Published ahead of print on December 13, 2007, doi:10.1164/rccm.200705-771OC
© 2008 American Thoracic Society doi: 10.1164/rccm.200705-771OC
Serodiagnosis of Mycobacterium avium–Complex Pulmonary Disease Using an Enzyme Immunoassay Kit1 Department of Internal Medicine, National Hospital Organization (NHO) National Toneyama Hospital, Toyonaka-shi, Osaka, Japan; 2 Department of Immunology, National Institute of Infectious Diseases, Shinjuku-ku, Tokyo, Japan; 3 Department of Clinical Laboratory Medicine, Graduate School of Medicine, Kyoto University, Kyoto-shi, Kyoto, Japan; 4 Department of Internal Medicine, NHO Kinki-chuo Chest Medical Center, Sakai-shi, Osaka, Japan; 5 Department of Medicine, Osaka Prefectural Medical Center for Respiratory and Allergic Diseases, Habikino-shi, Osaka, Japan; 6 Department of Laboratory Medicine, Saitama Medical Center, Jichi Medical University, Saitama-shi, Saitama, Japan; and 7 Department of Internal Medicine, Sakamoto Hospital, Toyonaka-shi, Osaka, Japan Correspondence and requests for reprints should be addressed to Seigo Kitada, M.D., Department of Internal Medicine, National Hospital Organization National Toneyama Hospital, 5-1-1 Toneyama, Toyonaka-shi, Osaka 560-8552, Japan. E-mail: kitadas{at}toneyama.hosp.go.jp
Rationale: The diagnosis of Mycobacterium avium–complex pulmonary disease (MAC-PD) and/or its discrimination from pulmonary tuberculosis (TB) is sometimes complicated and time consuming. Objectives: We investigated in a six-institution multicenter study whether a serologic test based on an enzyme immunoassay (EIA) kit was useful for diagnosing MAC-PD and for distinguishing it from other lung diseases. Methods: An EIA kit detecting serum IgA antibody to glycopeptidolipid core antigen specific for MAC was developed. Antibody levels were measured in sera from 70 patients with MAC-PD, 18 with MAC contamination, 37 with pulmonary TB, 45 with other lung diseases, and 76 healthy subjects. Measurements and Main Results: Significantly higher serum IgA antibody levels were detected in patients with MAC-PD than in the other groups (P < 0.0001). Setting the cutoff point at 0.7 U/ml resulted in a sensitivity and specificity of the kit for diagnosing MAC-PD of 84.3 and 100%, respectively. Significantly higher antibody levels were also found in patients with nodular-bronchiectatic disease compared with fibrocavitary disease in MAC-PD (P < 0.05). There was a positive correlation between the extent of disease on chest computed tomography scans and the levels of antibody (r = 0.43, P < 0.05) in patients with MAC-PD. Conclusions: The EIA kit is useful for the rapid diagnosis of MAC-PD and for differentiating MAC-PD from pulmonary TB and, if validated by studies in other populations, could find wide application in clinical practice.
Key Words: nontuberculous mycobacteria • immunocompetence • sensitivity and specificity
The prevalence of disease due to nontuberculous mycobacteria has been increasing recently (1–5). In Japan, Mycobacterium avium complex (MAC) accounts for approximately 70% of nontuberculous mycobacterial disease (6). MAC is now widely recognized as an important pathogen that causes chronic and progressive pulmonary disease even in immunocompetent patients and not only in those who are immunosuppressed. The diagnosis of MAC-PD is complicated because, in contrast to Mycobacterium tuberculosis, MAC contamination of clinical specimens can come from environmental sources such as water, dust, and soil, and because this organism may colonize the respiratory tract without any accompanying invasive disease (4). Thus, isolation of MAC from sputa is often of no clinical significance. Diagnosis of pulmonary disease due to MAC is complicated and time consuming when made according to the guidelines of the American Thoracic Society (ATS) (1), because MAC is ubiquitous in nature and the diagnosis requires clinical findings and its repeated isolation from sputum. In addition, it is also difficult to discriminate MAC-PD from infection due to other mycobacteria in the absence of culture results, because clinical features, such as symptomatic or radiographic findings, are very similar in mycobacterial diseases. In the context of infection control, it is particularly important to distinguish between MAC-PD and pulmonary tuberculosis (TB). To overcome these difficulties, we have developed a serologic test for the glycopeptidolipid (GPL) antigen specific for MAC, and have reported its clinical usefulness (7–9). The levels of antibody to GPL core were measured by an enzyme immunoassay (EIA) using sera of immunocompetent patients with MAC-PD. MAC-PD could be discriminated from pulmonary TB, Mycobacterium kansasii pulmonary disease and MAC colonization/contamination using this serologic test. Healthy subjects were seronegative. Of the different immunoglobulin (Ig) subclasses, best results were obtained by the measurement of IgA, with a sensitivity of 92.5% and specificity of 95.1%. These results suggest that the test is useful as a diagnostic aid. In the present study, to apply this test widely in clinical practice, we developed an EIA kit detecting serum IgA antibody specific for GPL core and investigated its usefulness in a multicenter study.
See the online supplement for additional methodologic details.
Patients and Serum Samples Fibrocavitary disease was defined as the presence of cavitary forms in upper lobes. NBE disease was defined as the presence of bronchiectasis and multiple nodular shadows on CCT. Disease conforming to neither of these types was considered unclassifiable. Forty-five patients underwent CCT and serodiagnosis at the same time. A correlation between the extent of disease and antibody levels was investigated. The extent of disease was expressed as the number of MAC-involved CCT segments, as described in the previous study (9). The studies in human subjects were approved by the research and ethical committees of the NHO National Toneyama Hospital, and written, informed consent was obtained from all subjects.
EIA Kit
Statistical Analysis
Study Subjects The characteristics of the subjects are shown in Table 1. Patients with pulmonary TB and healthy subjects were younger than patients with MAC-PD (P < 0.001), and there was a larger proportion of females in the latter group (P < 0.001). Of the 70 patients with MAC-PD, 15 had underlying pulmonary disease, all of which were the sequelae of pulmonary TB. Of the 18 individuals with MAC contamination, 15 had underlying pulmonary diseases (8 patients with the sequelae of pulmonary TB, 2 with lung cancer, 2 with chronic obstructive pulmonary disease, 1 with emphysema, 1 with pneumoconiosis, and 1 with sarcoidosis. Of the patients with MAC-PD, 19 were classified as having fibrocavitary disease, and 35 as having NBE disease, with 16 patients unclassifiable. The MAC-PD group included infections with M. avium (n = 56), Mycobacterium intracellulare (n = 12), or both (n = 2). The MAC contamination group included M. avium (n = 16) and M. intracellulare (n = 2).
Level of GPL Core IgA Antibody The level of serum IgA antibody to GPL core was quantified using the EIA kit (Figure 1). As expected, patients with MAC-PD had significantly higher levels than patients with MAC contamination, those with pulmonary TB, those with other lung diseases, and healthy subjects—namely, 10.7 ± 7.9, 0.2 ± 0.1, 0.1 ± 0.1, 0.0 ± 0.1, and 0.0 ± 0.0 U/ml, respectively (P < 0.0001). A receiver operating characteristic (ROC) curve was constructed for MAC-PD and the other groups to establish the best cutoff value (Figure 2). Setting the cutoff value at 0.7 U/ml resulted in 100% specificity, at a sensitivity of 84.3% (Table E1). Using the EIA kit allowed clear discrimination between patients with MAC-PD and MAC contamination, pulmonary TB, and other lung diseases, as well as healthy subjects.
Next, we compared levels of serum IgA antibody to GPL core in fibrocavitary disease and NBE disease of MAC-PD. Significantly higher levels were found in NBE (P < 0.05) (Figure 3). With the cutoff value set at 0.7 U/ml, positivity in NBE and fibrocavitary disease was 91.4 and 63.2%, respectively. In contrast, in patients with MAC-PD, no significant differences between M. avium and M. intracellulare as causative agents were observed (P = 0.403). The erythrocyte sedimentation rate in MAC-PD was 32.6 ± 28.6 mm/hour and there was a significant positive correlation between the erythrocyte sedimentation rate and antibody levels in patients with MAC-PD (r = 0.294, P < 0.05).
Radiographic Severity and the Level of GPL Core Antibody Forty-five patients with MAC-PD (10 with fibrocavitary disease, 26 with NBE disease, and 9 with unclassifiable type disease) underwent CCT and serodiagnosis at the same time. Four patients with unclassifiable type disease were excluded from the investigation because it was hard to discriminate between MAC lesions and underlying pulmonary disease. There was a positive correlation between the extent of disease and the levels of the antibody (r = 0.43, P < 0.05) (Figure 4). The total numbers of involved segments were not different (7.8 ± 4.9 and 7.9 ± 4.2 in fibrocavitary and NBE disease, respectively). Of 26 patients with NBE disease, 9 had small thin wall cavities. A tendency toward elevated GPL core antibody levels was found in NBE patients with cavities compared with those without, but this trend was not statistically significant (P = 0.08).
We previously established a serologic test for MAC-PD using a mixture of GPLs and GPL core antigen, and reported the clinical application of the EIA method for quantifying antibody levels (7, 8). GPL is an antigen located on the surface of the MAC cell wall and determines the serotype. At present, 31 distinct serotype-specific GPLs have been identified, of which the complete structures of 14 have been identified (10–12). GPL consists of a core common to all MAC serotypes and a serotype-specific oligosaccharide. In the initial study to establish the serodiagnosis of MAC-PD, we used the whole GPL antigen, a mixture of 11 serotype-specific GPLs (7). We then found that the GPL core was the dominant antigenic epitope of GPL, and subsequently developed a serologic test using GPL core antigen (8). In the previous study, GPL core antibody (IgG, IgA, and IgM) levels were found to be elevated in sera of patients with MAC-PD, but not pulmonary TB, M. kansasii–PD, MAC colonization/contamination, and healthy subjects. The study showed that this serologic test was useful for diagnosing MAC-PD and for differentiating it from pulmonary TB and M. kansasii–PD. Consistent with this, Fujita and colleagues (13) reported elevated levels of antibody against the GPL core antigen in patients with MAC-PD but not in those with pulmonary TB. In our previous study (8), of the different Ig classes, best results were obtained by IgA, including an association with CCT findings. Thus, a higher level of serum IgA antibody to GPL core indicated a wider extent of MAC disease and larger nodule formation on CCT (9). Therefore, we have attempted to develop and to assess an EIA kit for quantifying serum IgA antibody to GPL core in the present study. Optical density levels were converted to U/ml using standard serum samples, which provided reliable and reproducible results. In this multicenter study, using the EIA kit, it was confirmed that patients with MAC-PD could be clearly differentiated from those with pulmonary TB, those with MAC contamination, those with other lung diseases, and healthy subjects. Similar to our previous studies (7–9), the sensitivity and specificity for diagnosing MAC-PD by the kit was high and the level of the antibody correlated with the extent of MAC-PD assessed using CCT. Distinguishing pulmonary TB from MAC-PD in clinical practice using the EIA kit has proven useful. Differentiating TB from MAC is difficult because symptoms and radiographic findings are often similar among patients with pulmonary mycobacterial diseases. Patients with pulmonary TB require immediate treatment and isolation, whereas the diagnosis of MAC-PD does not necessitate rapidly starting antimicrobial therapy (1), and isolation is not required. GPL antigens, which are major cell surface antigens of MAC, are not present in the cell wall of M. tuberculosis complex (11). On the basis of this observation, patients with TB do not produce anti-GPL antibody. Indeed most patients with TB did not possess serum antibodies against GPLs (Figure 1) (7, 8). However, we cannot exclude the possibility that disease in patients with TB was of too short duration (MAC-PD, 4.8 ± 4.6 yr, vs. TB, 0.3 ± 0.2 yr) to have allowed immune responses and shed mycobacterial antigen. In this present study, with a cutoff level of 0.7 U/ml, all patients with TB were classified as seronegative. The levels of GPL core antibody in patients with pulmonary TB were very low or absent (0.1 ± 0.1 U/ml). In contrast, in previous studies (7, 8, 13), GPL seropositivity in patients with pulmonary TB ranged between 5.2 and 25%. One possible explanation for this previously reported lack of specificity may be that there was latent coinfection of MAC in patients with pulmonary TB. In the present study, however, we attempted to exclude patients with such latent coinfection because the entry criteria precluded patients having underlying lung disease or past history of pulmonary TB. Patients with lung diseases such as chronic obstructive pulmonary disease associated with smoking, bronchiectasis, previous mycobacterial disease, cystic fibrosis, and pneumoconiosis are prone to have MAC coinfection (1). In addition, future studies are needed to verify the cutoff value obtained from the ROC analysis using another sample of cases and controls on a much larger scale. MAC-PD has recently been classified into two distinct subtypes: fibrocavitary disease and NBE disease (1). Fibrocavitary disease, the most common manifestation of MAC-PD, is usually seen in middle-aged or elderly men predisposed to lung disease due to smoking and alcohol drinking. This subtype of disease, generally progressive, is similar to pulmonary TB on chest radiography. If left untreated, it can lead to extensive lung destruction and death. In contrast, NBE disease is mostly seen in nonsmoking middle-aged or elderly women without predisposing lung disease. The clinical course is usually slower and less dramatic. Patients with NBE are presumed to have had a long subclinical period before appearance of disease manifestations. Significantly higher levels of GPL core antibody were seen in NBE than in fibrocavitary disease (P < 0.05) and higher seropositivity was found in patients with the former (91.4% compared with 63.2%). There were no significant differences of extent of disease between the two groups in patients who underwent CCT and serodiagnosis at the same time. Therefore, the results suggested the possibility that the antibody levels tend not to elevate in patients with fibrocavitary disease. This may reduce the utility of serodiagnosis for discriminating cavitary MAC from cavitary TB. However, the antibody would probably be present at high levels in patients with extensive lesions in fibrocavitary disease as was indeed found in three patients (17.9 ± 5.9 U/ml) who had extensive lesions (more than 13 segments) (Figure 4). Further investigations are required for confirmation of this notion in a larger study. Of the 70 patients with MAC-PD, 64 had previously received combination chemotherapy, as recommended by the ATS guidelines (1). However, all had MAC-positive cultures at the time of serum collection, and were considered to have active MAC-PD. Thus, antibody levels were not changed by the failure of chemotherapy—that is, there was no conversion to seronegative from seropositive status (8); therefore, effects of the previous treatment on antibody levels were limited. Obviously, it would nonetheless be better to enroll chemotherapy-naive patients from diverse ethnic and racial populations and different geographic areas in future studies. At present, the diagnosis of MAC-PD is usually made according to the ATS guidelines, which include clinical, radiographic, and microbiological criteria (1). The latter requires multiple positive cultures for MAC from sputum, a positive culture from bronchial lavage or a lung biopsy specimen, together with the other diagnostic features. Although it is easy to meet the criteria in advanced-stage MAC-PD, it is often difficult in early-stage disease. In clinical routine, it is impractical to obtain multiple sputum samples or perform bronchoscopy to obtain bronchial washings or lung tissue in all patients. It is also time consuming, because a long duration is required before the results of multiple cultures are available. There are several rapid methods for identification of MAC, but they have some limitations. The liquid culture–based system using radiometry and fluorometry allows the detection of mycobacterial growth at an early stage, fewer than 7 days for nontuberculous mycobacteria. However, limitations of this system include the inability to observe colony morphology, difficulty in recognizing mixed cultures, overgrowth by contaminations, cost, and radioisotope disposal. Rapid identification of MAC is also possible using DNA hybridization, nucleic acid amplification, or high-pressure liquid chromatography (1). The use of molecular biological technology has shortened the time required to identify mycobacteria from several weeks to as little as 1 day. The overall sensitivity for detecting MAC varies between 70 and 100%, with a specificity greater than 98%. However, the inability to distinguish live and dead organisms precludes nucleic acid amplification for definite diagnosis of active disease (14). The EIA kit is a rapid (within a few hours) and noninvasive assay with high sensitivity (84.3%) and specificity (100%) for diagnosing MAC-PD. Using the EIA kit, as reported here, MAC-PD could be efficiently differentiated from MAC contamination. "MAC contamination" defined in the present study was considered to represent contamination from the environment, because patients were asymptomatic and revealed no significant CCT findings indicating active mycobacterial disease. Most of those people classified into the MAC contamination group were so categorized based on a single positive MAC culture by chance during the follow-up period after completion of chemotherapy for pulmonary TB or at routine examination on admission for other diseases. It is difficult to be certain that MAC contamination, as defined here, does not indicate subclinical infection because no confirmatory pathology was obtained. However, if MAC contamination does reflect subclinical infection, it is of little clinical importance and does not mandate therapy. There were 15.7% false-negative EIA determinations in patients with MAC-PD. In such cases, diagnosis of MAC-PD should be made according to the ATS guidelines, as previously described. There are several possible explanations for these false-negative results, including the following: (1) recently diagnosed disease; (2) change of GPL core antigenicity after chemotherapy; or (3) diversity of immune responses to GPL core in individual patients, potentially governed by HLA genes (15). Therefore, it might be expected that not all patients with MAC-PD are capable of producing antibody to GPL core. Although the specificity determined here for the EIA kit was high, there remains also the possibility of false-positive results in patients with disease due to other mycobacteria, such as Mycobacterium fortuitum, Mycobacterium chelonae, Mycobacterium abscessus, and Mycobacterium scrofulaceum, because these organisms also possess GPL on their cell wall surface (10, 11, 16). Indeed, we have detected seropositivity in several patients with culture-positive M. fortuitum (data not shown). The incidence of pulmonary disease due to these other mycobacteria is relatively low (<5%) in Japan and the United States (6, 17), but a report from South Korea documented a high incidence of pulmonary infection by M. abscessus or M. fortuitum (33 and 11%, respectively (18). Therefore, caution is necessary when interpreting the results of the EIA kit in locations where other mycobacterial infections are endemic. A recent study using high-resolution CT documented that characteristic findings with multiple small nodular shadows combined with bronchiectasis are predictive for culture-positive MAC with a relatively high probability. Swenson and colleagues (19) reported that, of 15 patients with these characteristic findings, 8 (53%) had cultures positive for MAC. Tanaka and coworkers (20) reported that, of 26 similar patients, 13 (50%) had positive cultures for MAC in bronchial washings. Therefore, combining positive results obtained by the EIA and the characteristic findings of high-resolution CT should yield a definitive diagnosis of MAC-PD even in patients with sputum culture–negative results for MAC. This approach may be useful especially in elderly patients with complications, in whom bronchoscopy cannot be performed. In summary, the EIA kit for detection of serum IgA antibody specific for GPL core antigen is useful for rapid and accurate serodiagnosis of MAC-PD. Taken together with clinical, radiographic, and microbiological criteria, the kit may be a valuable tool for the diagnosis of MAC-PD. Validation of the EIA kit in the diagnosis of MAC-PD requires a larger controlled study in diverse populations.
Supported by grants from the Ministry of Health, Labor, and Welfare (Research on Emerging and Re-emerging Infectious Diseases, Health Sciences research grants); the Ministry of Education, Culture, Sports, Science, and Technology; Tauns Laboratory, Inc.; and the Osaka Tuberculosis Research Foundation. This article contains an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org Originally Published in Press as DOI: 10.1164/rccm.200705-771OC on December 13, 2007 Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form May 25, 2007; accepted in final form December 13, 2007
Related articles in AJRCCM:
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
AT A GLANCE COMMENTARY
