibault and Francoise Maher for preparation of the manuscript. References 1. Pepys J. Hypersensitivity diseases of the lung due to fungi and organic dusts. Monogr Allergy 1969; 4:1-145. 2. Schuyler M, Salvaggio J. Hypersensitivity pneumonitis. Semin Respir Med 1984; 5:246-54. 3. Fink IN. Hypersensitivity pneumonitis. J Allergy Clin Immunol 1984; 74:1-9. 4. Pepys J, Jenkins PA. Precipitin test in farmer's lung. Thorax 1965; 20:21-35. 5. Keller RH, Swartz S, Schlueter DP, Bar-Sela S, Fink IN. Immunoregulation in hypersensitivity pneumonitis: phenotypic and functional studies of bronchoalveolar lavage lymphocytes. Am Rev Respir Dis 1984; 130:766-71. 6. Semenzato G, Agostini C, Zambello R, et al. Lung T cells in hypersensitivity pneumonitis: phenotypic and functional analyses. J Immunol 1986; 137:1164-72. 7. Schuyler M, Cook C, Listrom M, FenoglioPreiser C. Blast cells transfer experimental hypersensitivity pneumonitis. Am Rev Respir Dis 1988; 137:1449-55. 8. Fournier E, Tonnel AB, Gosset P, Wallaert B, Ameisen JC, Voisin C. Early neutrophil alveolitis after antigen inhalation in hypersensitivity pneumonitis. Chest 1985; 88:563-6. 9. Cormier Y, Belanger J, Durand P. Factors influencing the development of serum precipitins to farmer's lung antigen in Quebec dairy farmers. Thorax 1985; 40:138-42. 10. Stokes rc, Thrton CWG, Turner-Warwick M. A study of immunoglobulin G subclasses in patients with farmer's lung. Clin Allergy 1981; 11:201-7. 11. Solal-Celigny P, Laviolette M, Hebert J, Cormier Y. Immune reactions in the lungs of asymptomatic dairy farmers. Am Rev Respir Dis 1982; 126:964-7. 12. Cormier Y, Belanger J, Beaudoin J, Laviolette M, Beaudoin R, Hebert J. Abnormal bronchoalveolar lavage in asymptomatic dairy farmers. Study of lymphocytes. Am Rev Respir Dis 1984; 130: 1046-9. 13. Moore VL, Pedersen GM, Hauser WC, Fink IN. A study of lung lavage materials in patients
860 with hypersensitivitypneumonitis: in vitro response to mitogen and antigen in pigeon breeders' disease. J Allergy Clin Immunol 1980; 65:365-70. 14. Leatherman JW, Michael AF, Schwartz BA, Hoidal JR. Lung T cells in hypersensitivity pneumonitis. Ann Intern Med 1984; 100:390-2. 15. Leblanc P, Belanger J, Laviolette M, Cormier Y. Relationship among antigen contact, alveolitis and clinical status in farmer's lung disease. Arch Intern Med 1986; 146:153-7. 16. Cormier Y, Belanger J, Laviolette M. Persistent bronchoalveolar lymphocytosis in asymptomatic farmers. Am Rev Respir Dis 1986; 133:843-7. 17. Cormier Y, Belanger J, Laviolette M. Prognostic significanceof bronchoalveolar lymphocytes in farmer's lung. Am Rev Respir Dis 1987; 135:692-5. 18. Soler P, Nioche S, Valeyre D, et al. Role of mast cells in the pathogenesis of hypersensitivity pneumonitis. Thorax 1987; 42:565-72.
LAVIOLETTE, CORMIER, LOISEAU, SOLER, LEBLANC, AND HANCE
19. Haslam PL, DewarA, Butchers P, Primett ZS, Newman-TaylorA, Turner-Warwick M. Mast cells, atypical lymphocytes,and neutrophils in bronchoalveolar lavage in extrinsic allergic alveolitis. Am Rev Respir Dis 1987; 135:35-47. 20. Kawanami 0, Basset F, Barrios R, Lacronique JG, Ferrans VJ, Crystal RG. Hypersensitivity pneumonitis in man. Light- and electron-microscopic studies of 18 lung biopsies. Am J Pathol 1983; 1l0:277-91. 21. Takizawa H, Ohta K, Hirai K, Misaki Y, Shiga J, Miyamoto T. Importance of mast cells in hypersensitivity pneumonitis (abstract). J Allergy CIin Immunol 1989; 83:283. 22. Bjerrner L, Engstrom-Laurent A, Lundgren R, Rosenhall L, Hallgren R. Bronchoalveolar mastocytosis in farmer's lung is related to the disease activity. Arch Intern Med 1988; 148:1362-5. 23. Cormier Y, Belanger J, Tardif A, Leblanc P, Laviolette M. Relationships between radiographic
change, pulmonary function, and bronchoalveolar lavage fluid lymphocytes in farmer's lung disease. Thorax 1986; 41:28-33. 24. Cormier Y, Belanger J, Leblanc P, Laviolette M. Bronchoalveolar lavage in farmer's lung disease: diagnostic and physiologic significance. Br J Ind Med 1986; 43:401-5. 25. Laviolette M. Lymphocytefluctuation in bronchoalveolar lavage of normal volunteers. Thorax 1985; 40:651-6. 26. Agius RM, Godfrey RC, Holgate ST. Mast cell histamine content of human bronchoalveolar lavage fluid. Thorax 1985; 40:760-7. 27. Zar JH. Biostatistical analysis. 2nd ed. Englewood Cliffs, NJ: Prentice Hall, 1984. 28. Gariepy L, Cormier Y, Laviolette M, Tardif A. Predictive value of bronchoalveolar lavage cells and serum precipitins in asymptomatic dairy farmers. Am Rev Respir Dis 1989; 14:1386-9.
