INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Mycobacterium avium-intracellulare (MAI) pulmonary infection may occur in subjects with no predisposing lung disease and no known immunodeficiency (1), showing the multiple small nodules and bronchiectasis in chest computed tomographic (CT) scans (4). There have remained some unresolved issues as to the pathogenesis of the disorder. Because MAI is widely distributed in the environment, especially in water sources (10, 11), it can infect anybody through inhalation or ingestion. The disorder occurs predominantly in older and nonsmoking women. Some patients show a deterioration in symptoms, radiographic appearance, and sputum bacteriology, whereas others remained unchanged or occasionally improved (1, 8). To clarify these problems, we examined whether immunogenetic susceptibility is present in the disorder. We performed human leukocyte antigen (HLA) typing for A, B, C, DR, and DQ alleles in patients and compared the results with those of control subjects.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

We identified 64 patients in whom MAI pulmonary disease was diagnosed, and who visited the Shinshu University Hospital or the medical institutions in Nagano Prefecture with complaints of respiratory symptoms and/or a further examination of chest radiographic abnormalities; the majority of the Shinshu cases were recently reported (8). As described previously (6), we applied the American Thoracic Society diagnostic criteria (12). All patients had abnormal findings on chest CT scans, suggesting MAI infection, namely centrilobular small nodules associated with bronchiectasis. MAI was cultured in sputum on at least two separate occasions or in samples obtained from the region using a sterilized fiberoptic bronchoscope (FBS). The patients had had no other distinct lung disease. None had any abnormal findings that suggested immunodeficiency, including human immunodeficiency virus (HIV) infection.

MAI Bacteriology

Expectorated sputum and samples obtained by FBS were examined by Ziehl-Neelsen stain and cultured for mycobacteria using 3% Ogawa egg medium and other bacteria and fungi by routine method. Mycobacteria were identified and differentiated by growth characteristics and conventional biochemical tests (4).

Informed Consent

We explained the features of the disorder to all patients in whom MAI disease was diagnosed. At least a 12-mo observation period is necessary without treatment. All patients gave written consent to participate in the HLA study after they had been informed of the nature of study.

Estimation of Outcome

We estimated the outcome of the patients based on chest CT scans during at least a 12-mo interval and bacteriology of sputum. The CT changes were interpreted by two chest radiologists who were blind as to the patients' clinical course. Two groups were identified: (1) a deteriorated group and (2) a not deteriorated group. In the former, CT findings showed progression of nodules and/or consolidation, and negative cultures for MAI became positive in sputum.

HLA Typing

We performed HLA typing, as described previously (13). Briefly, venous blood samples were obtained from MAI patients and control subjects. Lymphocytes were isolated from peripheral blood by a density gradient centrifugation technique (14), and B cells were separated with Lympho-Kwik reagent (One Lambda). The mononuclear cells were subjected to HLA-A, -B, and -C typing with a standard microcytotoxicity test (15), whereas HLA-DR and -DQ typing were performed on the B cells by a similar technique but with prolonged incubation times (16). Healthy control subjects for HLA typing were all born in and resided in Japan. The control group was age- and sex-matched with the patient group.

Statistics

Data in the text and tables are expressed as mean ± standard deviation (SD). Comparison of variables between the subgroups was made by Student's unpaired t test. HLA data were analyzed by the standard statistical procedure of ² contingency table analysis and Fisher exact test. The odds ratio (OR) was provided by the cross-product ratio (incidence ratio) of the entries in the 2 × 2 table (i.e., ad/bc). A value of p < 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients' Characteristics

Patients' profiles and data in the text and tables are retrospective compilations of those at the start of the observation. As shown in Tables 1 and 2, 64 patients with MAI pulmonary infection were divided into 37 (57.8%) not deteriorated and 27 (42.2%) deteriorated groups after a mean observation period of 30 ± 15 mo (from 12 to 50 mo). The mean ages of two groups were 66 ± 10 and 67 ± 11 yr, respectively. There were few smokers in either group; Patient 28 (47 pack-years) in the former, and Patient 3 (15 pack-years) or Patient 26 (44 pack-years) in the latter.

Most of the deteriorated patients complained of respiratory symptoms, and only five patients (18.5%) had no symptoms. Fourteen (51.9%) patients complaining of symptoms were positive for MAI in sputum. On the other hand, 18 (48.6%) of not deteriorated patients had no symptoms, 29 (78.3%) of them were positive for MAI in bronchoalveolar lavage (BAL) samples, and only eight (21.6%) were positive for MAI in sputum.

HLA Typing

Neither 37 not deteriorated nor 27 deteriorated patients had significant frequencies of HLA-B (B 13, 35, 44, 46, 48, 51, 52, 54, 55, 56, 60, 61, and 70) or -C (Cw 1, 3, 4, and 7) alleles. Tables 3 and 4 show the frequencies for the HLA-A, -DR, and -DQ alleles in not deteriorated and deteriorated patients. The two groups revealed completely different frequencies for HLA typing.

HLA-DR6 and -DQ4 were significantly more frequent in 37 not deteriorated patients than in 100 control subjects (Table 3). DR6 was positive in 14 (37.8%) patients but in only 16 (16.0%) control subjects (p = 0.0061, OR = 3.20). DQ-4 was positive in 10 (27.0%) patients but in only 10 (10.0%) control subjects (p = 0.0122, OR = 3.33). On the other hand, HLA-A26 was significantly more frequent in 27 deteriorated patients than in 100 control subjects (Table 4). A26 was positive in 14 (51.9%) patients but in only 21 (21.0%) control subjects (p = 0.0015, OR = 4.05).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The two most important results of the present study were that the disorder consisted of two distinct groups, not deteriorated and deteriorated, and that the two groups revealed different frequencies of the HLA typing.

Some studies described the outcome of the disorder after therapy, such as antituberculous drug(s), clarithromycin, ciprofloxacin, and surgical resection. Prince and coworkers (1) reported that among 19 patients, three died due to deteriorated disease and five showed relapse or recurrence, whereas 11 (57.9%) patients improved. Kennedy and Weber (3) reported that among 15 patients, two died and two did not improve, whereas 11 (73.4%) patients showed improvement or remission. Reich and Johnson (2) observed a favorable response to therapy in more than 90% of 16 patients. Recently, we (8) demonstrated that erythrocyte sedimentation rate, the serum levels of C-reactive protein and carbohydrate antigen 19-9, and the neutrophil cell counts in the BAL fluid were significantly elevated in the deteriorated MAI infection. In the present study, after the mean observation interval of 30 ± 15 mo, 27 of 64 patients (42.2%) showed deterioration in chest CT findings. They complained of respiratory symptoms, such as cough, sputum, bloody sputum, and chest pain, and showed more frequently positiveness for MAI in sputum. Therefore, their MAI infection clearly became deteriorated. On the other hand, 37 (57.8%) showed no deterioration during the observation period.

MAI is well known to be one of the most common pathogens in immunodeficiency, such as AIDS. However, it is unclear why healthy subjects, mainly elderly women, are infected with the opportunistic pathogen. Some hypotheses have been proposed to explain the development of pulmonary MAI disease in this population. The previous theories have focused on increased host susceptibility owing to altered anatomical defense, such as an unidentified systemic connective tissue disorder (17), or voluntary cough suppression causing bronchiectasis in dependent lobes predisposing to MAI colonization (18). However, it is reasonable to suppose that disease susceptibility could also be explained by an immune deficiency caused by alterations on a genetic rather than anatomical level.

A genetic basis for increased susceptibility to nontuberculous mycobacteria has been suggested in a report of familial childhood disease (19). Huang and coworkers (20) have tried to determine if pulmonary MAI disease, which occurs in a population of older and HIV-negative women without preexisting lung disease, happens in subjects with a defect in their natural resistance-associated macrophages protein (NRAMP1) or interferon- receptor 1 (IFN-R1) genes. There is no evidence of a genetic defect in NRAMP1 or IFN-R1 that correlates with the disease.

The present study is the first investigation seeking the HLA typing correlating with MAI pulmonary disease, showing the association with HLA-DR6 and -DQ4. Furthermore, the subjects with HLA-A26 antigen have a probability of deterioration of MAI infection. In tuberculosis, the associations have been found between class I HLA antigens A10 and B8 and class II antigen DR2, although these associations have not been consistently demonstrated and could account for only a small part of the significant component in tuberculosis susceptibility identified by the twin studies (21). We need further studies to detail the role of the genetic disorder in the susceptibility and deterioration of MAI infection in the lungs.