- Email: [email protected]
The Absence of a Relationship between Serum Precipitins and Pulmonary Disease in a Community
The Absenc·e of a Relationship between Serum Precipitins and Pulmonary Disease in a Community* Russell R. Dodge, M.D.; Charles E. Reed, M.D., F.C.C.P.; and Robert A. Barbee, M .D.
To study the role of serum prec:ipitins in resplntory illness in a community, sera obtained from 3,047 residents of Tucson, Ariz, were tested for the presence of predpitating antibody to a battery of antigens. Positive reactions were obtained in 54 subjects (1.8 percent), a lower incidence than has been reported previously. The majority of these subjects were older than 54 yean of age, an age distribution significantly older than the entire sample (P < 0.01). Pulmonary function amo111 the subjects with positive predpitin reactions was not sipificantIy different from that of the asymptomatic nonsmokers of the entire sample. None of the subjects who were life10111 residents of Arizona had llel'Ulll precipitins to any of the thermophllic actinomycetes antigens which were used in the testing. These antigens have been found in 8880Ciation with eminsic allergic alveoHtis, ID08t fre-
quently amoq patients livi8I in the north central United States and were derived from strains of MicrO#Jolyspora faeni, Thennoactinomycetes candid.us and vulgaris. Each subject with precipitins to one or more of the tested antigens was matched by age, sex, and socioeconomic class with two subjects from the sample who had negative prec:ipitin reactions. The groups did not ditler in their prevalence of resplntory symptoms or abnormalities of pulmonary function. We conclude that the presence of precipitating serum antibodies amo111 subjects in a community is not indicative of the presence of immunologic pulmonary disease but merely reftects previons exposure to the tested antigen. In addition, individuals whose sera contain precipitating antibody appear to have no increased tendency to develop other types of pulmonary disease.
pepys1 demonstrated in 1961 that extracts of moldy hay formed precipitins when incubated with the sera of patients with farmer's lung. It has since been well established that these precipitins are antigen-antibody complexes and that the antigens originate with the thermophilic actinomycetes.2•3 Other antigens, derived from such sources as pigeon droppings and fungi, have been found to be responsible for a variety of pulmonary diseases. 4 Positive precipitin tests are also found among persons with no evidence of pulmonary disease who have been exposed to organic dusts containing these antigens.'°7 Positive precipitin reactions have even been found occasionally in persons who have no known history of exposure to antigenic organic dusts. These persons have included office workers, patients at a chest clinic, hospitalized patients, and volunteers. s-u
These latter two findings have cast doubt upon the diagnostic usefulness of the test. Some researchers believe that a positive precipitin test is unrelated to the presence of pulmonary disease and only reflects exposure to the tested antigen. 9 While this conclusion may be justified for the samples studied, the biases inherent in examination of occupational groups and volunteers make extrapolation to the general population hazardous. The purpose of this report is to determine the prevalence of serum preclpitin reactions in a stratified randomly sampled population from a community and to explore the relationship of these reactions to the history of respiratory illness and environmental exposure.
°From the Westend Laboratories, College of Medicine, University of Arizona, Tucson ( Drs. Dodge and Barbee), and the University of Wisconsin Medical Center, Madison (Dr. Reed). Supported by Special Center Research grants HL 15389 and HL 14136 from the National Heart, Lung, and Blood Institute. Manuscript received June 9; revision accepted August 29. Reprint requests: Dr. Dodge, University of Arizona Health Sciences Center, Tucson 85724
608 DODGE, REED, BARBEE
MATERIALS AND METHODS
The design of the Tucson Epidemiologic Study of Obstructive Lung Diseases has been described in detail elsewhere.12 It is a stratified, random clustered sample of nonMexican-American white households in the area of Tucson, Ariz. Approximately 19 percent of those households that were initially approached refused to participate in the full study. The information available on nonparticipating households shows no significant differences from participants with regard to age, social status, geographic location, characteristics pf the household, or the presence or absence of serious illness.
CHEST, 73: 5, MAY, 1978
The population under study includes 3,485 Anglo-white subjects from 1,655 households. Historic data were collected by self-administered questionnaires and interviews. Maximum expiratory flow-volume curves have been obtained on all subjects more than 5 years old. Percent predicted values for the forced expiratory volume in one second ( FEV 1 ) , the forced vital capacity ( FVC), and the instantaneous forced expiratory flow after exhalation of 75 percent of the FVC ( Vmax751) were based on prediction equations derived from asymptomatic nonsmokers in the population. Samples of blood were obtained from 3,047 subjects (94.2 percent of those eligible). Of these, 2,900 samples were tested for the presence of precipitin reactions, the others being lost or damaged. The detection of precipitating antibodies was done by the gel immunodiHusion technique of Flaherty and co-workers.13,14 The techniques have been described in detail elsewhere.13-Ui The antigens were prepared in six pools by combining extracts of three strains of Aspergillus fumigatus (pool 1 ) ; pigeon serum (pool 2); two strains of T oulgaris (pool 3); one strain each of M faeni and T candidus isolated from an air conditioner (pool 4); one strain each of alternaria, pullularia, Penicillium nibrum, and P casei (pool 5) ; and one strain each of Trichoderma, Cephalosporium, and Phoma violacea (pool 6). Data were stored and processed on a computer (CDC 6400). The comparison of age distributions between subjects with a positive precipitin reaction and the entire sample was made by x2 analysis. All precipitin-negative subjects from the entire sample who matched each precipitin-positive subject in regard to age, sex, and socioeconomic status were identified. Two precipitin-negative subjects were then randomly selected and paired with the respective precipitinpositive subject with whom they were matched. Comparisons of various indications of respiratory health between the precipitin-positive and precipitin-negative subjects were then made using x2 analysis and analysis of variance as the methods of statistical analysis. REsULTS
Sera from 54 of the 2,900 tested subjects formed a precipitin reaction to one or more of the tested antigens, an incidence of 1.8 percent. Exactly onehaH of the precipitin-positive subjects were males and haH were females. To aid in analysis of data, the entire sample of subjects was divided into the following three groups according to age: those under 15 years old, those between 15 and 54 years old, and those older than 54 years. Figure 1 demonstrates the distribution of both the entire sample and the precipitin-positive subjects among the age groups. The age distribution of the precipitin-positive subjects was significantly older than the whole sample, with over one-half of the subgroup being over 54 years of age (P < 0.01 ). The results of tests of pulmonary function in the precipitin-positive subjects were compared to those of all nonsmoking asymptomatic persons in the entire sample by expressing the results of the subgroup as percent predicted of the asymptomatic nonsmokers. The mean values (plus or minus standard error) of the FEVi. FVC, and Vmax 7 ~ of the CHEST, 73: 5, MAY, 1978
-·-
I Entire Community Sample 0 Precipitin-Positive Subjects
Cf)
0
Cl>
.c
60
::::J
en 0 40·
52
46.8
-
~
c:
35
Cl>
~
8!.
20·
32.4
20.7
r3
0-14
15-54
>54
Age Distribution
1. Percentage of precipitin-positive subjects and of entire community sample within each age group. Difference is significant for subjects older than 54 years (P < 0.01). FIGURE
precipitin-positive subgroup were 100.2 ± 3.2 percent, 96.9 ± 5.4 percent, and 97.3 + 6.2 percent, respectively. Thus, there was no evidence of decreased pulmonary function among the precipitinpositive subjects. Table I--Compari1on of Smolri,.. Hi1torie1, Re1piratory Symp1om1, and Tei,. of Pulmonary Function•
Data
PrecipitinPrecipitinNegative Positive Subjects Subjects
Total No. of subjects
54
Current smokers
27 (50)
58 (54)
Asymptomatic subjects
28 (52)
63 (58)
4 (7) 1 (2)
12 (11) 4 (4)
4 (7)
2 (2)
7 (13) 3 (6)
11 (10) 8 (7)
7 (13)
4 (4)
0
4 (4)
No. with complaint Cough only Cough and wheeze Cough and shortness of breath Cough, wheeze, and shortness of breath Wheeze only Wheeze and shortness of breath Shortness of breath only FEV1, percent of predicted**t
100.2±3.2
108
94.1±2.3
FVC, percent of predicted**t Vmax1.%, percent of predicted ••t
96.9±5.4 96.6±3.8 97.3±6.2
88.1 ±4.4
*Table values are numbers of subjects (unless otherwise stated); numbers within parentheses are percentages. **Differences between precipitin-positive and precipitin-negative subjects are not significant (P>0.05). tMean ±SE.
SERUM PRECIPITINS AND PULMONARY DISEASE 609
Table 2-Conapariaon
of Trpe of Hona..cooli1111
u.-•
Data Total No. of subjects Central refrigeration Room refrigeration Central evaporative Room evaporative C',ombination
PrecipitinPositive Subjects 54
6 (11)
1 (2) 38 (70) 3 (6) 6 (11)
Unia
PrecipitinNegative Subjects 108 29 (27) 3 (3) 68 (63) 3 (3)
5 (5)
*Differences between precipitin-positive and precipitin-negative subjects are not aignificant (P>0.05). Table values are number of subjects; number within parentheses are percentr ages.
To further determine whether or not differences in respiratory health exist between precipitin-positive and precipitin-negative individuals, each positive subject was matched with two negative subjects from the sample in regard to age, sex, and socioeconomic status. Table 1 demonstrates that there were no significant differences between the precipitin-positive subjects and precipitin-negative subjects in the prevalence of smoking or in the results of tests of pulmonary function ( P > 0.05). The prevalence of specific respiratory symptoms was too small to be subjected to statistical analysis. Differences in the history of exposure might explain why some subjects in the sample developed serum precipitins and others did not. In an attempt to discover differences in histories of exposure between the precipitin-positive and precipitin-negative subjects, comparisons were made of the types of home-cooling system presently used, the current occupation, and the areas of the country where the subjects had resided. Table 3--Conapamon of Oecupadonal Cla.i/ieatio••
Data Total No. of subjects Classification Group 1 (professional) Group 2 (clerical; sales) Group 3 (craftsmen; foremen) Group 4 (operatives) Group 5 (farmers) Group 6 (personal services) Unavailable
Precipitin-PrecipitinPositive Negative Subjects Subjects 54
34 (32)
7 (13)
22 (20)
9 (17) 9 (17) 1 (2)
13 (12) 15 (14) 2 (2)
11 (20)
14 (13) 8 (7)
*Differences between precipitin-positive and precipitin-negative subjects are not significant (P >0.05). Table values are numbers of subjects; numbers within parentheses are percentr ages.
610 DODGE,REED,BARBEE
DlscussION
The first type of extrinsic allergic alveolitis to be described was farmer's lung, which was reported by Campben1s in 1932. By the mid-1960s, it had been shown by Pepys et al, 1 Barbee et al, 8 and others17 that environmental exposure to antigens of the thermophilic actinomycetes was responsible for the clinical syndrome of farmer's lung. Because most patients with farmer's lung demonstrated a precipitin reaction when their sera were incubated with antigen from the actinomycetes, it was postulated that these complexes of antigen-antibody might be responsible for pulmonary injury in these patients. Evidence for this mechanism of injury came from studies of serum sickness in animals. 18 In these experiments, antigen-antibody complexes were trapped beneath endothelial cells where the complexes activated complement; subsequently, polymorphonuclear cells were attracted, lysosomes were Table 4--Conapariaon of Percent of Life Spena in Eaeh Geopaplaie Repon •
108
13 (24)
4 (7)
Table 2 demonstrated that roughly two-thirds of both precipitin-positive and precipitin-negative subjects used central evaporative cooling in their homes. Table 3 reveals no statistically significant differences in occupation between precipitin-positive and precipitin-negative subjects (P > 0.05). Table 4 shows that in the two groups, no significant differences exist in the mean percent of life spent in various regions of the United States ( P > 0.05). Analysis of the information on history of residency revealed that six of the precipitin-positive subjects were lifelong residents of Arizona. None of the six had positive precipitins to any of the thermophilic actinomycetes antigens. These antigens were derived from strains of M faeni, T candidus, and T oulgaris, the organisms reported to be the cause of extrinsic allergic alveolitis most frequently among patients living in the north central United States.
Data Total No. of subjects Mean percent of life spent in: Arizona Far West Rockies Middle West South East Outside United States
Precipitin-PrecipitinPositive Negative Subjects Subjects 54
108
44
41 6 3 27 4 17 2
6 3 28 7 10 2
*Differences between precipitin-positive and precipitin-negative subjects are not significant (P >0.05).
CHEST, 73: 5, MAY, 1978
released, and tissue injury occurred. Further evidence for the importance of the precipitin reaction included the similar time lag of four to eight hours between exposure and injury in both farmer's lung and experimentally-induced serum sickness. 18 The discovery that persons exposed to a wide variety of organic dusts may develop a rather hom0geneous syndrome of acute alveolitis has encouraged the viewpoint that these diseases all have a similar pathogenesis. The finding of precipitins when these patients' sera were incubated with extracts of organic dust seemed to be the common etiologic thread among them and added to the argument that these antigen-antibody complexes were responsible for this type of pulmonary disease. Subsequent reports have raised doubts about the causal importance of the precipitin reaction. Pepys19 has reported that up to 20 percent of his patients with typical farmer's lung do not have positive precipitin reactions. These findings, combined with the discovery that many persons among those exposed to organic dusts have serum precipitins but no signs of pulmonary disease, have led to the conclusion that precipitins reflect exposure but not necessarily respiratory disease among occupational groups. Interest in other, perhaps more important, immunologic mechanisms has focused on cellmediated hypersensitivity. Moore and his co-workers20 have developed an animal model of hypersensitivity-induced pneumonitis in which generation of cellular hypersensitivity to mycobacteria is necessary for lesions to occur in rabbits immunized with pigeon antigens. Precipitating antibodies alone, without concomitant cell-mediated hypersensitivity, were not sufficient to produce the lesions. The results of the present study confirm the conclusion that precipitins reflect exposure and not respiratory disease. We found that 1.8 percent of our sample had a positive precipitin reaction, a much lower incidence than that found among volunteers, office workers, patients at a clinic, and hospitalized patients.s..u This observation may reflect the fact that the organisms most commonly reported as the etiologic agents of allergic alveolitis do not thrive in an arid climate. Exposure to them may be infrequent in the southwestern United States. The absence of precipitin reactions to the thermophilic actinomycetes endemic to the north central United States among the lifelong residents of Arizona strengthens the positive association that exists between exposure and the precipitin test. Since the chances of exposure to these organisms probably increases as a person becomes older, the older age distribution of precipitin-positive subjects is further CHEST, 73: 5, MAY, 1978
evidence that the presence of precipitating antibody reflects exposure to the organic antigen used in the test. When the precipitin-positive subjects were compared to the precipitin-negative control subjects, no diHerences were detected in any of the tested measurements of pulmonary disease or the history of exposure. The prevalence of respiratory symptoms was the same in both groups. Testing of pulmonary function, which included measurements of FEV1, FVC, and Vmax11%, yielded similar results in both groups. H the development of a positive precipitin reaction was associated with changes in pulmonary compliance or resistance in small airways, then decreased flow at low pulmonary volumes might occur without changes in FEV1 or FVC. Our results suggest that this possibility has not occurred in our precipitin-positive subjects. H precipitin-positive individuals diHer from precipitin-negative control subjects only in having been exposed to the tested antigens, the precipitin-positive persons might diHer from the precipitin-negative subjects in their histories of exposure; however, no diHerences were detected in the types of occupations in which the subjects were engaged or in the percentage of lifetime the individuals spent in the various geographic regions of the United States, suggesting that these factors are not important in the development of serum precipitins in the general population. Because of the numerous reports relating alveolitis and precipitin reactions to home heating and cooling units, 9.21.22 we examined the relationships between the subjects' home cooling systems and the prevalence of precipitins and respiratory disease. The southwestern United States is the only area of the country where the evaporative cooler system is widely used. These devices draw warm atmospheric air through water-cooled pads, leading to evaporation of the water and subsequent cooling of the air. The air is then forced indoors to cool the home. These moist conditions can serve as a source of fungal growth and distribution of spores. In a previous study, Dworin23 found almost twice as many spores during a one-year count in a house cooled by this system as in two homes cooled by central air conditioning. We found no increased prevalence of serum precipitins or abnormalities of pulmonary function among the residents of homes with evaporative coolers, suggesting these devices are not responsible for the development of the specific antibodies tested for pulmonary disease among this population. In conclusion, we have examined a group of subjects with positive serum precipitin reactions. These SERUM PRECIPITINS AND PULMONARY DISEASE 611
individuals were part of a general sample of population. While the prevalence of a positive test, the age of the subjects with positive tests, and the pattern of reactions found in subjects who were lifelong residents of Arizona suggest that a positive reaction reflects exposure to the tested antigen, there was no evidence that a positive reaction is an indicator of pulmonary disease.
II 12
ACKNOWLEDGMENT: We thank Ms. Martha G. Cline for her technical assistance.
13
I Pepys J, Riddell RW, Citron KM et al: Precipitins against extracts of hay and fungi in the serum of patients with farmer's lung. Acta Allergol 16:76-77, 1961 2 Pepys J, Jenkins PA, Festenstein GN, et al: Farmer's lung: Thermophilic actinomycetes as a source of "farmer's lung hay" antigen. Lancet 2:607-611, 1963 3 Barbee RA, Dickie HA, Rankin J: Pathogenicity of specific glycopeptide antigen in farmer's lung. Proc Soc Exp Biol Med 118:546-550, 1965 4 Nicholson DP: Extrinsic allergic pneumonias. Am J Med 53:131-136, 1972 5 Wenzel FJ, Gray RL, Roberts RC et al: Serologic studies in farmer's lung. Am Rev Respir Dis 109:464-468, 1974 6 Barbee RA, Callies Q, Dickie HA, et al: The long-term prognosis in farmer's lung. Am Rev Respir Dis 97:223231, 196S 7 Fink N, Sosman AJ, Barboriak JJ, et al: Pigeon breeder's disease: A clinical study of a hypersensitivity pneumonitis. Ann Intern Med 68:1200-1219, 1968 8 Phanuphak P, Salvaggio J, Fink J, et al: Incidence of serum precipitins against organic-dust antigens in different population by counterimmunoelectrophoresis. Chest 68:753-758, 1975 9 do Pico GA, Reddan WG, Chmelik F, et al: The valve of precipitating antibodies in screening for hypersensitivity pneumonitis. Am Rev Respir Dis 113:451-455, 1976 10 Chmelik F, Flaherty DK, Reed CE: Precipitating anti-
14
812 DODGE,REED,BARBEE
15 16 17 18
19 20 21 22 23
bodies in office workers and hospitalized patients direCtea towards antigens causing hypersensitivity pneumonitis. Am Rev Respir Dis 111:201-2<>.5, 1975 Wenzel FS, Gray RL, Roberts RC, et al: Serologic studies in farmer's lung: Precipitins to the thermophilic actinomycetes. Am Rev Respir Dis 108:464-468, 1974 Lebowitz MD, Knudson RJ, Burrows B: Tucson epidemiologic study of obstructive lung diseases: I. Methodology and prevalence of disease. Am J Epidemiol 102:137-152, 1975 Flaherty DK, Barboriak JJ, Emanuel D, et al: Multilaboratory comparison of the three immunodiffusion methods used for the detection of precipitating antibodies in hypersensitivity pneumonitis. J Lab Clin Med 84:298-306, 1974 Flaherty DK, Murray HD, Reed CE: Cross reactions to antigens causing hypersensitivity pneumonitis. J Allergy Clin Immunol 53:329-335, 1974 Fink N, Barboriak JJ, Sosman AJ: Immunologic studies of pigeon breeder's disease. J Allergy 39:214-221, 1967 Campbell JM: Acute symptoms following work with hay. Br Med J 2:1143-1144, 1932 Wenzel FJ, Emanuel DA, Lawton BR, et al: Isolation of the causative agent of farmer's lung. Ann Allergy 22:533540, 1964 Cochrane CG, Dixon FS: Cell and tissue damage through antigen-antibody complexes. In Miescher PA, MullerEberhard HJ (eds) : Textbook of Immunopathology (vol I ) • New York, Grune and Stratton, Inc, 1968, p 94 Pepys J: Hypersensitivity diseases of the lungs due to fungi and organ dusts. Monogr Allergy 4:85, 1969 Moore VL, Hensley GT, Fink JN: An animal model of hypersensitivity pneumonitis in the rabbit. J Clin Invest, 56:937-944, 1975 Banaszak EF, Thiede WH, Fink JN: Hypersensitivity pneumonitis due to contamination of an air-conditioner. N Engl J Med 283:271-276, 1970 Fink JN, Banaszak EA, Thiede WH: lntersitital pneumonitis due to hypersensitivity to an organism contaminating a heating system. Ann Intern Med 74:80-83, 1971 Dworin M: A study of atmospheric mold spores in Tucson, Arizona. Ann Allergy 24:31-36, 1966
~HEST,
73: 5, MAY, 1978
To study the role of serum prec:ipitins in resplntory illness in a community, sera obtained from 3,047 residents of Tucson, Ariz, were tested for the presence of predpitating antibody to a battery of antigens. Positive reactions were obtained in 54 subjects (1.8 percent), a lower incidence than has been reported previously. The majority of these subjects were older than 54 yean of age, an age distribution significantly older than the entire sample (P < 0.01). Pulmonary function amo111 the subjects with positive predpitin reactions was not sipificantIy different from that of the asymptomatic nonsmokers of the entire sample. None of the subjects who were life10111 residents of Arizona had llel'Ulll precipitins to any of the thermophllic actinomycetes antigens which were used in the testing. These antigens have been found in 8880Ciation with eminsic allergic alveoHtis, ID08t fre-
quently amoq patients livi8I in the north central United States and were derived from strains of MicrO#Jolyspora faeni, Thennoactinomycetes candid.us and vulgaris. Each subject with precipitins to one or more of the tested antigens was matched by age, sex, and socioeconomic class with two subjects from the sample who had negative prec:ipitin reactions. The groups did not ditler in their prevalence of resplntory symptoms or abnormalities of pulmonary function. We conclude that the presence of precipitating serum antibodies amo111 subjects in a community is not indicative of the presence of immunologic pulmonary disease but merely reftects previons exposure to the tested antigen. In addition, individuals whose sera contain precipitating antibody appear to have no increased tendency to develop other types of pulmonary disease.
pepys1 demonstrated in 1961 that extracts of moldy hay formed precipitins when incubated with the sera of patients with farmer's lung. It has since been well established that these precipitins are antigen-antibody complexes and that the antigens originate with the thermophilic actinomycetes.2•3 Other antigens, derived from such sources as pigeon droppings and fungi, have been found to be responsible for a variety of pulmonary diseases. 4 Positive precipitin tests are also found among persons with no evidence of pulmonary disease who have been exposed to organic dusts containing these antigens.'°7 Positive precipitin reactions have even been found occasionally in persons who have no known history of exposure to antigenic organic dusts. These persons have included office workers, patients at a chest clinic, hospitalized patients, and volunteers. s-u
These latter two findings have cast doubt upon the diagnostic usefulness of the test. Some researchers believe that a positive precipitin test is unrelated to the presence of pulmonary disease and only reflects exposure to the tested antigen. 9 While this conclusion may be justified for the samples studied, the biases inherent in examination of occupational groups and volunteers make extrapolation to the general population hazardous. The purpose of this report is to determine the prevalence of serum preclpitin reactions in a stratified randomly sampled population from a community and to explore the relationship of these reactions to the history of respiratory illness and environmental exposure.
°From the Westend Laboratories, College of Medicine, University of Arizona, Tucson ( Drs. Dodge and Barbee), and the University of Wisconsin Medical Center, Madison (Dr. Reed). Supported by Special Center Research grants HL 15389 and HL 14136 from the National Heart, Lung, and Blood Institute. Manuscript received June 9; revision accepted August 29. Reprint requests: Dr. Dodge, University of Arizona Health Sciences Center, Tucson 85724
608 DODGE, REED, BARBEE
MATERIALS AND METHODS
The design of the Tucson Epidemiologic Study of Obstructive Lung Diseases has been described in detail elsewhere.12 It is a stratified, random clustered sample of nonMexican-American white households in the area of Tucson, Ariz. Approximately 19 percent of those households that were initially approached refused to participate in the full study. The information available on nonparticipating households shows no significant differences from participants with regard to age, social status, geographic location, characteristics pf the household, or the presence or absence of serious illness.
CHEST, 73: 5, MAY, 1978
The population under study includes 3,485 Anglo-white subjects from 1,655 households. Historic data were collected by self-administered questionnaires and interviews. Maximum expiratory flow-volume curves have been obtained on all subjects more than 5 years old. Percent predicted values for the forced expiratory volume in one second ( FEV 1 ) , the forced vital capacity ( FVC), and the instantaneous forced expiratory flow after exhalation of 75 percent of the FVC ( Vmax751) were based on prediction equations derived from asymptomatic nonsmokers in the population. Samples of blood were obtained from 3,047 subjects (94.2 percent of those eligible). Of these, 2,900 samples were tested for the presence of precipitin reactions, the others being lost or damaged. The detection of precipitating antibodies was done by the gel immunodiHusion technique of Flaherty and co-workers.13,14 The techniques have been described in detail elsewhere.13-Ui The antigens were prepared in six pools by combining extracts of three strains of Aspergillus fumigatus (pool 1 ) ; pigeon serum (pool 2); two strains of T oulgaris (pool 3); one strain each of M faeni and T candidus isolated from an air conditioner (pool 4); one strain each of alternaria, pullularia, Penicillium nibrum, and P casei (pool 5) ; and one strain each of Trichoderma, Cephalosporium, and Phoma violacea (pool 6). Data were stored and processed on a computer (CDC 6400). The comparison of age distributions between subjects with a positive precipitin reaction and the entire sample was made by x2 analysis. All precipitin-negative subjects from the entire sample who matched each precipitin-positive subject in regard to age, sex, and socioeconomic status were identified. Two precipitin-negative subjects were then randomly selected and paired with the respective precipitinpositive subject with whom they were matched. Comparisons of various indications of respiratory health between the precipitin-positive and precipitin-negative subjects were then made using x2 analysis and analysis of variance as the methods of statistical analysis. REsULTS
Sera from 54 of the 2,900 tested subjects formed a precipitin reaction to one or more of the tested antigens, an incidence of 1.8 percent. Exactly onehaH of the precipitin-positive subjects were males and haH were females. To aid in analysis of data, the entire sample of subjects was divided into the following three groups according to age: those under 15 years old, those between 15 and 54 years old, and those older than 54 years. Figure 1 demonstrates the distribution of both the entire sample and the precipitin-positive subjects among the age groups. The age distribution of the precipitin-positive subjects was significantly older than the whole sample, with over one-half of the subgroup being over 54 years of age (P < 0.01 ). The results of tests of pulmonary function in the precipitin-positive subjects were compared to those of all nonsmoking asymptomatic persons in the entire sample by expressing the results of the subgroup as percent predicted of the asymptomatic nonsmokers. The mean values (plus or minus standard error) of the FEVi. FVC, and Vmax 7 ~ of the CHEST, 73: 5, MAY, 1978
-·-
I Entire Community Sample 0 Precipitin-Positive Subjects
Cf)
0
Cl>
.c
60
::::J
en 0 40·
52
46.8
-
~
c:
35
Cl>
~
8!.
20·
32.4
20.7
r3
0-14
15-54
>54
Age Distribution
1. Percentage of precipitin-positive subjects and of entire community sample within each age group. Difference is significant for subjects older than 54 years (P < 0.01). FIGURE
precipitin-positive subgroup were 100.2 ± 3.2 percent, 96.9 ± 5.4 percent, and 97.3 + 6.2 percent, respectively. Thus, there was no evidence of decreased pulmonary function among the precipitinpositive subjects. Table I--Compari1on of Smolri,.. Hi1torie1, Re1piratory Symp1om1, and Tei,. of Pulmonary Function•
Data
PrecipitinPrecipitinNegative Positive Subjects Subjects
Total No. of subjects
54
Current smokers
27 (50)
58 (54)
Asymptomatic subjects
28 (52)
63 (58)
4 (7) 1 (2)
12 (11) 4 (4)
4 (7)
2 (2)
7 (13) 3 (6)
11 (10) 8 (7)
7 (13)
4 (4)
0
4 (4)
No. with complaint Cough only Cough and wheeze Cough and shortness of breath Cough, wheeze, and shortness of breath Wheeze only Wheeze and shortness of breath Shortness of breath only FEV1, percent of predicted**t
100.2±3.2
108
94.1±2.3
FVC, percent of predicted**t Vmax1.%, percent of predicted ••t
96.9±5.4 96.6±3.8 97.3±6.2
88.1 ±4.4
*Table values are numbers of subjects (unless otherwise stated); numbers within parentheses are percentages. **Differences between precipitin-positive and precipitin-negative subjects are not significant (P>0.05). tMean ±SE.
SERUM PRECIPITINS AND PULMONARY DISEASE 609
Table 2-Conapariaon
of Trpe of Hona..cooli1111
u.-•
Data Total No. of subjects Central refrigeration Room refrigeration Central evaporative Room evaporative C',ombination
PrecipitinPositive Subjects 54
6 (11)
1 (2) 38 (70) 3 (6) 6 (11)
Unia
PrecipitinNegative Subjects 108 29 (27) 3 (3) 68 (63) 3 (3)
5 (5)
*Differences between precipitin-positive and precipitin-negative subjects are not aignificant (P>0.05). Table values are number of subjects; number within parentheses are percentr ages.
To further determine whether or not differences in respiratory health exist between precipitin-positive and precipitin-negative individuals, each positive subject was matched with two negative subjects from the sample in regard to age, sex, and socioeconomic status. Table 1 demonstrates that there were no significant differences between the precipitin-positive subjects and precipitin-negative subjects in the prevalence of smoking or in the results of tests of pulmonary function ( P > 0.05). The prevalence of specific respiratory symptoms was too small to be subjected to statistical analysis. Differences in the history of exposure might explain why some subjects in the sample developed serum precipitins and others did not. In an attempt to discover differences in histories of exposure between the precipitin-positive and precipitin-negative subjects, comparisons were made of the types of home-cooling system presently used, the current occupation, and the areas of the country where the subjects had resided. Table 3--Conapamon of Oecupadonal Cla.i/ieatio••
Data Total No. of subjects Classification Group 1 (professional) Group 2 (clerical; sales) Group 3 (craftsmen; foremen) Group 4 (operatives) Group 5 (farmers) Group 6 (personal services) Unavailable
Precipitin-PrecipitinPositive Negative Subjects Subjects 54
34 (32)
7 (13)
22 (20)
9 (17) 9 (17) 1 (2)
13 (12) 15 (14) 2 (2)
11 (20)
14 (13) 8 (7)
*Differences between precipitin-positive and precipitin-negative subjects are not significant (P >0.05). Table values are numbers of subjects; numbers within parentheses are percentr ages.
610 DODGE,REED,BARBEE
DlscussION
The first type of extrinsic allergic alveolitis to be described was farmer's lung, which was reported by Campben1s in 1932. By the mid-1960s, it had been shown by Pepys et al, 1 Barbee et al, 8 and others17 that environmental exposure to antigens of the thermophilic actinomycetes was responsible for the clinical syndrome of farmer's lung. Because most patients with farmer's lung demonstrated a precipitin reaction when their sera were incubated with antigen from the actinomycetes, it was postulated that these complexes of antigen-antibody might be responsible for pulmonary injury in these patients. Evidence for this mechanism of injury came from studies of serum sickness in animals. 18 In these experiments, antigen-antibody complexes were trapped beneath endothelial cells where the complexes activated complement; subsequently, polymorphonuclear cells were attracted, lysosomes were Table 4--Conapariaon of Percent of Life Spena in Eaeh Geopaplaie Repon •
108
13 (24)
4 (7)
Table 2 demonstrated that roughly two-thirds of both precipitin-positive and precipitin-negative subjects used central evaporative cooling in their homes. Table 3 reveals no statistically significant differences in occupation between precipitin-positive and precipitin-negative subjects (P > 0.05). Table 4 shows that in the two groups, no significant differences exist in the mean percent of life spent in various regions of the United States ( P > 0.05). Analysis of the information on history of residency revealed that six of the precipitin-positive subjects were lifelong residents of Arizona. None of the six had positive precipitins to any of the thermophilic actinomycetes antigens. These antigens were derived from strains of M faeni, T candidus, and T oulgaris, the organisms reported to be the cause of extrinsic allergic alveolitis most frequently among patients living in the north central United States.
Data Total No. of subjects Mean percent of life spent in: Arizona Far West Rockies Middle West South East Outside United States
Precipitin-PrecipitinPositive Negative Subjects Subjects 54
108
44
41 6 3 27 4 17 2
6 3 28 7 10 2
*Differences between precipitin-positive and precipitin-negative subjects are not significant (P >0.05).
CHEST, 73: 5, MAY, 1978
released, and tissue injury occurred. Further evidence for the importance of the precipitin reaction included the similar time lag of four to eight hours between exposure and injury in both farmer's lung and experimentally-induced serum sickness. 18 The discovery that persons exposed to a wide variety of organic dusts may develop a rather hom0geneous syndrome of acute alveolitis has encouraged the viewpoint that these diseases all have a similar pathogenesis. The finding of precipitins when these patients' sera were incubated with extracts of organic dust seemed to be the common etiologic thread among them and added to the argument that these antigen-antibody complexes were responsible for this type of pulmonary disease. Subsequent reports have raised doubts about the causal importance of the precipitin reaction. Pepys19 has reported that up to 20 percent of his patients with typical farmer's lung do not have positive precipitin reactions. These findings, combined with the discovery that many persons among those exposed to organic dusts have serum precipitins but no signs of pulmonary disease, have led to the conclusion that precipitins reflect exposure but not necessarily respiratory disease among occupational groups. Interest in other, perhaps more important, immunologic mechanisms has focused on cellmediated hypersensitivity. Moore and his co-workers20 have developed an animal model of hypersensitivity-induced pneumonitis in which generation of cellular hypersensitivity to mycobacteria is necessary for lesions to occur in rabbits immunized with pigeon antigens. Precipitating antibodies alone, without concomitant cell-mediated hypersensitivity, were not sufficient to produce the lesions. The results of the present study confirm the conclusion that precipitins reflect exposure and not respiratory disease. We found that 1.8 percent of our sample had a positive precipitin reaction, a much lower incidence than that found among volunteers, office workers, patients at a clinic, and hospitalized patients.s..u This observation may reflect the fact that the organisms most commonly reported as the etiologic agents of allergic alveolitis do not thrive in an arid climate. Exposure to them may be infrequent in the southwestern United States. The absence of precipitin reactions to the thermophilic actinomycetes endemic to the north central United States among the lifelong residents of Arizona strengthens the positive association that exists between exposure and the precipitin test. Since the chances of exposure to these organisms probably increases as a person becomes older, the older age distribution of precipitin-positive subjects is further CHEST, 73: 5, MAY, 1978
evidence that the presence of precipitating antibody reflects exposure to the organic antigen used in the test. When the precipitin-positive subjects were compared to the precipitin-negative control subjects, no diHerences were detected in any of the tested measurements of pulmonary disease or the history of exposure. The prevalence of respiratory symptoms was the same in both groups. Testing of pulmonary function, which included measurements of FEV1, FVC, and Vmax11%, yielded similar results in both groups. H the development of a positive precipitin reaction was associated with changes in pulmonary compliance or resistance in small airways, then decreased flow at low pulmonary volumes might occur without changes in FEV1 or FVC. Our results suggest that this possibility has not occurred in our precipitin-positive subjects. H precipitin-positive individuals diHer from precipitin-negative control subjects only in having been exposed to the tested antigens, the precipitin-positive persons might diHer from the precipitin-negative subjects in their histories of exposure; however, no diHerences were detected in the types of occupations in which the subjects were engaged or in the percentage of lifetime the individuals spent in the various geographic regions of the United States, suggesting that these factors are not important in the development of serum precipitins in the general population. Because of the numerous reports relating alveolitis and precipitin reactions to home heating and cooling units, 9.21.22 we examined the relationships between the subjects' home cooling systems and the prevalence of precipitins and respiratory disease. The southwestern United States is the only area of the country where the evaporative cooler system is widely used. These devices draw warm atmospheric air through water-cooled pads, leading to evaporation of the water and subsequent cooling of the air. The air is then forced indoors to cool the home. These moist conditions can serve as a source of fungal growth and distribution of spores. In a previous study, Dworin23 found almost twice as many spores during a one-year count in a house cooled by this system as in two homes cooled by central air conditioning. We found no increased prevalence of serum precipitins or abnormalities of pulmonary function among the residents of homes with evaporative coolers, suggesting these devices are not responsible for the development of the specific antibodies tested for pulmonary disease among this population. In conclusion, we have examined a group of subjects with positive serum precipitin reactions. These SERUM PRECIPITINS AND PULMONARY DISEASE 611
individuals were part of a general sample of population. While the prevalence of a positive test, the age of the subjects with positive tests, and the pattern of reactions found in subjects who were lifelong residents of Arizona suggest that a positive reaction reflects exposure to the tested antigen, there was no evidence that a positive reaction is an indicator of pulmonary disease.
II 12
ACKNOWLEDGMENT: We thank Ms. Martha G. Cline for her technical assistance.
13
I Pepys J, Riddell RW, Citron KM et al: Precipitins against extracts of hay and fungi in the serum of patients with farmer's lung. Acta Allergol 16:76-77, 1961 2 Pepys J, Jenkins PA, Festenstein GN, et al: Farmer's lung: Thermophilic actinomycetes as a source of "farmer's lung hay" antigen. Lancet 2:607-611, 1963 3 Barbee RA, Dickie HA, Rankin J: Pathogenicity of specific glycopeptide antigen in farmer's lung. Proc Soc Exp Biol Med 118:546-550, 1965 4 Nicholson DP: Extrinsic allergic pneumonias. Am J Med 53:131-136, 1972 5 Wenzel FJ, Gray RL, Roberts RC et al: Serologic studies in farmer's lung. Am Rev Respir Dis 109:464-468, 1974 6 Barbee RA, Callies Q, Dickie HA, et al: The long-term prognosis in farmer's lung. Am Rev Respir Dis 97:223231, 196S 7 Fink N, Sosman AJ, Barboriak JJ, et al: Pigeon breeder's disease: A clinical study of a hypersensitivity pneumonitis. Ann Intern Med 68:1200-1219, 1968 8 Phanuphak P, Salvaggio J, Fink J, et al: Incidence of serum precipitins against organic-dust antigens in different population by counterimmunoelectrophoresis. Chest 68:753-758, 1975 9 do Pico GA, Reddan WG, Chmelik F, et al: The valve of precipitating antibodies in screening for hypersensitivity pneumonitis. Am Rev Respir Dis 113:451-455, 1976 10 Chmelik F, Flaherty DK, Reed CE: Precipitating anti-
14
812 DODGE,REED,BARBEE
15 16 17 18
19 20 21 22 23
bodies in office workers and hospitalized patients direCtea towards antigens causing hypersensitivity pneumonitis. Am Rev Respir Dis 111:201-2<>.5, 1975 Wenzel FS, Gray RL, Roberts RC, et al: Serologic studies in farmer's lung: Precipitins to the thermophilic actinomycetes. Am Rev Respir Dis 108:464-468, 1974 Lebowitz MD, Knudson RJ, Burrows B: Tucson epidemiologic study of obstructive lung diseases: I. Methodology and prevalence of disease. Am J Epidemiol 102:137-152, 1975 Flaherty DK, Barboriak JJ, Emanuel D, et al: Multilaboratory comparison of the three immunodiffusion methods used for the detection of precipitating antibodies in hypersensitivity pneumonitis. J Lab Clin Med 84:298-306, 1974 Flaherty DK, Murray HD, Reed CE: Cross reactions to antigens causing hypersensitivity pneumonitis. J Allergy Clin Immunol 53:329-335, 1974 Fink N, Barboriak JJ, Sosman AJ: Immunologic studies of pigeon breeder's disease. J Allergy 39:214-221, 1967 Campbell JM: Acute symptoms following work with hay. Br Med J 2:1143-1144, 1932 Wenzel FJ, Emanuel DA, Lawton BR, et al: Isolation of the causative agent of farmer's lung. Ann Allergy 22:533540, 1964 Cochrane CG, Dixon FS: Cell and tissue damage through antigen-antibody complexes. In Miescher PA, MullerEberhard HJ (eds) : Textbook of Immunopathology (vol I ) • New York, Grune and Stratton, Inc, 1968, p 94 Pepys J: Hypersensitivity diseases of the lungs due to fungi and organ dusts. Monogr Allergy 4:85, 1969 Moore VL, Hensley GT, Fink JN: An animal model of hypersensitivity pneumonitis in the rabbit. J Clin Invest, 56:937-944, 1975 Banaszak EF, Thiede WH, Fink JN: Hypersensitivity pneumonitis due to contamination of an air-conditioner. N Engl J Med 283:271-276, 1970 Fink JN, Banaszak EA, Thiede WH: lntersitital pneumonitis due to hypersensitivity to an organism contaminating a heating system. Ann Intern Med 74:80-83, 1971 Dworin M: A study of atmospheric mold spores in Tucson, Arizona. Ann Allergy 24:31-36, 1966
~HEST,
73: 5, MAY, 1978