In January 1999, an HIV-seronegative 53-year-old man was hospitalized with a solid lesion in the upper lobe of the right lung, first identified in 1989. Computer tomography (CT) of the chest showed two dense infiltrates, 3x2 cm and 6x5 cm, without cavitation. The patient reported nonproductive cough and weight loss of 3 kg in 2 years. He was a cigarette smoker (approximately 1 pack/day for 30 years). Chronic abuse of alcohol began in 1992, after the initial diagnosis. An elevated erythrocyte sedimentation rate was the only abnormal laboratory finding. A tuberculin skin test (5 units of PPD) had a 7-mm induration. Sputum and bronchoalveolar lavage fluid were smear negative for acid-fast bacilli (AFB). Tuberculous granuloma were found in bronchial and peribronchial lung tissue obtained by bronchoscopy; however, no AFB or signs of malignancy were detected. Anti-TB therapy was started with isoniazid, rifampin, and pyrazinamide. The patient's clinical condition improved, and by November 1999 the pulmonary infiltrate had almost completely resolved. In interviews, the patient said that he had no extended exposure to household or farm animals. He lives in his own apartment in a town (population 14,000) in central Germany. No tuberculin skin test-positive contact persons have been identified.

Figure 1

Figure 1. Spoligotype patterns of the isolates obtained from patient 1 (lane 1) and patient 2 (lane 2) and control strain H37Rv (lane 3).

One mycobacterial culture was grown from bronchoalveolar lavage fluid in liquid media (12B vial, BACTEC460TB, Becton Dickinson Microbiology Systems, Cockeysville, MD, USA) after 5 weeks of incubation. Limited growth was observed on Stonebrink medium (small colonies visible after >6 weeks of incubation), and no growth was detected on Loewenstein-Jensen medium. The isolate was identified as M. tuberculosis complex by gene probes (ACCUProbe, GenProbe, San Diego, USA) and as llama-type M. microti by the characteristic spoligotype pattern (Figure 1) (1). After reculturing, the strain showed normal growth in BACTEC 460TB (Becton Dickinson), allowing drug-susceptibility testing that indicated susceptibility to isoniazid, rifampin, pyrazinamide, ethambutol, and streptomycin. Because of the limited growth on solid media, classical biochemical tests could not be performed. The bacteria showed normal cell morphologic features in several Ziehl-Neelsen-stained microscopic preparations from cultures.

A severely ill 58-year-old man sought medical attention in April 1999. The patient was in poor general condition, with weight loss (> 8 kg in the last year), night sweats, occasional fever, fatigue, and productive cough. He had a history of smoking (10 cigarettes/day for at least 20 years) and diabetes mellitus (diagnosed in 1980). A chest X-ray in December 1983 showed no abnormalities of the heart and lung. The current illness had onset in early 1998, after which the patient was examined intensively by a number of physicians. Several X-rays of the chest indicated pulmonary disease with progression to bilateral lesions, predominantly in the right lung (Figure 2a); however, diagnosis was inconclusive. In September 1998, the general condition of the patient had so deteriorated that he was unable to work. Cytopathologic examination of a hemorrhagic pleural effusion in October 1998 showed lymphocytic-granulocytic pleuritis. X-ray examination (Figure 2b) and CT scan (data not shown) in April 1999 showed bilateral fibrosis of the lung parenchyma with a partial honeycomb pattern. Comparison with the X-ray in early 1998 documented a clear progression of the changes in the right lung.

Laboratory values were normal except for a slightly elevated white blood cell count. A tuberculin skin test (Mendel-Mantoux) and tests for HIV-1 and HIV-2 were negative. Fiber-optic bronchoscopy showed chronic atrophic bronchitis without signs of malignancy. For the first time, a bronchial secretion specimen was smear positive for AFB, and severe pulmonary TB was diagnosed. After treatment was begun with isoniazid, ethambutol, rifampin, streptomycin, and pyrazinamide, the patient's clinical condition improved, he recovered his normal weight, and the clinical symptoms resolved. Despite restrictive alterations in ventilatory function that correlated with radiologic and histomorphologic findings, the patient's condition improved and he has returned to work. An X-ray in January 2000 showed a consistent picture of the fibrosing lung disease.

One liquid culture (MGIT, Becton Dickinson) from a gastric aspiration obtained in April 1999 was positive for Mycobacteria in mid-June after 9 weeks of incubation. M. tuberculosis complex-specific gene probes (ACCUProbe) were positive. By the spoligotyping method, the isolate was unequivocally identified as vole-type M. microti (Figure 1). In contrast to the llama-type isolate, even after reculturing, the vole isolate showed very slow growth in liquid media and no growth on solid media, making drug-susceptibility testing and biochemical characterization impossible. In the Ziehl-Neelsen-stained microscopic preparation of both the bronchial secretion and several cultures, the bacterial cells showed normal morphologic features. In subsequent interviews, the patient said that he had not been exposed to household or farm animals. He lives in a village in a rural area in southern Germany. No tuberculin skin test-positive contact persons have been identified.

We describe the first two cases of pulmonary TB due to M. microti of the llama and vole types, respectively, in German patients. M. microti caused severe pulmonary TB with extensive histopathologic changes in the lung of an HIV-seronegative immunocompetent patient. As in the case reported by van Soolingen et al. (1), our data confirm the potential of M. microti to cause clinical illness in immunocompetent patients.

In contrast to the case reported by Foudraine and co-workers (5), both patients responded well to treatment with isoniazid, rifampin, and pyrazinamide or isoniazid, ethambutol, rifampin, streptomycin, and pyrazinamide, which indicates that regular anti-TB therapy may be appropriate for patients with M. microti infection.

As reported previously (1), primary isolation and biochemical characterization of M. microti are complicated by the slow growth of the bacteria. The resultant diagnostic delay, especially if atypical pulmonary lesions are present, may lead to false diagnosis and inadequate treatment. Thus, infection with M. microti should be considered in the differential diagnosis of patients with typical signs of TB but no bacteriologic proof. In these cases, spoligotyping is a fast, reliable technique for identification and differentiation of M. microti and other members of the M. tuberculosis complex (1,6,7).

The vole-type strain showed only limited growth during primary isolation and reculturing, impeding further characterization. In contrast, our results indicate that the llama-type M. microti may show normal growth in BACTEC 460TB. Moreover, we demonstrated that drug-susceptibility testing of clinical isolates of this subtype may be performed in liquid media.

Transmission from rodents to humans has been discussed as a possible pathway for the acquisition of M. microti infection (1). Although patients were interviewed, no source of infection could be identified.

Because of the difficulty with primary isolation and differentiation, the prevalence and clinical importance of M. microti may have been underestimated. Further studies based on molecular methods are needed to characterize the epidemiology of this emerging human pathogen.