Nonspecificity of the Anda A60-tb ELISA test for serodiagnosis of mycobacterial disease
Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by Iowa State University on 11/10/14 For personal use only.
Microbiology Laboratories, King Faisal Specialist Hospital and Research Centre, P. 0. Box 3354, Riyadh 11211, Saudi Arabia Received October 23, 1991 Revision received January 27, 1992 Accepted February 26, 1992 QADRI,S. M. H., and SMITH,K. K. 1992. Nonspecificity of the Anda A60-tb ELISA test for serodiagnosis of mycobacterial disease. Can. J. Microbiol. 38: 804-806. The conventional methods for the laboratory diagnosis of tuberculosis and other mycobacterial diseases are time consuming and beyond the scope of most of the small and medium-sized hospital facilities. Therefore, there has been considerable interest in the development of a serological method for the detection of antibodies against mycobacteria. We recently evaluated a commercially available ELISA test (Anda Biologicals, Strasbourg, France) that measures antibody levels to A60 antigen, a membrane glycoprotein that is found in most mycobacteria. Of the 123 patients with positive pulmonary.cultures for Mycobacterium tuberculosis, 82% had detectable antibodies against the kit antigen. Of the 68 patients with extrapulmonary tuberculosis, 59% yielded positive results. Specimens from 2 of the 12 patients that grew Mycobacterium avium-intracellulare complex, and one each with Mycobacterium fortuitum and Mycobacterium chelonei, were considered significant on the basis of medical history and repeated isolation of the bacterium from clinical specimens, and these patients yielded positive serology. Of the healthy, normal PPD positive and PPD negative controls, 24% gave false positive results. Key words: Anda A60-tb ELISA,serodiagnosis of mycobacteria, mycobacterium. QADRI,S. M. H., et SMITH, K. K. 1992. Nonspecificity of the Anda A60-tb ELISA test for serodiagnosis of mycobacterial disease. Can. J . Microbiol. 38 : 804-806. Les methodes usuelles de laboratoire pour la diagnostic de la tuberculose et autres infections reliees aux mycobacteries sont longues et ne sont pas a la portee des hdpitaux de petite ou de moyenne taille. Cette situation explique le grand interet pour les methodes serologiques permettant la detection d'anticorps diriges contre les mycobacteries. Nous avons evalue une trousse ELISA commerciale (Anda Biologicals, Strasbourg, France) qui mesure les titres d'anticorps contre l'antigkne A60, une glycoproteine membranaire retrouvee chez la plupart des mycobacteries. La trousse a decele des anticorps chez 82% de 123 patients presentant des cultures pulmonaires positives a Mycobacterium tuberculosis. Chez 68 patients atteints de tuberculose extrapulmonaire, 59% ont donne des resultats positifs. Les prelkvements de 2 parmi 12 patients qui avaient une culture positive au groupe Mycobacterium avium-intracellulare et un positif a Mycobacterium fortuitum et un a Mycobacterium chelonei ont ete consideres comme significatifs d'aprks le dossier medical et l'isolement repet6 de la bacterie dans les prelkvements et ces patients avaient une serologic positive. Chez un groupe d'individus sains dont le PPD de contrale etait positif normal ou negatif, 24% ont donne des resultats serologiques faussement positifs. Mots cles : epreuve ELISA Anda A60-tb, serodiagnostic de la tuberculose, Mycobacterium. [Traduit par la redaction]
Introduction Mycobacterium tuberculosis and several other members of this genus are capable of causing human disease, producing slowly developing destructive granulomatous lesions, with or without caseation. Although the disease can occur in any part of the body, pulmonary manifestations are most common. The organisms can disseminate from lungs to other sites, where they cause extrapulmonary manifestations. Worldwide, nearly 8 million new cases and 3 million deaths result from tuberculosis each year. After a steady decline for several years, the disease is showing a slight increase in the United States, with 22 436 new cases recorded by the Centers for Disease Control, Atlanta, Ga., in 1988. Ability of mycobacteria to cause destructive lesions and to disseminate to other sites in the body necessitates rapid diagnosis for optimal patient care. Successful isolation of these bacteria is dependent upon collection of proper speci-
h his work was presented in part at the Annual Meeting of the American Society for Microbiology, Dallas, Texas, 1991. 2 ~ u t h oto r whom all correspondence should be addressed. Pr~nredIn Canada / lmprime au Canada
mens, is time consuming, and is beyond the scope of many cli"ica1 laboratories. Therefore, ~er010gicaldiagnosis Can be a viable alternative (Grange 1984; Ivany et 01. 1988). In this Paper, we report the evaluation of a commercially available EL1SA test that detects antibody levels to A60 antigen of mycobacteria.
Materials and methods Anda A60-tb test (Anda Biologicals, Strasbourg, France) was performed according to the manufacturer's directions. Briefly, the sera to be tested were diluted in specimen dilution buffer, distributed to the antigen A60 coated wells of microtiter plates, and incubated for 60 min in a 37°C water bath. The wells were washed 5 times with kit wash buffer, using an automatic plate washer (Titertek S8/12, Eflab, Finland), and blotted dry. Peroxidase labelled anti-human IgG was added, and the plates were incubated for 30 min at 37°C. The plates were washed and blotted dry, the substrate (tetramethylbenzene containing hydrogen peroxide) was added to the wells, and the plates were left at 37°C for 15 min. Color development was stopped using 0.5 M H2S04and the final color was read at 450 nm (Electronucleonics, Fairfield, N.J.). Results were calculated according to the directions provided by the manufacturer.
805
QADRI AND SMITH
Mycobacteria were isolated using conventional culture techniques with Lowenstein Jensen and Middlebrook 7H10 media, and also by direct inoculation of BACTEC 12B bottles run on a BACTEC 460 (Becton, Dickinson, Towson, Md .). Preliminary identification was performed using BACTEC p-nitro-a-amino-P-hydroxypropiophenone (Siddiqi 1989; Sommers and Good 1985), and the identification was confirmed by standard techniques (Kent and Kubica 1985; Sommers and Good 1985; Vestal 1975). Statistical analysis Sensitivity, specificity, and predictive values were calculated according to the method described by Galen (1986).
Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by Iowa State University on 11/10/14 For personal use only.
Results Serum samples from 227 individuals were tested for the presence of antibodies to mycobacterial antigen using the Anda A60 ELISA kit. Results were interpreted as follows as per manufacturer's recommendations: < 125 units, negative; 126- 199 units, equivocal; > 200 units, positive. Of the 227 subjects, 191 had culture-proven tuberculosis caused by M. tuberculosis. Serum samples from 73% of these patients gave positive results, ranging between 200 and 1600 units (Table I), and samples from four patients with atypical mycobacterial disease also yielded positive ELISA results. Sera from 12 patients whose specimens grew atypical mycobacteria, in the absence of clinical disease, remained negative. Of the atypical mycobacteria, 8 were Mycobacterium avium-intracellulare isolated from colon washings (I), urine (2), and sputum (5); 2 Mycobacterium terrae were isolated from sputum and 2 Mycobacterium gordoneae from fine-needle aspirate (FNA) (1) and sputum (I). Twenty patients with pulmonary lesions, where sputa were repeatedly culture negative, showed 55% false positives and 20% gave equivocal results. Serum samples from 84 normal healthy adults were also tested, half of them being PPD positive and the other half PPD negative. Of the PPD-positive individuals four showed 320 units and three 130 units. Sixteen of the 42 PPD-negative controls gave positive results, ranging between 200 and 680 units. None of the control group had any signs or symptoms or positive x-ray findings for mycobacterial disease. Of the 123 patients who had pulmonary disease, 82% showed the evidence of antibody presence to A60 antigen (Table 2). However, only 59% of the 68 patients with extrapulmonary disease yielded ELISA positive results.
Discussion Definitive diagnosis of mycobacterial disease by culture in extrapulmonary tuberculosis is difficult as a result of nonspecific signs and symptoms, slow clinical course, and relative inaccessibility of infected tissue, while in pulmonary disease it can be time consuming, especially when the patient does not expectorate the needed lo4 organisms/mL needed for microscopic detection (Kim et al. 1984). Although serological diagnosis is widely used for many infectious diseases, such tests for mycobacterial infections are confined to few research laboratories. Ivany et al. (1988) reported the use of murine monoclonal antibodies in developing immunodiagnostic assays for tuberculosis and leprosy. Recently, his group also evaluated the potential use of serology in extrapulmonary tuberculosis (Wilkins and Ivanyi 1990). They detected antibodies to the 38-kDa M. tuberculosis antigen in 73% of cases of extrapulmonary and 70% of smearnegative cases of pulmonary tuberculosis. Until recently such
TABLE1. Mycobacterial antibody detection by Anda A60 enzyme immunoassay (EIA) test EIA (n) Subject classification (S/C)
Positive
Equivocal
Culture positive M. tuberculosis (191)* M. aviurn-intracellulare (2)* M. avium-intracellulare (8) M. fortuitum-chelonei (2)* M. gordonae (2) M. terrae (2) Culture negative With pulmonary lesions (20) Normal healthy controls P P D positive (42) P P D negative (42) *Clinically proven disease.
tests were restricted to a few specialized research laboratories beause of the unavailability of commercial kits. We evaluated a recently introduced commercial ELISA kit that is designed for the detection of IgG antibodies in human serum against clinically significant mycobacteria. It uses an antigen prepared from the cytoplasm of Mycobacterium bovis BCG and referred to as A60 (Cocito et al. 1987). Probably the first agglutination test against tubercle bacilli was used by Arloing (1898); it had 57% sensitivity in patients with pulmonary TB, but 11% of controls also yielded positive reactions. Since then a number of crude, semipurified, and purified antigens have been used to determine antibody response to mycobacterial infections. Different investigators have reported sensitivities of 31-93% and specificities of 78-100% (Baelden et al. 1990; Benjamin and Daniel 1982; Daniel and Debanne 1987; Krambovitis et al. 1986; Levy et al. 1988; Thongkrajai et al. 1989; Turneer et al. 1988). Using in-house purified A60 antigen in an ELISA test, Baelden et a l . (1990) showed 100% negative sera in neonates and healthy adults and 6.4% false positives among 124 tuberculin-negative patients, mostly with pleural or bronchopulmonary disease other than tuberculosis. Eighty-three percent of their active postprimary and 43% of the active primary cases gave positive results. Although the sensitivity of the commercial test in our hands was similar to that of Baelden et al. (1990), we found that 24% of our normal, healthy adults, and 75% of our culture-negative patients with pulmonary lesions, gave false positive or equivocal results. Sensitivity of the test was 82070 in pulmonary tuberculosis (TB), but only 59% in the extrapulmonary disease, a form that is notoriously difficult to diagnose by culture. Furthermore, a negative skin test had no bearing on the rate of false positivity. We would like to mention that we followed the methodology according to the manufacturer's directions, as most diagnostic laboratories would, without any in-house modifications. Calculations, excluding equivocal results, gave a positive predictive value (PPV) of 89070, negative predictive value (NPV) of 59070, and an overall sensitivity of 80% and specificity of 65070, using the manufacturer's cutoff values. By raising the cuttoff to < 300 units for negative and > 300 for positive, a PPV of 90070, NPV of 29Y0, sensitivity of 68070, and specificity of 71 % were found. To exclude the PPD false positive controls, a cutoff of 700 units was used, which gave
CAN. J.
MICROBIOL. VOL. 38, 1992
TABLE 2. Sensitivity of Anda A60 test in clinical and culture-proven pulmonary and extrapulmonary TB EIA (n)
Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by Iowa State University on 11/10/14 For personal use only.
Sensitivity Specimen type
Patients (S/C)
Pulmonary FNAB* spine Cerebrospinal fluid Urine Ascitic fluid Hip Peritoneum Miscellaneous (blood, bone marrow, cervical lymph node, tissue)
123 14 9 10 6 5 6
Positive
Equivocal
Negative
("70)
18
*FNAB, fine needle aspiration biopsy.
a PPV of 90070, but the sensitivity dropped to 14%. Three patients with pulmonary lesions but repeated negative sputum cultures had > 700 units. Methods are being developed to measure antibody levels in specimens other than sera, such as cerebral spinal fluid and bronchial washings (Chandramuki et al. 1989; Levy et al. 1988), which when used in conjunction with serological tests may increase specificity. Sensitivity has also been increased by combining antigen detection with antibody measurements (Krambovitis et al. 1986). Our results suggest that the test by itself, in its present form, does not have acceptable predictive value in diagnosis of mycobacterial disease.
Acknowledgements The authors thank Ms. Linda M. Umali for secretarial assistance and Mrs. Marilyn Smith for the literature search. Arloing, S. 1898. Agglutination de bacille de la tuberculose vraie. C.R. Acad. Sci. 126: 1398-1400. Baelden, M.C., Vanderelst, B., Dieng, M., et al. 1990. Serological analysis of human tuberculosis by ELISA with mycobacterial antigen 60. Scand. J. Infect. Dis. 22: 63-73. Benjamin, R.G., and Daniel, T.M. 1982. Serodiagnosis of tuberculosis using the enzyme-linked immunoblast assay (ELISA) of antibody to Mycobacterium tuberculosis antigen 5. Am. Rev. Respir. Dis. 126: 1013-1016. Chandramuki, A., Bothamley, G.H., Brennan, P . J., and Ivanji, J. 1989. Level of antibody to defined antigens of Mycobacteriurn tuberculosis in tuberculous meningitis. J. Clin. Microbiol. 27: 82 1-825.
Cocito, C., Baelden, M.C., and Benoit, C. 1987. Immunological properties of antigen 60 of BCG. Scand. J. Immunol. 25: 579-585.
Daniel, T.M., and Debanne, S.M. 1987. The serodiagnosis of tuberculosis and other mycobacterial diseases by enzyme-linked immunosorbent assay. Am. Rev. Respir. Dis. 135: 1137-1151. Galen, R.S. 1986. Use of predictive value theory in clinical
immunology. In Manual of clinical laboratory immunology. 3rd ed. Edited by N.R. Rose, H. Freidman, and J.L. Fahey. American Society for Microbiology, Washington, D.C. pp. 966-970. Grange, J.M. 1984. The humoral immune response in tuberculosis: its nature, biological role and diagnostic usefulness. Adv. Tuberc. Res. 21: 1-78. Ivanyi, J., Bothamley, G.H., and Jackett, P.S. 1988. Immunodiagnostic assays for tuberculosis and leprosy. Br.Med. Bull. 44: 645-649.
Kent, P.T., and Kubica, G.P. 1985. Public health mycobacteriology. A guide for the level I11 laboratory. Centers for Disease Control, Atlanta, Ga. Kim, T.C., Blackman, R.S., Heatwole, K.M., et al. 1984. Acidfast bacilli in sputum smears of patients with tuberculosis. Am. Rev. Respir. Dis. 129: 264-268. Krambovitis, E., Harris, M., and Hughes, D.T.D. 1986. Improved serodiagnosis of tuberculosis using two assay test. J. Clin. Pathol. 39: 779-785.
Levy, H., Wadee, A.A., Feldman, C., and Rabson, A.B. 1988. Enzyme-linked immunosorbent assay for the detection of antibodies against Mycobacterium tuberculosis in bronchial washings and serum. Chest, 93: 762-766. Siddiqi, S.H. 1989. Bactec TB system, product and procedure manual. Becton Dickinson, Towson, Md. Sommers, H.M., and Good, R.C. 1985. Mycobacterium. In Manual of clinical microbiology. Edited by E.M. Lennette, A. Balows, W. J. Hausler, and H.G. Shadomy. American Society for Microbiology, Washington, D.C. pp. 2 16-248. Thongkrajai, P., Lulitanon, V., and Chamnanvanakit, C. 1989. Improved ELISA with immunoabsorbent-purified mycobacterial antigen for serodiagnosis of tuberculosis. J. Med. Microbiol. 39: 101-104.
Turneer, M., Van Vooren, J.P., De Bruyn, J., et al. 1988. Humoral immune response in human tuberculosis. J. Clin. Microbiol. 26: 1714-1719.
Vestal, A.L. 1975. Procedures for the isolation and identification of mycobacteria. Center for Disease Control, Atlanta, Ga. Wilkins, E.G.L., and Ivanyi, J. 1990. Potential value of serology for diagnosis of extrapulmonary tuberculosis. Lancet, 336: 641-644.
Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by Iowa State University on 11/10/14 For personal use only.
Microbiology Laboratories, King Faisal Specialist Hospital and Research Centre, P. 0. Box 3354, Riyadh 11211, Saudi Arabia Received October 23, 1991 Revision received January 27, 1992 Accepted February 26, 1992 QADRI,S. M. H., and SMITH,K. K. 1992. Nonspecificity of the Anda A60-tb ELISA test for serodiagnosis of mycobacterial disease. Can. J. Microbiol. 38: 804-806. The conventional methods for the laboratory diagnosis of tuberculosis and other mycobacterial diseases are time consuming and beyond the scope of most of the small and medium-sized hospital facilities. Therefore, there has been considerable interest in the development of a serological method for the detection of antibodies against mycobacteria. We recently evaluated a commercially available ELISA test (Anda Biologicals, Strasbourg, France) that measures antibody levels to A60 antigen, a membrane glycoprotein that is found in most mycobacteria. Of the 123 patients with positive pulmonary.cultures for Mycobacterium tuberculosis, 82% had detectable antibodies against the kit antigen. Of the 68 patients with extrapulmonary tuberculosis, 59% yielded positive results. Specimens from 2 of the 12 patients that grew Mycobacterium avium-intracellulare complex, and one each with Mycobacterium fortuitum and Mycobacterium chelonei, were considered significant on the basis of medical history and repeated isolation of the bacterium from clinical specimens, and these patients yielded positive serology. Of the healthy, normal PPD positive and PPD negative controls, 24% gave false positive results. Key words: Anda A60-tb ELISA,serodiagnosis of mycobacteria, mycobacterium. QADRI,S. M. H., et SMITH, K. K. 1992. Nonspecificity of the Anda A60-tb ELISA test for serodiagnosis of mycobacterial disease. Can. J . Microbiol. 38 : 804-806. Les methodes usuelles de laboratoire pour la diagnostic de la tuberculose et autres infections reliees aux mycobacteries sont longues et ne sont pas a la portee des hdpitaux de petite ou de moyenne taille. Cette situation explique le grand interet pour les methodes serologiques permettant la detection d'anticorps diriges contre les mycobacteries. Nous avons evalue une trousse ELISA commerciale (Anda Biologicals, Strasbourg, France) qui mesure les titres d'anticorps contre l'antigkne A60, une glycoproteine membranaire retrouvee chez la plupart des mycobacteries. La trousse a decele des anticorps chez 82% de 123 patients presentant des cultures pulmonaires positives a Mycobacterium tuberculosis. Chez 68 patients atteints de tuberculose extrapulmonaire, 59% ont donne des resultats positifs. Les prelkvements de 2 parmi 12 patients qui avaient une culture positive au groupe Mycobacterium avium-intracellulare et un positif a Mycobacterium fortuitum et un a Mycobacterium chelonei ont ete consideres comme significatifs d'aprks le dossier medical et l'isolement repet6 de la bacterie dans les prelkvements et ces patients avaient une serologic positive. Chez un groupe d'individus sains dont le PPD de contrale etait positif normal ou negatif, 24% ont donne des resultats serologiques faussement positifs. Mots cles : epreuve ELISA Anda A60-tb, serodiagnostic de la tuberculose, Mycobacterium. [Traduit par la redaction]
Introduction Mycobacterium tuberculosis and several other members of this genus are capable of causing human disease, producing slowly developing destructive granulomatous lesions, with or without caseation. Although the disease can occur in any part of the body, pulmonary manifestations are most common. The organisms can disseminate from lungs to other sites, where they cause extrapulmonary manifestations. Worldwide, nearly 8 million new cases and 3 million deaths result from tuberculosis each year. After a steady decline for several years, the disease is showing a slight increase in the United States, with 22 436 new cases recorded by the Centers for Disease Control, Atlanta, Ga., in 1988. Ability of mycobacteria to cause destructive lesions and to disseminate to other sites in the body necessitates rapid diagnosis for optimal patient care. Successful isolation of these bacteria is dependent upon collection of proper speci-
h his work was presented in part at the Annual Meeting of the American Society for Microbiology, Dallas, Texas, 1991. 2 ~ u t h oto r whom all correspondence should be addressed. Pr~nredIn Canada / lmprime au Canada
mens, is time consuming, and is beyond the scope of many cli"ica1 laboratories. Therefore, ~er010gicaldiagnosis Can be a viable alternative (Grange 1984; Ivany et 01. 1988). In this Paper, we report the evaluation of a commercially available EL1SA test that detects antibody levels to A60 antigen of mycobacteria.
Materials and methods Anda A60-tb test (Anda Biologicals, Strasbourg, France) was performed according to the manufacturer's directions. Briefly, the sera to be tested were diluted in specimen dilution buffer, distributed to the antigen A60 coated wells of microtiter plates, and incubated for 60 min in a 37°C water bath. The wells were washed 5 times with kit wash buffer, using an automatic plate washer (Titertek S8/12, Eflab, Finland), and blotted dry. Peroxidase labelled anti-human IgG was added, and the plates were incubated for 30 min at 37°C. The plates were washed and blotted dry, the substrate (tetramethylbenzene containing hydrogen peroxide) was added to the wells, and the plates were left at 37°C for 15 min. Color development was stopped using 0.5 M H2S04and the final color was read at 450 nm (Electronucleonics, Fairfield, N.J.). Results were calculated according to the directions provided by the manufacturer.
805
QADRI AND SMITH
Mycobacteria were isolated using conventional culture techniques with Lowenstein Jensen and Middlebrook 7H10 media, and also by direct inoculation of BACTEC 12B bottles run on a BACTEC 460 (Becton, Dickinson, Towson, Md .). Preliminary identification was performed using BACTEC p-nitro-a-amino-P-hydroxypropiophenone (Siddiqi 1989; Sommers and Good 1985), and the identification was confirmed by standard techniques (Kent and Kubica 1985; Sommers and Good 1985; Vestal 1975). Statistical analysis Sensitivity, specificity, and predictive values were calculated according to the method described by Galen (1986).
Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by Iowa State University on 11/10/14 For personal use only.
Results Serum samples from 227 individuals were tested for the presence of antibodies to mycobacterial antigen using the Anda A60 ELISA kit. Results were interpreted as follows as per manufacturer's recommendations: < 125 units, negative; 126- 199 units, equivocal; > 200 units, positive. Of the 227 subjects, 191 had culture-proven tuberculosis caused by M. tuberculosis. Serum samples from 73% of these patients gave positive results, ranging between 200 and 1600 units (Table I), and samples from four patients with atypical mycobacterial disease also yielded positive ELISA results. Sera from 12 patients whose specimens grew atypical mycobacteria, in the absence of clinical disease, remained negative. Of the atypical mycobacteria, 8 were Mycobacterium avium-intracellulare isolated from colon washings (I), urine (2), and sputum (5); 2 Mycobacterium terrae were isolated from sputum and 2 Mycobacterium gordoneae from fine-needle aspirate (FNA) (1) and sputum (I). Twenty patients with pulmonary lesions, where sputa were repeatedly culture negative, showed 55% false positives and 20% gave equivocal results. Serum samples from 84 normal healthy adults were also tested, half of them being PPD positive and the other half PPD negative. Of the PPD-positive individuals four showed 320 units and three 130 units. Sixteen of the 42 PPD-negative controls gave positive results, ranging between 200 and 680 units. None of the control group had any signs or symptoms or positive x-ray findings for mycobacterial disease. Of the 123 patients who had pulmonary disease, 82% showed the evidence of antibody presence to A60 antigen (Table 2). However, only 59% of the 68 patients with extrapulmonary disease yielded ELISA positive results.
Discussion Definitive diagnosis of mycobacterial disease by culture in extrapulmonary tuberculosis is difficult as a result of nonspecific signs and symptoms, slow clinical course, and relative inaccessibility of infected tissue, while in pulmonary disease it can be time consuming, especially when the patient does not expectorate the needed lo4 organisms/mL needed for microscopic detection (Kim et al. 1984). Although serological diagnosis is widely used for many infectious diseases, such tests for mycobacterial infections are confined to few research laboratories. Ivany et al. (1988) reported the use of murine monoclonal antibodies in developing immunodiagnostic assays for tuberculosis and leprosy. Recently, his group also evaluated the potential use of serology in extrapulmonary tuberculosis (Wilkins and Ivanyi 1990). They detected antibodies to the 38-kDa M. tuberculosis antigen in 73% of cases of extrapulmonary and 70% of smearnegative cases of pulmonary tuberculosis. Until recently such
TABLE1. Mycobacterial antibody detection by Anda A60 enzyme immunoassay (EIA) test EIA (n) Subject classification (S/C)
Positive
Equivocal
Culture positive M. tuberculosis (191)* M. aviurn-intracellulare (2)* M. avium-intracellulare (8) M. fortuitum-chelonei (2)* M. gordonae (2) M. terrae (2) Culture negative With pulmonary lesions (20) Normal healthy controls P P D positive (42) P P D negative (42) *Clinically proven disease.
tests were restricted to a few specialized research laboratories beause of the unavailability of commercial kits. We evaluated a recently introduced commercial ELISA kit that is designed for the detection of IgG antibodies in human serum against clinically significant mycobacteria. It uses an antigen prepared from the cytoplasm of Mycobacterium bovis BCG and referred to as A60 (Cocito et al. 1987). Probably the first agglutination test against tubercle bacilli was used by Arloing (1898); it had 57% sensitivity in patients with pulmonary TB, but 11% of controls also yielded positive reactions. Since then a number of crude, semipurified, and purified antigens have been used to determine antibody response to mycobacterial infections. Different investigators have reported sensitivities of 31-93% and specificities of 78-100% (Baelden et al. 1990; Benjamin and Daniel 1982; Daniel and Debanne 1987; Krambovitis et al. 1986; Levy et al. 1988; Thongkrajai et al. 1989; Turneer et al. 1988). Using in-house purified A60 antigen in an ELISA test, Baelden et a l . (1990) showed 100% negative sera in neonates and healthy adults and 6.4% false positives among 124 tuberculin-negative patients, mostly with pleural or bronchopulmonary disease other than tuberculosis. Eighty-three percent of their active postprimary and 43% of the active primary cases gave positive results. Although the sensitivity of the commercial test in our hands was similar to that of Baelden et al. (1990), we found that 24% of our normal, healthy adults, and 75% of our culture-negative patients with pulmonary lesions, gave false positive or equivocal results. Sensitivity of the test was 82070 in pulmonary tuberculosis (TB), but only 59% in the extrapulmonary disease, a form that is notoriously difficult to diagnose by culture. Furthermore, a negative skin test had no bearing on the rate of false positivity. We would like to mention that we followed the methodology according to the manufacturer's directions, as most diagnostic laboratories would, without any in-house modifications. Calculations, excluding equivocal results, gave a positive predictive value (PPV) of 89070, negative predictive value (NPV) of 59070, and an overall sensitivity of 80% and specificity of 65070, using the manufacturer's cutoff values. By raising the cuttoff to < 300 units for negative and > 300 for positive, a PPV of 90070, NPV of 29Y0, sensitivity of 68070, and specificity of 71 % were found. To exclude the PPD false positive controls, a cutoff of 700 units was used, which gave
CAN. J.
MICROBIOL. VOL. 38, 1992
TABLE 2. Sensitivity of Anda A60 test in clinical and culture-proven pulmonary and extrapulmonary TB EIA (n)
Can. J. Microbiol. Downloaded from www.nrcresearchpress.com by Iowa State University on 11/10/14 For personal use only.
Sensitivity Specimen type
Patients (S/C)
Pulmonary FNAB* spine Cerebrospinal fluid Urine Ascitic fluid Hip Peritoneum Miscellaneous (blood, bone marrow, cervical lymph node, tissue)
123 14 9 10 6 5 6
Positive
Equivocal
Negative
("70)
18
*FNAB, fine needle aspiration biopsy.
a PPV of 90070, but the sensitivity dropped to 14%. Three patients with pulmonary lesions but repeated negative sputum cultures had > 700 units. Methods are being developed to measure antibody levels in specimens other than sera, such as cerebral spinal fluid and bronchial washings (Chandramuki et al. 1989; Levy et al. 1988), which when used in conjunction with serological tests may increase specificity. Sensitivity has also been increased by combining antigen detection with antibody measurements (Krambovitis et al. 1986). Our results suggest that the test by itself, in its present form, does not have acceptable predictive value in diagnosis of mycobacterial disease.
Acknowledgements The authors thank Ms. Linda M. Umali for secretarial assistance and Mrs. Marilyn Smith for the literature search. Arloing, S. 1898. Agglutination de bacille de la tuberculose vraie. C.R. Acad. Sci. 126: 1398-1400. Baelden, M.C., Vanderelst, B., Dieng, M., et al. 1990. Serological analysis of human tuberculosis by ELISA with mycobacterial antigen 60. Scand. J. Infect. Dis. 22: 63-73. Benjamin, R.G., and Daniel, T.M. 1982. Serodiagnosis of tuberculosis using the enzyme-linked immunoblast assay (ELISA) of antibody to Mycobacterium tuberculosis antigen 5. Am. Rev. Respir. Dis. 126: 1013-1016. Chandramuki, A., Bothamley, G.H., Brennan, P . J., and Ivanji, J. 1989. Level of antibody to defined antigens of Mycobacteriurn tuberculosis in tuberculous meningitis. J. Clin. Microbiol. 27: 82 1-825.
Cocito, C., Baelden, M.C., and Benoit, C. 1987. Immunological properties of antigen 60 of BCG. Scand. J. Immunol. 25: 579-585.
Daniel, T.M., and Debanne, S.M. 1987. The serodiagnosis of tuberculosis and other mycobacterial diseases by enzyme-linked immunosorbent assay. Am. Rev. Respir. Dis. 135: 1137-1151. Galen, R.S. 1986. Use of predictive value theory in clinical
immunology. In Manual of clinical laboratory immunology. 3rd ed. Edited by N.R. Rose, H. Freidman, and J.L. Fahey. American Society for Microbiology, Washington, D.C. pp. 966-970. Grange, J.M. 1984. The humoral immune response in tuberculosis: its nature, biological role and diagnostic usefulness. Adv. Tuberc. Res. 21: 1-78. Ivanyi, J., Bothamley, G.H., and Jackett, P.S. 1988. Immunodiagnostic assays for tuberculosis and leprosy. Br.Med. Bull. 44: 645-649.
Kent, P.T., and Kubica, G.P. 1985. Public health mycobacteriology. A guide for the level I11 laboratory. Centers for Disease Control, Atlanta, Ga. Kim, T.C., Blackman, R.S., Heatwole, K.M., et al. 1984. Acidfast bacilli in sputum smears of patients with tuberculosis. Am. Rev. Respir. Dis. 129: 264-268. Krambovitis, E., Harris, M., and Hughes, D.T.D. 1986. Improved serodiagnosis of tuberculosis using two assay test. J. Clin. Pathol. 39: 779-785.
Levy, H., Wadee, A.A., Feldman, C., and Rabson, A.B. 1988. Enzyme-linked immunosorbent assay for the detection of antibodies against Mycobacterium tuberculosis in bronchial washings and serum. Chest, 93: 762-766. Siddiqi, S.H. 1989. Bactec TB system, product and procedure manual. Becton Dickinson, Towson, Md. Sommers, H.M., and Good, R.C. 1985. Mycobacterium. In Manual of clinical microbiology. Edited by E.M. Lennette, A. Balows, W. J. Hausler, and H.G. Shadomy. American Society for Microbiology, Washington, D.C. pp. 2 16-248. Thongkrajai, P., Lulitanon, V., and Chamnanvanakit, C. 1989. Improved ELISA with immunoabsorbent-purified mycobacterial antigen for serodiagnosis of tuberculosis. J. Med. Microbiol. 39: 101-104.
Turneer, M., Van Vooren, J.P., De Bruyn, J., et al. 1988. Humoral immune response in human tuberculosis. J. Clin. Microbiol. 26: 1714-1719.
Vestal, A.L. 1975. Procedures for the isolation and identification of mycobacteria. Center for Disease Control, Atlanta, Ga. Wilkins, E.G.L., and Ivanyi, J. 1990. Potential value of serology for diagnosis of extrapulmonary tuberculosis. Lancet, 336: 641-644.
