Elevated Bronchoalveolar Lavage Fluid Histamine Levels in Allergic Asthmatics Are Associated with Increased Airway Obstruction 1-3
NIZAR N. JARJOUR, WILLIAM J. CALHOUN, LAWRENCE B. SCHWARTZ, and WILLIAM W. BUSSE
Introduction Bronchial asthma is characterized by airway hyperresponsiveness. Although the precise mechanisms by which airway hyperresponsiveness develops in asthma have yet to be established, a number of investigators have proposed that airway inflammation is a factor. By use ofbronchoalveolar lavage (BAL), deMonchyand coworkers (1)and Metzger and colleagues (2) have shown that a significant cellular influx occurs during the late asthmatic reaction to antigen challenge, an event usually associated with airway hyperresponsiveness. Moreover, other investigators (3) have used bronchial biopsy to show that airway inflammation is present in asthma and that it may relate to the level of hyperresponsiveness. Bronchial biopsy has some advantages compared with BAL in assessingcellular infiltration associated with chronic inflammation. However, mediator concentrations, and possibly their cellular sources, can be determined using BAL techniques. For example, mast cells release both histamine and tryptase, whereas basophils release only histamine. This differential pattern of mediator release can give insight into the source of BAL histamine in allergic asthma, as local airway allergen challenge has been associated with parallel increases in both histamine and tryptase, suggesting in vivo activation of pulmonary mast cells (4). The purpose of this study was to establish the levels of histamine and tryptase in the BAL fluid of allergic subjects, to compare them with healthy control fluids, and to determine the relationships among those mediators, airway function, and airway responsiveness. Methods Study Subjects All subjects gave informed consent to a protocol approved by the University of Wiscon-
SUMMARY Bronchoalveolar lavage (SAL) fluid was evaluated for histamine and tryptaae levels In 61 samples (46 samples from 24 atopic asthmatics, seven samples from seven patients with allergic rhinitis, and eight samples from eight normal volunteers). Asthmatics and patients with allergic rhinitis had significantly higher BAL histamine than did normal subjects (169 ± 22, 141 ± 23, 42 ± 6 pg/ml, respectively; p < 0.05, both comparisons). SAL fluid tryptaae levels were also higher In asthmatics and patients with allergic rhinitis than In normal subjects (0.36 ± 0.03, 0.38 ± 0.05, 0.23 ± 0.04 nglml, respectively; p < 0.05, both comparisons); however, levels of tryptaae and histamine In SAL were not correlated (r 0.03 In the group as a whole, r 0.12 In the asthmatic group). SAL concentration of histamine correlated Inversely with FEY1 percent predicted In the asthmatic group (r 0.44, P < 0.005).Asthmatlcs with high SAL fluid histamine (~ 100 pg/ml, n 23) had lower FEY1 percent predicted (80 ± 3% versus 98 ± 3%, P 0.0005), lower FEY1/FYC ratio (72 ± 1% versus 77 ± 2%, P < 0.05), higher percentage of SAL eoslnophlls (2.2 ± 0.4% versus 0.6 ± 0.1%, P < 0.002), and greater airway responslvene88 (lower PD20 (13.1 ± 3.4 versus 41.5 ± 13.7 cumulative breath units, p < 0.05]) compared with asthmatics with low SAL fluid histamine « 100 pg/ml, n 23). These data Indicate that SAL histamine concentrations are correlated with airway obstruction In asthma and may serve as an Index of severity. The lack of correlation between SAL tryptase and histamine may Indicate that pulmonary basophlls, rather than mast cells, could be the source for histamine In these Patients. ' AM REV RESPIR DIS 1991; 144:83-87
=-
=-
=-
=
=
=
sin Human Subjects Committee. Thirty-nine subjects (16 women) between the ages of 18 and 51yr were recruited. These volunteers included seven subjects with allergic rhinitis (AR) (3 women), 24 allergic asthmatics (AA) (10women), and eight nonatopic healthy normal volunteers (NY) (3 women). All werenonsmokers and otherwise healthy. The 24 subjects with asthma underwent a total of 46 BAL procedures because some subjects participated in more than one study, and some studies required more than one BAL in each subject. However, in this report we included only results of BAL analyses performed during a period of quiescence and with no preceding challenge or intervention. When more than one BAL was performed in the same person, the two BAL specimens were obtained from different segments, and the two procedures were separated by a minimum of 24 h. This time separation has been shown to avoid artifactual effects of a prior BAL in subjects with mild asthma (5). Asthmatics wereexcluded if they were receiving topical or systemic steroids or cromolyn. The allergic state was defined as a positive prick skin test to one or more common aeroallergens:extract of ragweed, grass, trees, house dust mite, alternaria, cat, and dog (Greer Laboratories, Lenoir, NC).
Pulmonary Functions and Bronchoprovocation Spirometry was performed using a portable spirometer meeting American Thoracic Society criteria, with the best of several FVC maneuvers used to obtain a FEV 1 and FVC (6). The FEV 1 was analyzed and reported as percent predicted. Methacholine and histamine bronchoprovocations were performed by a graded nebulization technique using a DeVilbissnebulizer (DeVilbissCo., Somerset, PA) and a Rosenthal-French dosimeter (Lab-
(Received in original form October 25, 1990 and in revised form March 1, 1991) 1 From the Sections of Pulmonary/Critical Care and Allergy/Clinical Immunology, Department of Medicine, School of Medicine, University of Wisconsin, Madison, Wisconsin, and the Department of Medicine, Medical College of Virginia, Richmond, Virginia. 2 Supported in part by grants from the Department of Medicine, University of Wisconsin, by Grants K08 01828, AI 26609, and UW GCRC RR 03186from the National Institutes of Health, and by the American Lung Association. 3 Correspondence and requests for reprints should be addressed to Nizar N. Jarjour, M.D., 600 Highland Ave. (H6/380), Madison, WI 53792.
83
84
oratory for Applied Immunology, Baltimore, MD). After an initial spirometry, subjects were given five breaths of diluent, and spirometry was rechecked 5 min later. Five breaths of increasing concentration of agonist were given, and spirometry was rechecked 5 min after each dose until the FEV 1 fell more than 20010 from the postdiluent value. The provocation dose giving a 20010 drop in FEV 1 (PD l o ) was calculated by a log-linear interpolation using a commercially available analysis package (PD20, Madison Scientific Software, Madison, WI), and expressed in cumulative breath units (CBU) (1 breath unit = 1 inhalation of 1 mg/ml solution of histamine or methacholine). All challenges were done at least 2 h after bronchoscopy and BAL. Because methacholine and histamine challenges when done by similar technique gave similar results, we combined the results from both challenges (7, 8).
Bronchoscopy and Bronchoalveolar Lovage Bronchoscopy and BAL were performed as previously reported (5). Spirometry was performed in each subject immediately before bronchoscopy and BAL. With one exception, the FEV 1 was ~ 60010 predicted, in accordance with guidelines for the investigative use of BAL in asthmatic volunteers (9, 10). In that single individual, whose FEV 1 was 46010 predicted, BAL was done uneventfully. All subjects received parenteral atropine sulfate (0.6 mg) and midazolam (1 to 1.5 mg) as premedication. Asthmatic subjects wereasked to stop their p-agonist inhalers for 8 h, and theophylline for 24 h prior to BAL. 1\vopuffs of an albuterol metered-dose inhaler wereadministered immediately prior to bronchoscopy. Bronchoscopy was accomplished using an Olympus fiberoptic bronchoscope (Olympus Corp., New Hyde Park, NY) that was wedged into a subsegment of the right middle lobe (RML) or lingula. BAL was performed using four aliquots (60-mleach) of sterile, nonpyrogenic 0.9010 NaCI, warmed to body temperature. Each aliquot was instilled and immediately recovered by hand suction. BAL Fluid Processing The lavage return was pooled, and BAL cells were recovered by centrifugation. Supernatant was frozen at -70 0 C. The cell pellet was washed once in phosphate-buffered saline and resuspended in Hanks' balanced salt solution (HBSS). Cells were counted using a hemocytometer and were brought to a final dilution of 2 x lQ6 cells/ml for eytofuge preparation and cell functional assays. Cytofuge slides were air-dried, fixed in methanol, and stained (Diff-Quik; Scientific Products, Chicago, IL). Differential cellcounts wereperformed by identifying ~ 300 cells as lymphocytes, neutrophils, eosinophils, or alveolar macrophages. HAL Total Protein and Mediators Assays BAL fluid protein was measured by the Lowry
JARJOUR, CALHOUN, SCHWARTZ, AND BUSSE
assay (11) modified for microtiter equipment as described (12). BAL fluid histamine was measured with a singleisotope radioenzymatic assay using purified rat kidney histamine N-methyltransferase (13). In our laboratory this method has a linear coefficient of correlation of 0.99, with a lower limit of sensitivity of 30 pg/mI. BAL histamine levels lower than 30 pg/ml were analyzed in this report as 30 pg/ml. BAL fluid tryptase was measured in whole BAL fluid by a sandwich enzyme-linked immunosorbent assay (ELISA) method based on a modification (14) of a previously described procedure (4, 15). The murine monoclonal antibody, 05, for capture, and the bionylated murine antibody, 04, for detection, were used (14). The assay lower limit of detection is 0.05 ng of tryptase/40 tJ,1 sample. Samples of lavage fluid were concentrated to 20-fold by ultrafiltration using a Centricon10 chamber (Amicon Corp., Danvers, MA).
Data Analysis Data were first analyzed in an electronic spreadsheet (SuperCalc 4; Computer Associates, San Jose, CA). Subsequent analysis was done using a microcomputer statistics package (Abstat 6.05;Anderson-Bell, Parker, CO). Initial comparisons betweenAR, AA, and NY groups were made by performing a nonparametric analysis of variance (Kruskal-Wallis test: KW). Only when a significant KW was obtained were further statistical comparisons between the groups performed to determine which group(s) differed. Close comparisons were made using a Mann-Whitney test (MW). Correlations were decided by Pearson's test. A p value < 0.05 was considered significant (16).All data wereexpressedas mean ± SEM. Results
Physiologic Findings Allergic asthmatics demonstrated a low-
er FEV l (88 ± 2010, p < 0.05 versus AR or NV) than did either AR (112 ± 3) or NY (106 ± 4) subjects (p < 0.01, KW for effect of subject group on FEY 1). Moreover, the ratio of FEV 1 to FVC in asthmatics was significantly less than in NV subjects (0.75 ± 0.01 versus 0.87 ± 0.03, p 100 pg/ml) have demonstrated
86
JAAJOUR, CALHOUN, SCHWARTZ, AND BUSSE
TABLE 2 COMPARISON OF ASTHMATICS WITH HIGH AND LOW BAL HISTAMINE* Histamine
Histamine 100 pglml
Parameter
< 100 pglml
~
Samples, n Histamine, pglml Tryptase, nglml FEV1 , % pred FEV1/FVC PD2o, cumulative breath units Volume return, ml SAL total protein, ~glml Total cells, 10' AM, % Eos, % Neu, % Lym, %
23 48 :t 4 0.39:t 0.05 96 :t 3 rr e 2 41.5 ± 13.7
23 291 ::t 26 0.33 :t: 0.03
180 ± 3 68 ± 3 15.5 ± 1.8 93.0 ± 1.0
0.6::t 0.1 0.7::t 0.1 5.4::t 0.9
8O:t: 71 :t: 13.1 :t: 149 :t: 8O:t:
3
1 3.4 7 6
12.3 ::t 1.9 90.1 :t: 1.4 2.2 :t 0.4 2.1 :t 0.7 5.6 :t 1.1
p Valuet
0.ססOO
NS* 0.0003 0.04 0.0455 0.0002
NS NS
NS 0.0011 0.0488
NS
For definition of abbreviations, see table 1. • Data are means ± SEM. t p values are based on independent t test. Not significant; p > 0.05 confirmed by Mann-Whitney test.
*
greater airway hyperresponsiveness to methacholine than have those with lower levels « 100 pg/ml) of this mediator (17). These data, coupled with observations from the present study, suggest that BAL histamine concentrations may be an indicator of asthma severity among subjects with stable asthma. However, Casale and coworkers (17) did not identify a difference in FEV 1 volume between the two groups. Possible explanations for the lack of correlation between airflow limitation and BAL histamine concentrations in that study may be due to differences in the severity of asthma, sample size, or their use of actual FEV 1 values for analysis as opposed to FEV 1 percent predicted. The use of actual FEV 1 values can be misleading in small studies because of differences in subjects' predicted values.
BAL Histamine - Tryptase Correlation Allergic Asthmatics 1.2
-e
NIZAR N. JARJOUR, WILLIAM J. CALHOUN, LAWRENCE B. SCHWARTZ, and WILLIAM W. BUSSE
Introduction Bronchial asthma is characterized by airway hyperresponsiveness. Although the precise mechanisms by which airway hyperresponsiveness develops in asthma have yet to be established, a number of investigators have proposed that airway inflammation is a factor. By use ofbronchoalveolar lavage (BAL), deMonchyand coworkers (1)and Metzger and colleagues (2) have shown that a significant cellular influx occurs during the late asthmatic reaction to antigen challenge, an event usually associated with airway hyperresponsiveness. Moreover, other investigators (3) have used bronchial biopsy to show that airway inflammation is present in asthma and that it may relate to the level of hyperresponsiveness. Bronchial biopsy has some advantages compared with BAL in assessingcellular infiltration associated with chronic inflammation. However, mediator concentrations, and possibly their cellular sources, can be determined using BAL techniques. For example, mast cells release both histamine and tryptase, whereas basophils release only histamine. This differential pattern of mediator release can give insight into the source of BAL histamine in allergic asthma, as local airway allergen challenge has been associated with parallel increases in both histamine and tryptase, suggesting in vivo activation of pulmonary mast cells (4). The purpose of this study was to establish the levels of histamine and tryptase in the BAL fluid of allergic subjects, to compare them with healthy control fluids, and to determine the relationships among those mediators, airway function, and airway responsiveness. Methods Study Subjects All subjects gave informed consent to a protocol approved by the University of Wiscon-
SUMMARY Bronchoalveolar lavage (SAL) fluid was evaluated for histamine and tryptaae levels In 61 samples (46 samples from 24 atopic asthmatics, seven samples from seven patients with allergic rhinitis, and eight samples from eight normal volunteers). Asthmatics and patients with allergic rhinitis had significantly higher BAL histamine than did normal subjects (169 ± 22, 141 ± 23, 42 ± 6 pg/ml, respectively; p < 0.05, both comparisons). SAL fluid tryptaae levels were also higher In asthmatics and patients with allergic rhinitis than In normal subjects (0.36 ± 0.03, 0.38 ± 0.05, 0.23 ± 0.04 nglml, respectively; p < 0.05, both comparisons); however, levels of tryptaae and histamine In SAL were not correlated (r 0.03 In the group as a whole, r 0.12 In the asthmatic group). SAL concentration of histamine correlated Inversely with FEY1 percent predicted In the asthmatic group (r 0.44, P < 0.005).Asthmatlcs with high SAL fluid histamine (~ 100 pg/ml, n 23) had lower FEY1 percent predicted (80 ± 3% versus 98 ± 3%, P 0.0005), lower FEY1/FYC ratio (72 ± 1% versus 77 ± 2%, P < 0.05), higher percentage of SAL eoslnophlls (2.2 ± 0.4% versus 0.6 ± 0.1%, P < 0.002), and greater airway responslvene88 (lower PD20 (13.1 ± 3.4 versus 41.5 ± 13.7 cumulative breath units, p < 0.05]) compared with asthmatics with low SAL fluid histamine « 100 pg/ml, n 23). These data Indicate that SAL histamine concentrations are correlated with airway obstruction In asthma and may serve as an Index of severity. The lack of correlation between SAL tryptase and histamine may Indicate that pulmonary basophlls, rather than mast cells, could be the source for histamine In these Patients. ' AM REV RESPIR DIS 1991; 144:83-87
=-
=-
=-
=
=
=
sin Human Subjects Committee. Thirty-nine subjects (16 women) between the ages of 18 and 51yr were recruited. These volunteers included seven subjects with allergic rhinitis (AR) (3 women), 24 allergic asthmatics (AA) (10women), and eight nonatopic healthy normal volunteers (NY) (3 women). All werenonsmokers and otherwise healthy. The 24 subjects with asthma underwent a total of 46 BAL procedures because some subjects participated in more than one study, and some studies required more than one BAL in each subject. However, in this report we included only results of BAL analyses performed during a period of quiescence and with no preceding challenge or intervention. When more than one BAL was performed in the same person, the two BAL specimens were obtained from different segments, and the two procedures were separated by a minimum of 24 h. This time separation has been shown to avoid artifactual effects of a prior BAL in subjects with mild asthma (5). Asthmatics wereexcluded if they were receiving topical or systemic steroids or cromolyn. The allergic state was defined as a positive prick skin test to one or more common aeroallergens:extract of ragweed, grass, trees, house dust mite, alternaria, cat, and dog (Greer Laboratories, Lenoir, NC).
Pulmonary Functions and Bronchoprovocation Spirometry was performed using a portable spirometer meeting American Thoracic Society criteria, with the best of several FVC maneuvers used to obtain a FEV 1 and FVC (6). The FEV 1 was analyzed and reported as percent predicted. Methacholine and histamine bronchoprovocations were performed by a graded nebulization technique using a DeVilbissnebulizer (DeVilbissCo., Somerset, PA) and a Rosenthal-French dosimeter (Lab-
(Received in original form October 25, 1990 and in revised form March 1, 1991) 1 From the Sections of Pulmonary/Critical Care and Allergy/Clinical Immunology, Department of Medicine, School of Medicine, University of Wisconsin, Madison, Wisconsin, and the Department of Medicine, Medical College of Virginia, Richmond, Virginia. 2 Supported in part by grants from the Department of Medicine, University of Wisconsin, by Grants K08 01828, AI 26609, and UW GCRC RR 03186from the National Institutes of Health, and by the American Lung Association. 3 Correspondence and requests for reprints should be addressed to Nizar N. Jarjour, M.D., 600 Highland Ave. (H6/380), Madison, WI 53792.
83
84
oratory for Applied Immunology, Baltimore, MD). After an initial spirometry, subjects were given five breaths of diluent, and spirometry was rechecked 5 min later. Five breaths of increasing concentration of agonist were given, and spirometry was rechecked 5 min after each dose until the FEV 1 fell more than 20010 from the postdiluent value. The provocation dose giving a 20010 drop in FEV 1 (PD l o ) was calculated by a log-linear interpolation using a commercially available analysis package (PD20, Madison Scientific Software, Madison, WI), and expressed in cumulative breath units (CBU) (1 breath unit = 1 inhalation of 1 mg/ml solution of histamine or methacholine). All challenges were done at least 2 h after bronchoscopy and BAL. Because methacholine and histamine challenges when done by similar technique gave similar results, we combined the results from both challenges (7, 8).
Bronchoscopy and Bronchoalveolar Lovage Bronchoscopy and BAL were performed as previously reported (5). Spirometry was performed in each subject immediately before bronchoscopy and BAL. With one exception, the FEV 1 was ~ 60010 predicted, in accordance with guidelines for the investigative use of BAL in asthmatic volunteers (9, 10). In that single individual, whose FEV 1 was 46010 predicted, BAL was done uneventfully. All subjects received parenteral atropine sulfate (0.6 mg) and midazolam (1 to 1.5 mg) as premedication. Asthmatic subjects wereasked to stop their p-agonist inhalers for 8 h, and theophylline for 24 h prior to BAL. 1\vopuffs of an albuterol metered-dose inhaler wereadministered immediately prior to bronchoscopy. Bronchoscopy was accomplished using an Olympus fiberoptic bronchoscope (Olympus Corp., New Hyde Park, NY) that was wedged into a subsegment of the right middle lobe (RML) or lingula. BAL was performed using four aliquots (60-mleach) of sterile, nonpyrogenic 0.9010 NaCI, warmed to body temperature. Each aliquot was instilled and immediately recovered by hand suction. BAL Fluid Processing The lavage return was pooled, and BAL cells were recovered by centrifugation. Supernatant was frozen at -70 0 C. The cell pellet was washed once in phosphate-buffered saline and resuspended in Hanks' balanced salt solution (HBSS). Cells were counted using a hemocytometer and were brought to a final dilution of 2 x lQ6 cells/ml for eytofuge preparation and cell functional assays. Cytofuge slides were air-dried, fixed in methanol, and stained (Diff-Quik; Scientific Products, Chicago, IL). Differential cellcounts wereperformed by identifying ~ 300 cells as lymphocytes, neutrophils, eosinophils, or alveolar macrophages. HAL Total Protein and Mediators Assays BAL fluid protein was measured by the Lowry
JARJOUR, CALHOUN, SCHWARTZ, AND BUSSE
assay (11) modified for microtiter equipment as described (12). BAL fluid histamine was measured with a singleisotope radioenzymatic assay using purified rat kidney histamine N-methyltransferase (13). In our laboratory this method has a linear coefficient of correlation of 0.99, with a lower limit of sensitivity of 30 pg/mI. BAL histamine levels lower than 30 pg/ml were analyzed in this report as 30 pg/ml. BAL fluid tryptase was measured in whole BAL fluid by a sandwich enzyme-linked immunosorbent assay (ELISA) method based on a modification (14) of a previously described procedure (4, 15). The murine monoclonal antibody, 05, for capture, and the bionylated murine antibody, 04, for detection, were used (14). The assay lower limit of detection is 0.05 ng of tryptase/40 tJ,1 sample. Samples of lavage fluid were concentrated to 20-fold by ultrafiltration using a Centricon10 chamber (Amicon Corp., Danvers, MA).
Data Analysis Data were first analyzed in an electronic spreadsheet (SuperCalc 4; Computer Associates, San Jose, CA). Subsequent analysis was done using a microcomputer statistics package (Abstat 6.05;Anderson-Bell, Parker, CO). Initial comparisons betweenAR, AA, and NY groups were made by performing a nonparametric analysis of variance (Kruskal-Wallis test: KW). Only when a significant KW was obtained were further statistical comparisons between the groups performed to determine which group(s) differed. Close comparisons were made using a Mann-Whitney test (MW). Correlations were decided by Pearson's test. A p value < 0.05 was considered significant (16).All data wereexpressedas mean ± SEM. Results
Physiologic Findings Allergic asthmatics demonstrated a low-
er FEV l (88 ± 2010, p < 0.05 versus AR or NV) than did either AR (112 ± 3) or NY (106 ± 4) subjects (p < 0.01, KW for effect of subject group on FEY 1). Moreover, the ratio of FEV 1 to FVC in asthmatics was significantly less than in NV subjects (0.75 ± 0.01 versus 0.87 ± 0.03, p 100 pg/ml) have demonstrated
86
JAAJOUR, CALHOUN, SCHWARTZ, AND BUSSE
TABLE 2 COMPARISON OF ASTHMATICS WITH HIGH AND LOW BAL HISTAMINE* Histamine
Histamine 100 pglml
Parameter
< 100 pglml
~
Samples, n Histamine, pglml Tryptase, nglml FEV1 , % pred FEV1/FVC PD2o, cumulative breath units Volume return, ml SAL total protein, ~glml Total cells, 10' AM, % Eos, % Neu, % Lym, %
23 48 :t 4 0.39:t 0.05 96 :t 3 rr e 2 41.5 ± 13.7
23 291 ::t 26 0.33 :t: 0.03
180 ± 3 68 ± 3 15.5 ± 1.8 93.0 ± 1.0
0.6::t 0.1 0.7::t 0.1 5.4::t 0.9
8O:t: 71 :t: 13.1 :t: 149 :t: 8O:t:
3
1 3.4 7 6
12.3 ::t 1.9 90.1 :t: 1.4 2.2 :t 0.4 2.1 :t 0.7 5.6 :t 1.1
p Valuet
0.ססOO
NS* 0.0003 0.04 0.0455 0.0002
NS NS
NS 0.0011 0.0488
NS
For definition of abbreviations, see table 1. • Data are means ± SEM. t p values are based on independent t test. Not significant; p > 0.05 confirmed by Mann-Whitney test.
*
greater airway hyperresponsiveness to methacholine than have those with lower levels « 100 pg/ml) of this mediator (17). These data, coupled with observations from the present study, suggest that BAL histamine concentrations may be an indicator of asthma severity among subjects with stable asthma. However, Casale and coworkers (17) did not identify a difference in FEV 1 volume between the two groups. Possible explanations for the lack of correlation between airflow limitation and BAL histamine concentrations in that study may be due to differences in the severity of asthma, sample size, or their use of actual FEV 1 values for analysis as opposed to FEV 1 percent predicted. The use of actual FEV 1 values can be misleading in small studies because of differences in subjects' predicted values.
BAL Histamine - Tryptase Correlation Allergic Asthmatics 1.2
-e
