- Email: [email protected]
The Detection by Immunoassay of Antibody to Mycobacterial Antigens and Mycobacterial Antigens in Bronchoalveolar Lavage Fluid from Patients with Tuberculosis and Control Subjects
The Detection by Immunoassay of Antibody to Mycobacterial Antigens and Mycobacterial Antigens in Bronchoalveolar Lavage Fluid from Patients with Tuberculosis and Control Subjects· Alamelu Raja, Ph.D.;t Robert P. Baughman, M.D., F.G.G.P.; and Thomas M. Daniel, M.D., F.G.C.P.
Bronchoalveolar lavage (BAL) samples from patients with pulmonary tuberculosis and from control subjects were studied using direct and inhibition enzyme-linked immunosorbent assay (ELISA) for the detection of IgG and IgA antibody to Mycobacterium tuberculosU culture 6ltrate and purified antigen 5 and for the detection of mycobacterial antigens. The IgG and IgA antibodies were found in the majority of samples from tuberculous patients. The IgG titers were higher than were IgA titers. Substantial non-
speci6city was observed for antibodies to crude M tuberculoBia culture 6ltrate. Antibodies to purified M tuberculoBia
antigen 5 were more speci6c. The detection of antibody in BAL did not offer advantages over previously described serodiagnostic assays. Antigen was found in some BAL samples from tuberculous patients. Greater speci&city was observed with an assay based on puri6ed antigen 5 than on crude mycobacterial culture 6ltrate.
The diagnosis of pulmonary tuberculosis presently rests on microbiologic techniques that have not changed substantially during the past century Although many patients with tuberculosis are examined via the bronchoscope during the course of diagnostic investigation, bronchoscopic specimens are currently studied only by classic microbiologic means. While accurate, these methods are limited. Direct smear identification of mycobacteria is of limited sensitivity and does not differentiate among various species of mycobacteria. Culture is both sensitive and specific, but usually requires several weeks for specific identification, even using rapid radiometric methods. In this study we report our experience using direct and inhibition enzyme-linked immunoabsorbent assay (E LISA)for the detection ofantibody to mycobacterial antigens and for the detection of mycobacterial antigens in bronchoalveolar lavage (BAL) samples from patients with pulmonary tuberculosis and from control subjects.
patients. Patients were divided into five groups based on clinical diagnosis. Patients were classified as having tuberculosis. disease due to mycobacteria other than Mycobacterium tuberculosis (MOIT) (two cases infected with M kansasii, two with M avium complex), or histoplasmosis on the basis of culture of sputum or specimens obtained at bronchoscopy Patients were classified as having sarcoidosis if they had a compatible clinical histo~ chest roentgenogram, and lung biopsy Control patients were those individuals who underwent bronchoscopy for localized lesions. Bronchoscopy and BAL were performed in a standard manner. A flexible fiberoptic bronchoscope was advanced to a distal bronchus either in the area of infiltrate or, in patients with diffuse disease, to the right middle lobe. In the control patients. the lavage was performed in an area not near the localized lesion, either the right middle lobe or the lingula of the left upper lobe. Once the bronchoscope was wedged, two to four aliquots of 60 ml of normal saline solution were introduced and immediately aspirated using a hand-held syringe. The aspirated fluid was pooled. An aliquot of the fluid was removed for total and differential cell counts. Slides were prepared of each BAL specimen by spinning 0.2 to 0.4 ml aliquots of unconcentrated BAL fluid onto each glass slide using a cytocentrifuge. The slides were air dried and stained with a modified Wright-Giemsa stain (Diff-Quik, American Scientific Products. McGaw Park, IL). Differential cell counts were performed on 200 nucleated cells per slide.
Study Subjects and Bronchoscopic Procedures
EUSA Measurmaent of Antibody
METHODS
Bronchoscopy with bronchoalveolar lavage was performed in 52
·From the Department of Medicine, Cue Weitern Reserve Universttyand University Hospitals, Cleveland, and the Department of Medicine, University of Cincinnati, Cinoinnati. tCurrently at Tuberculosis Research Centre, Chetput, Madras. India. Supportedby Grant AI15102 from the National InstituteofAllergy and Infectious Diseases, Bethesda, MD, and by a grant from the Vitek Corporatton, St. Louis. MO Dr. Raja was the r6Cipient of a fellowship from the World Health Orpniution during these studies. Manuscript received November 3; revision acceptedJanuary 4. Rsprint ""q~"': Dr. DanHJI. DBfJ!lrlmBnt of M"dWifUl, Un'ver,'ty H0&p'tal&, 2014 AbIngton Road. C~land 44106
Antibody was measured by the ELISA method of Benjamin and
Daniell with minor modificationI. In brief, polyvinyl chloride mlerottter plates (Becton.. Dickinson, Oxnard, CA) were sensitized with unheated crude culture filtrate of M tubBrculom. H37&' or puri8ed M tubf1rCUlo.u antigen 53 and then quenched with bovine serum albumin. Duplicate aliquot. of undiluted HAL samples were incubatedin the microtiter plate wells. after which the wells were washed. Affinity purified specific anti-human IIG or IgA conjugated to alkaline phosphatase (Sigma Chemical Co., St. Louis, MO) at a previously determined optimaldilutionwasthen incubatedin each well. Afterwaahing, p-nttrophenyl phosphate substrate wu added to eachwelland colordevelopment allowed to proceed. Whencolor CHEST I .. I 1 I JUL~
1_
113
Table I-Patient Age and Cellular CharacteriBtica of42 Bronchoalveolar Lavage Fluw Bronchoalveolar Lavage Differential Cell Count (%± SEM)
Group
No. of Patients
Age (Years ± SEM)
Macrophages
Lymphocytes
Neutrophils
Thberculosis Mycobacterial pulmonary infections other than tuberculosis Histoplasmosis Sarcoidosis Control subjects
9 4 4 8 17
52±5.4 5O± 1.0 41 ±8.9 4O±4.0 54±3.0
68.9±5.2 75.5± 1.9 87.0±4.3 67.4±4.2 89.5± 1.3
21.3±3.6* 17.0±2.7 11.2±3.7 29.5±4.5* 6.4±0.8
9.1±3.7 5.5±2.3 1.8± 1.0 1.6±0.6 2.8±0.6
*Differs from controls, p
The inhibition ELISA was performed using an antiserum raised in a goat immunized by the repetitive subcutaneous injection of purified M tuberculosis antigen 5 without Freunds adjuvants. It reacted principally with antigen 5. Plates for this assay were sensitized with either M tuberculosis crude unheated culture filtrate! or purified antigen 5. 3 Antiserum in a dilution of 1:5,000 was preincubated with equal volumes of BAL sample that had been concentrated tenfold by lyophilization and reconstitution. Inhibition ELISA was then completed in the standard fashion, 1 and the antigen concentration determined from a standard curve. This assay was found to have a sensitivity of 5 ng/ml when used to detect purified antigen 5 with the minimum positive level taken at a point 3 standard deviations from the control mean.
Statistical Methods
Analysis of variance (ANOVA) was used for comparisons of means using the Microstat statistical software (Ecosoft, Inc, Indianapolis, IN~
RESULTS
Characteristics of BAL Fluids The cellular characteristics of BAL fluids obtained from each diagnostic category of patient are presented in Table 1. Patients with tuberculosis and sarcoidosis had a lymphocytosis.
Detection ofAntibody in BAL Both IgG and IgA antibody were measured by ELISA in all of the BAL samples, and each sample was tested for each antibody using solid-phase absorbents coated with two mycobacterial antigen preparations. The results are presented in Figure 1 to 4. The IgG antibody to crude unheated culture filtrate of M tuberculosis (Fig 1) was found in BAL from patients with tuberculosis and from patients with
SAL - IQG ANTIBODY TO CULTURE FILTRATE BAL - IgG ANTIBODY TO ANTIGEN 5 ~2.0
••
~2.0 ~
l-
en
Z W 0 ..J
1.5
1.0
U
~
a..
0
.5
0
••• •••
_.-·
• • •
•••
• ••
•
T8
MOTT
••
.•••. ...• ... · ----...· •
• I I • I
HISTO
SARCOID
1.5
w
0
...J
1.0
t=
a..
0
....•.
~
U
NORMAL
FIGURE 1. IgG antibody to crude unheated culture filtrate of Mycobacterium tuberculosis in BAL samples from patients with pulmonary tuberculosis (TB), mycobacterial pulmonary infections other than tuberculosis (MOTT), pulmonary histoplasmosis (HIS TO), pulmonary sarcoidosis (SARCOID), and normal volunteers (NORMAL). Each BAL sample result is represented by a single solid point, and means for groups are shown by horizontal lines. Results of ELISA assays are given as the optical density developed in the ELISA and measured using an automated plate reader.
134
~
l(j) Z
.5
0
··
T8
··• .·.
•
..L..!..
MOTT
HISTO
SARCOID
• •• .A. •• ••• ••
•
NORMAL
2. IgG antibody to purified Mycobacterium tuberculosis antigen 5 in BAL samples from patients with pulmonary tuberculosis (TB), mycobacterial pulmonary infections other than tuberculosis (MarT), pulmonary histoplasmosis (HISTO), pulmonary sarcoidosis (SARCOID), and normal volunteers (NORMAL~ Each BAL sample result is represented by a single solid point, and means for groups are shown by horizontallines. Results of ELISA assays are given as the optical density developed in the ELISA and measured using an automated plate reader. FICURE
Mycob8ctet1aI Antigen and Tuberculosis (Raja, Baughman, DanIel)
SAL - lOA ANTIBODY TO CULTURE FILTRATE 2.0
>t:: en z w
0 -I
~IOOO
.s 0 ~
1.5
ex
• 1.0
o
0
~C'
z
i= a.
..
SAL - ANTIGEN DETECTION WITH CULTURE FILTRATE ASSAY
.5
0
••
·•• TB
•
• --L..
• • MOTT
HISTO
....··• •
• • -+• ••• SARCOID
..:•.. NORMAL
FIGURE 3. IgA antibody to crude unheated culture filtrate of Mycobacterium tuberculosis in BAL samples from patients with pulmonary tuberculosis (TB), mycobacterial pulmonary infections other than tuberculosis (MOTT), pulmonary histoplasmosis (HIS TO), pulmonary sarcoidosis (SARCOID), and normal volunteers (NORMAL). Each BAL sample result is represented by a single solid point, and means for groups are shown by horizontal lines. Results of ELISA assays are given as the optical density developed in the ELISA and measured using an automated plate reader.
mycobacterial infections other than tuberculosis, histoplasmosis, sarcoidosis, and normal control subjects with no significant difference in the mean titers in these groups. Although the mean titer in normal subjects was lower than the mean titer in the other groups, it was not significantly lower than the mean for tuberculosis patients (p = 0.07). The IgG antibody to purified M tuberculosis antigen 5 was found in lower titer than antibody to culture filtrate (Fig 2). In this case, the difference in mean titer between tuberSAL - IgA ANTIBODY TO ANTIGEN 5 ~2.0
a:::
I-
800 600
z
su W
u
400
w
z
200
c:r
<5
C)
t= Z
TB
MOTT
HISTO
SARCOID
NORMAL
5. Mycobacterial antigen concentration in tenfold concentrated BAL samples from patients with pulmonary tuberculosis (TB), mycobacterial pulmonary infections other than tuberculosis (MarT), pulmonary histoplasmosis (HISTO), pulmonary sarcoidosis (SARCOID), and normal volunteers (NORMAL) as measured using ELISA with plates sensitized with crude unheated culture filtrate of Mycobacterium tuberculosis. Each BAL sample result is represented by a single solid point. Antigen concentration is expressed in nanograms per milliliter. FIGURE
culosis patients and control subjects was significant (p
Detection ofAntigen in BAL
FIGURE 4. IgA antibody to purified Mycobacterium tuberculosis antigen 5 in BAL samples from patients with pulmonary tuberculosis (TB), mycobacterial pulmonary infections other than tuberculosis (MarT), pulmonary histoplasmosis (HIS TO), pulmonary sarcoidosis (SARCOID), and normal volunteers (NORMAL). Each BAL sample result is represented by a single solid point, and means for groups are shown by horizontal lines. Results of ELISA assays are given as the optical density developed in the ELISA and measured using an automated plate reader.
Inhibition ELISA for the measurement of mycobacterial antigen in BAL was performed using an antiserum principally reactive with antigen 5. The assay was performed both using microtiter plates sensitized with culture filtrate and with antigen 5. Results are shown in Figures 5 and 6 for each assay system. The culture filtrate assay (Fig 5) proved to be nonspecific, with positive results occurring in large numbers of control samples. The antigen 5 assay (Fig 6) was more specific. Antigen was detected in six of nine BAL samples from patients with tuberculosis. Very high levels were found in three samples from tuberculosis patients, and in none of the control specimens, However, antigen at moderate levels was detected in one of four samples from patients with histoplasmosis, two of seven samCHEST I 94 I 1 I JULY; 1988
135
pIes from patients with sarcoidosis, and one of 16 samples from normal control subjects. Thus, only detection of a very high level of antigen could be considered diagnostically useful. DISCUSSION
The measurement of serum antibody for the diagnosis of tuberculosis using ELISA techniques has been studied extensively" It is clear that serodiagnosis using ELI'SA provides reasonably favorable test characteristics. Both IgG and 19A serum antibody have been shown to correlate with active tuberculosis, but titers of IgG have generally been found to be higher than titers of IgA. The ELISA methods that use purified antigens are better than those that use crude culture filtrate or tuberculin purified protein derivative (PPD). Among those purified antigens that perform well in ELISA, the antigen 5 used in the present study has been evaluated extensively 1,4-8 Our results using BAL parallel the large experience with serum." Even though estimation of BAL antibody against antigen 5 did not yield a clear cut off point between patients with active tuberculosis and control subjects, antibody to this purified antigen proved to be more specific for tuberculosis than antibody to crude antigen. The detection of antibody in BAL samples raises interesting questions about the pathogenesis of tuberculosis that may be deserving of further study Although antibody-mediated immune responses are thought to have little to do with the immunopathogenesis of tuberculosis, antibody might play a local role in the bronchial tree. It is not inappropriate to ask why transbronchial spread of disease is a relatively SAL - ANTIGEN DETECTION WITH ANTIGEN 5 ASSAY
E
~IOOO
<,
...
REFERENCES
C'
.5
z
800
~
600
0
a::
t-
z
W
u
Z
0
400
U
z
w
C)
i=
z
«
200 <5
• TB
MOTT
HISTO
SARCOID
NORMAL
FIGURE 6. Mycobacterial antigen concentration in tenfold concentrated BAL samples from patients with pulmonary tuberculosis (TB), mycobacterial pulmonary infections other than tuberculosis tuott; pulmonary histoplasmosis (HIS TO), pulmonary sarcoidosis (SARCOID), and normal volunteers (NORMAL) as measured using EUSA with plates sensitized with Mycobacterium tuberculosis purified antigen 5. Each BAL sample result is represented by a single solid point. Antigen concentration is expressed in nanograms per milliliter.
138
uncommon event in patients whose bronchial secretions contain virulent tubercle bacilli and whether local antibody may be protective in this context. It is known that 19A antibody is the dominant antibody of the upper airway and IgG of the lower airway 9 so it is of interest that we found both classes of antibody in BAL samples. The IgG titers were higher, perhaps reflecting the predominantly lower airway source of the BAL samples. It also must be recognized that the lymphocytosis in our tuberculosis patients might be associated with nonspecific increase in BAL IgG.lO Such an increase would probably not affect our determinations of specific antibody and our sarcoidosis patients provide a control for this possibility Antigen has been previously detected in the sputum of patients with tuberculosis by radioimmunoassay!' and by ELISA. 12 The BAL samples have not previously been studied. An amount of antigen 5 equal to or greater than 1,000 ng/ml in BAL was specific for the diagnosis of tuberculosis. However, six of nine samples from tuberculosis patients did not contain such amounts of antigen. This might reflect the fact that the lavage was not usually performed in an obviously diseased pulmonary segment. The detection of antigen 5 was more specific for the diagnosis of tuberculosis than assay using crude antigen. The number of falsepositive antigen 5 assays of BAL was small, and none of the false-positive results were high values. Because cultural identification of mycobacteria is time consuming, there is need for rapid and specific alternative means of recognition of mycobacteria in clinical specimens. Immunoassays have the theoretic potential for such applications. Our experience in this study suggests that the approach of immunoassay of BAL for mycobacterial antigens might be useful, but further studies are needed. 1 Benjamin RG, Daniel TM. Serodiagnosis of tuberculosis using the enzyme-linked immunoabsorbent assay (ELISA) of antibody to Mycobacterium tuberculosis antigen 5. Am Rev Respir Dis 1982; 126:1013-16 2 Daniel TM, Ferguson LE. Purification and characterization of two proteins from culture filtrates of Mycobacterium tuberculosis H37Ra strain. Infect Immun 1970; 1:164-68 3 Daniel TM, Anderson PA. The isolation by immunoabsorbent affinity chromatography and physicochemical characterization of Mycobacterium tuberculosis antigen 5. Am Rev Respir Dis 1978; 117:533-39 4 Daniel TM, Debanne SM. The serodiagnosis of tuberculosis and other mycobacterial diseases by enzyme-linked immunosorbent assay Am Rev Respir Dis 1987; 135:1137-51 5 Balestrino EA, Daniel TM, de Latini MDS, Latini OA, Ma ~ ScocozzaJB. Serodiagnosis of pulmonary tuberculosis in Argentina by enzyme-linked immunosorbent assay (ELISA) of IgG antibody to Mycobacterium tuberculosis antigen 5 and tuberculin purified protein derivative. Bull World Health Org 1984; 62:755-61 6 Daniel TM, Debanne SM, van der Kuyp F. Enzyme-linked immunosorbent assay using Mycobacterium tuberculosis antigen Mycobacterial Antigen and Tuberculosis (Raja, Baughman, Daniel)
5 and PPD for the serodiagnosis of tuberculosis. Chest 1985; 88:388-92 7 Ma Y, Wang YM, Daniel TM. Enzyme-linked immunosorbent assay using Mycobacterium tuberculosis antigen 5 for the diagnosis of pulmonary tuberculosis in China. Am Rev Respir Dis 1986; 134:1273-75 8 AIde S, Mosca C, Gonzalez Montaner LJ. EI diagnostico de actividad tuberculosa por enzimoinmunoensayo en fase solida (ELISA) usando antigeno 5 del M. tuberculosis. La Semana Medica 1986; 169:251-55 9 Kaltreider HB. Expression of immune mechanisms in the lung.
Am Rev Resp Dis 1976; 113:347-79 10 Lawrence EC, Martin RR, Blaese RM, Teague RB, Awe RJ, Wilson RK, et aI. Increased bronchoalveolar IgG-secreting cells in interstitial lung diseases. N Eng} J Med 1980; 302:1186-88 11 Strauss E, Wu N, Quraishi MAH, Levine S. Clinical applications of the radioimmunoassay of secretory tuberculoprotein. Proc Natl Acad Sci USA 1981; 78:3214-17 12 Yanez MA, Coppola M~ Russo DA, Delaha E, Chaparas SD, Yeager H Jr. Determination of mycobacterial antigens in sputum by enzyme immunoassay J Clin Microbioll986; 23:822-25
CHEST I 94 I 1 I JUL~
1988
137
Bronchoalveolar lavage (BAL) samples from patients with pulmonary tuberculosis and from control subjects were studied using direct and inhibition enzyme-linked immunosorbent assay (ELISA) for the detection of IgG and IgA antibody to Mycobacterium tuberculosU culture 6ltrate and purified antigen 5 and for the detection of mycobacterial antigens. The IgG and IgA antibodies were found in the majority of samples from tuberculous patients. The IgG titers were higher than were IgA titers. Substantial non-
speci6city was observed for antibodies to crude M tuberculoBia culture 6ltrate. Antibodies to purified M tuberculoBia
antigen 5 were more speci6c. The detection of antibody in BAL did not offer advantages over previously described serodiagnostic assays. Antigen was found in some BAL samples from tuberculous patients. Greater speci&city was observed with an assay based on puri6ed antigen 5 than on crude mycobacterial culture 6ltrate.
The diagnosis of pulmonary tuberculosis presently rests on microbiologic techniques that have not changed substantially during the past century Although many patients with tuberculosis are examined via the bronchoscope during the course of diagnostic investigation, bronchoscopic specimens are currently studied only by classic microbiologic means. While accurate, these methods are limited. Direct smear identification of mycobacteria is of limited sensitivity and does not differentiate among various species of mycobacteria. Culture is both sensitive and specific, but usually requires several weeks for specific identification, even using rapid radiometric methods. In this study we report our experience using direct and inhibition enzyme-linked immunoabsorbent assay (E LISA)for the detection ofantibody to mycobacterial antigens and for the detection of mycobacterial antigens in bronchoalveolar lavage (BAL) samples from patients with pulmonary tuberculosis and from control subjects.
patients. Patients were divided into five groups based on clinical diagnosis. Patients were classified as having tuberculosis. disease due to mycobacteria other than Mycobacterium tuberculosis (MOIT) (two cases infected with M kansasii, two with M avium complex), or histoplasmosis on the basis of culture of sputum or specimens obtained at bronchoscopy Patients were classified as having sarcoidosis if they had a compatible clinical histo~ chest roentgenogram, and lung biopsy Control patients were those individuals who underwent bronchoscopy for localized lesions. Bronchoscopy and BAL were performed in a standard manner. A flexible fiberoptic bronchoscope was advanced to a distal bronchus either in the area of infiltrate or, in patients with diffuse disease, to the right middle lobe. In the control patients. the lavage was performed in an area not near the localized lesion, either the right middle lobe or the lingula of the left upper lobe. Once the bronchoscope was wedged, two to four aliquots of 60 ml of normal saline solution were introduced and immediately aspirated using a hand-held syringe. The aspirated fluid was pooled. An aliquot of the fluid was removed for total and differential cell counts. Slides were prepared of each BAL specimen by spinning 0.2 to 0.4 ml aliquots of unconcentrated BAL fluid onto each glass slide using a cytocentrifuge. The slides were air dried and stained with a modified Wright-Giemsa stain (Diff-Quik, American Scientific Products. McGaw Park, IL). Differential cell counts were performed on 200 nucleated cells per slide.
Study Subjects and Bronchoscopic Procedures
EUSA Measurmaent of Antibody
METHODS
Bronchoscopy with bronchoalveolar lavage was performed in 52
·From the Department of Medicine, Cue Weitern Reserve Universttyand University Hospitals, Cleveland, and the Department of Medicine, University of Cincinnati, Cinoinnati. tCurrently at Tuberculosis Research Centre, Chetput, Madras. India. Supportedby Grant AI15102 from the National InstituteofAllergy and Infectious Diseases, Bethesda, MD, and by a grant from the Vitek Corporatton, St. Louis. MO Dr. Raja was the r6Cipient of a fellowship from the World Health Orpniution during these studies. Manuscript received November 3; revision acceptedJanuary 4. Rsprint ""q~"': Dr. DanHJI. DBfJ!lrlmBnt of M"dWifUl, Un'ver,'ty H0&p'tal&, 2014 AbIngton Road. C~land 44106
Antibody was measured by the ELISA method of Benjamin and
Daniell with minor modificationI. In brief, polyvinyl chloride mlerottter plates (Becton.. Dickinson, Oxnard, CA) were sensitized with unheated crude culture filtrate of M tubBrculom. H37&' or puri8ed M tubf1rCUlo.u antigen 53 and then quenched with bovine serum albumin. Duplicate aliquot. of undiluted HAL samples were incubatedin the microtiter plate wells. after which the wells were washed. Affinity purified specific anti-human IIG or IgA conjugated to alkaline phosphatase (Sigma Chemical Co., St. Louis, MO) at a previously determined optimaldilutionwasthen incubatedin each well. Afterwaahing, p-nttrophenyl phosphate substrate wu added to eachwelland colordevelopment allowed to proceed. Whencolor CHEST I .. I 1 I JUL~
1_
113
Table I-Patient Age and Cellular CharacteriBtica of42 Bronchoalveolar Lavage Fluw Bronchoalveolar Lavage Differential Cell Count (%± SEM)
Group
No. of Patients
Age (Years ± SEM)
Macrophages
Lymphocytes
Neutrophils
Thberculosis Mycobacterial pulmonary infections other than tuberculosis Histoplasmosis Sarcoidosis Control subjects
9 4 4 8 17
52±5.4 5O± 1.0 41 ±8.9 4O±4.0 54±3.0
68.9±5.2 75.5± 1.9 87.0±4.3 67.4±4.2 89.5± 1.3
21.3±3.6* 17.0±2.7 11.2±3.7 29.5±4.5* 6.4±0.8
9.1±3.7 5.5±2.3 1.8± 1.0 1.6±0.6 2.8±0.6
*Differs from controls, p
The inhibition ELISA was performed using an antiserum raised in a goat immunized by the repetitive subcutaneous injection of purified M tuberculosis antigen 5 without Freunds adjuvants. It reacted principally with antigen 5. Plates for this assay were sensitized with either M tuberculosis crude unheated culture filtrate! or purified antigen 5. 3 Antiserum in a dilution of 1:5,000 was preincubated with equal volumes of BAL sample that had been concentrated tenfold by lyophilization and reconstitution. Inhibition ELISA was then completed in the standard fashion, 1 and the antigen concentration determined from a standard curve. This assay was found to have a sensitivity of 5 ng/ml when used to detect purified antigen 5 with the minimum positive level taken at a point 3 standard deviations from the control mean.
Statistical Methods
Analysis of variance (ANOVA) was used for comparisons of means using the Microstat statistical software (Ecosoft, Inc, Indianapolis, IN~
RESULTS
Characteristics of BAL Fluids The cellular characteristics of BAL fluids obtained from each diagnostic category of patient are presented in Table 1. Patients with tuberculosis and sarcoidosis had a lymphocytosis.
Detection ofAntibody in BAL Both IgG and IgA antibody were measured by ELISA in all of the BAL samples, and each sample was tested for each antibody using solid-phase absorbents coated with two mycobacterial antigen preparations. The results are presented in Figure 1 to 4. The IgG antibody to crude unheated culture filtrate of M tuberculosis (Fig 1) was found in BAL from patients with tuberculosis and from patients with
SAL - IQG ANTIBODY TO CULTURE FILTRATE BAL - IgG ANTIBODY TO ANTIGEN 5 ~2.0
••
~2.0 ~
l-
en
Z W 0 ..J
1.5
1.0
U
~
a..
0
.5
0
••• •••
_.-·
• • •
•••
• ••
•
T8
MOTT
••
.•••. ...• ... · ----...· •
• I I • I
HISTO
SARCOID
1.5
w
0
...J
1.0
t=
a..
0
....•.
~
U
NORMAL
FIGURE 1. IgG antibody to crude unheated culture filtrate of Mycobacterium tuberculosis in BAL samples from patients with pulmonary tuberculosis (TB), mycobacterial pulmonary infections other than tuberculosis (MOTT), pulmonary histoplasmosis (HIS TO), pulmonary sarcoidosis (SARCOID), and normal volunteers (NORMAL). Each BAL sample result is represented by a single solid point, and means for groups are shown by horizontal lines. Results of ELISA assays are given as the optical density developed in the ELISA and measured using an automated plate reader.
134
~
l(j) Z
.5
0
··
T8
··• .·.
•
..L..!..
MOTT
HISTO
SARCOID
• •• .A. •• ••• ••
•
NORMAL
2. IgG antibody to purified Mycobacterium tuberculosis antigen 5 in BAL samples from patients with pulmonary tuberculosis (TB), mycobacterial pulmonary infections other than tuberculosis (MarT), pulmonary histoplasmosis (HISTO), pulmonary sarcoidosis (SARCOID), and normal volunteers (NORMAL~ Each BAL sample result is represented by a single solid point, and means for groups are shown by horizontallines. Results of ELISA assays are given as the optical density developed in the ELISA and measured using an automated plate reader. FICURE
Mycob8ctet1aI Antigen and Tuberculosis (Raja, Baughman, DanIel)
SAL - lOA ANTIBODY TO CULTURE FILTRATE 2.0
>t:: en z w
0 -I
~IOOO
.s 0 ~
1.5
ex
• 1.0
o
0
~C'
z
i= a.
..
SAL - ANTIGEN DETECTION WITH CULTURE FILTRATE ASSAY
.5
0
••
·•• TB
•
• --L..
• • MOTT
HISTO
....··• •
• • -+• ••• SARCOID
..:•.. NORMAL
FIGURE 3. IgA antibody to crude unheated culture filtrate of Mycobacterium tuberculosis in BAL samples from patients with pulmonary tuberculosis (TB), mycobacterial pulmonary infections other than tuberculosis (MOTT), pulmonary histoplasmosis (HIS TO), pulmonary sarcoidosis (SARCOID), and normal volunteers (NORMAL). Each BAL sample result is represented by a single solid point, and means for groups are shown by horizontal lines. Results of ELISA assays are given as the optical density developed in the ELISA and measured using an automated plate reader.
mycobacterial infections other than tuberculosis, histoplasmosis, sarcoidosis, and normal control subjects with no significant difference in the mean titers in these groups. Although the mean titer in normal subjects was lower than the mean titer in the other groups, it was not significantly lower than the mean for tuberculosis patients (p = 0.07). The IgG antibody to purified M tuberculosis antigen 5 was found in lower titer than antibody to culture filtrate (Fig 2). In this case, the difference in mean titer between tuberSAL - IgA ANTIBODY TO ANTIGEN 5 ~2.0
a:::
I-
800 600
z
su W
u
400
w
z
200
c:r
<5
C)
t= Z
TB
MOTT
HISTO
SARCOID
NORMAL
5. Mycobacterial antigen concentration in tenfold concentrated BAL samples from patients with pulmonary tuberculosis (TB), mycobacterial pulmonary infections other than tuberculosis (MarT), pulmonary histoplasmosis (HISTO), pulmonary sarcoidosis (SARCOID), and normal volunteers (NORMAL) as measured using ELISA with plates sensitized with crude unheated culture filtrate of Mycobacterium tuberculosis. Each BAL sample result is represented by a single solid point. Antigen concentration is expressed in nanograms per milliliter. FIGURE
culosis patients and control subjects was significant (p
Detection ofAntigen in BAL
FIGURE 4. IgA antibody to purified Mycobacterium tuberculosis antigen 5 in BAL samples from patients with pulmonary tuberculosis (TB), mycobacterial pulmonary infections other than tuberculosis (MarT), pulmonary histoplasmosis (HIS TO), pulmonary sarcoidosis (SARCOID), and normal volunteers (NORMAL). Each BAL sample result is represented by a single solid point, and means for groups are shown by horizontal lines. Results of ELISA assays are given as the optical density developed in the ELISA and measured using an automated plate reader.
Inhibition ELISA for the measurement of mycobacterial antigen in BAL was performed using an antiserum principally reactive with antigen 5. The assay was performed both using microtiter plates sensitized with culture filtrate and with antigen 5. Results are shown in Figures 5 and 6 for each assay system. The culture filtrate assay (Fig 5) proved to be nonspecific, with positive results occurring in large numbers of control samples. The antigen 5 assay (Fig 6) was more specific. Antigen was detected in six of nine BAL samples from patients with tuberculosis. Very high levels were found in three samples from tuberculosis patients, and in none of the control specimens, However, antigen at moderate levels was detected in one of four samples from patients with histoplasmosis, two of seven samCHEST I 94 I 1 I JULY; 1988
135
pIes from patients with sarcoidosis, and one of 16 samples from normal control subjects. Thus, only detection of a very high level of antigen could be considered diagnostically useful. DISCUSSION
The measurement of serum antibody for the diagnosis of tuberculosis using ELISA techniques has been studied extensively" It is clear that serodiagnosis using ELI'SA provides reasonably favorable test characteristics. Both IgG and 19A serum antibody have been shown to correlate with active tuberculosis, but titers of IgG have generally been found to be higher than titers of IgA. The ELISA methods that use purified antigens are better than those that use crude culture filtrate or tuberculin purified protein derivative (PPD). Among those purified antigens that perform well in ELISA, the antigen 5 used in the present study has been evaluated extensively 1,4-8 Our results using BAL parallel the large experience with serum." Even though estimation of BAL antibody against antigen 5 did not yield a clear cut off point between patients with active tuberculosis and control subjects, antibody to this purified antigen proved to be more specific for tuberculosis than antibody to crude antigen. The detection of antibody in BAL samples raises interesting questions about the pathogenesis of tuberculosis that may be deserving of further study Although antibody-mediated immune responses are thought to have little to do with the immunopathogenesis of tuberculosis, antibody might play a local role in the bronchial tree. It is not inappropriate to ask why transbronchial spread of disease is a relatively SAL - ANTIGEN DETECTION WITH ANTIGEN 5 ASSAY
E
~IOOO
<,
...
REFERENCES
C'
.5
z
800
~
600
0
a::
t-
z
W
u
Z
0
400
U
z
w
C)
i=
z
«
200 <5
• TB
MOTT
HISTO
SARCOID
NORMAL
FIGURE 6. Mycobacterial antigen concentration in tenfold concentrated BAL samples from patients with pulmonary tuberculosis (TB), mycobacterial pulmonary infections other than tuberculosis tuott; pulmonary histoplasmosis (HIS TO), pulmonary sarcoidosis (SARCOID), and normal volunteers (NORMAL) as measured using EUSA with plates sensitized with Mycobacterium tuberculosis purified antigen 5. Each BAL sample result is represented by a single solid point. Antigen concentration is expressed in nanograms per milliliter.
138
uncommon event in patients whose bronchial secretions contain virulent tubercle bacilli and whether local antibody may be protective in this context. It is known that 19A antibody is the dominant antibody of the upper airway and IgG of the lower airway 9 so it is of interest that we found both classes of antibody in BAL samples. The IgG titers were higher, perhaps reflecting the predominantly lower airway source of the BAL samples. It also must be recognized that the lymphocytosis in our tuberculosis patients might be associated with nonspecific increase in BAL IgG.lO Such an increase would probably not affect our determinations of specific antibody and our sarcoidosis patients provide a control for this possibility Antigen has been previously detected in the sputum of patients with tuberculosis by radioimmunoassay!' and by ELISA. 12 The BAL samples have not previously been studied. An amount of antigen 5 equal to or greater than 1,000 ng/ml in BAL was specific for the diagnosis of tuberculosis. However, six of nine samples from tuberculosis patients did not contain such amounts of antigen. This might reflect the fact that the lavage was not usually performed in an obviously diseased pulmonary segment. The detection of antigen 5 was more specific for the diagnosis of tuberculosis than assay using crude antigen. The number of falsepositive antigen 5 assays of BAL was small, and none of the false-positive results were high values. Because cultural identification of mycobacteria is time consuming, there is need for rapid and specific alternative means of recognition of mycobacteria in clinical specimens. Immunoassays have the theoretic potential for such applications. Our experience in this study suggests that the approach of immunoassay of BAL for mycobacterial antigens might be useful, but further studies are needed. 1 Benjamin RG, Daniel TM. Serodiagnosis of tuberculosis using the enzyme-linked immunoabsorbent assay (ELISA) of antibody to Mycobacterium tuberculosis antigen 5. Am Rev Respir Dis 1982; 126:1013-16 2 Daniel TM, Ferguson LE. Purification and characterization of two proteins from culture filtrates of Mycobacterium tuberculosis H37Ra strain. Infect Immun 1970; 1:164-68 3 Daniel TM, Anderson PA. The isolation by immunoabsorbent affinity chromatography and physicochemical characterization of Mycobacterium tuberculosis antigen 5. Am Rev Respir Dis 1978; 117:533-39 4 Daniel TM, Debanne SM. The serodiagnosis of tuberculosis and other mycobacterial diseases by enzyme-linked immunosorbent assay Am Rev Respir Dis 1987; 135:1137-51 5 Balestrino EA, Daniel TM, de Latini MDS, Latini OA, Ma ~ ScocozzaJB. Serodiagnosis of pulmonary tuberculosis in Argentina by enzyme-linked immunosorbent assay (ELISA) of IgG antibody to Mycobacterium tuberculosis antigen 5 and tuberculin purified protein derivative. Bull World Health Org 1984; 62:755-61 6 Daniel TM, Debanne SM, van der Kuyp F. Enzyme-linked immunosorbent assay using Mycobacterium tuberculosis antigen Mycobacterial Antigen and Tuberculosis (Raja, Baughman, Daniel)
5 and PPD for the serodiagnosis of tuberculosis. Chest 1985; 88:388-92 7 Ma Y, Wang YM, Daniel TM. Enzyme-linked immunosorbent assay using Mycobacterium tuberculosis antigen 5 for the diagnosis of pulmonary tuberculosis in China. Am Rev Respir Dis 1986; 134:1273-75 8 AIde S, Mosca C, Gonzalez Montaner LJ. EI diagnostico de actividad tuberculosa por enzimoinmunoensayo en fase solida (ELISA) usando antigeno 5 del M. tuberculosis. La Semana Medica 1986; 169:251-55 9 Kaltreider HB. Expression of immune mechanisms in the lung.
Am Rev Resp Dis 1976; 113:347-79 10 Lawrence EC, Martin RR, Blaese RM, Teague RB, Awe RJ, Wilson RK, et aI. Increased bronchoalveolar IgG-secreting cells in interstitial lung diseases. N Eng} J Med 1980; 302:1186-88 11 Strauss E, Wu N, Quraishi MAH, Levine S. Clinical applications of the radioimmunoassay of secretory tuberculoprotein. Proc Natl Acad Sci USA 1981; 78:3214-17 12 Yanez MA, Coppola M~ Russo DA, Delaha E, Chaparas SD, Yeager H Jr. Determination of mycobacterial antigens in sputum by enzyme immunoassay J Clin Microbioll986; 23:822-25
CHEST I 94 I 1 I JUL~
1988
137