[Two cases of pulmonary infection by Mycobacterium abscessus in which drug susceptibility testing results conflicted with clinical courses].

We encountered two cases of pulmonary infection by Mycobacterium abscessus (M. abscessus). [Case 1] A 66-year-old man who had been treated for non-tuberculous mycobacterium in the past was admitted because of productive cough. His chest X-ray film showed cavitation and direct microscopy of sputum revealed positive acid-fast bacilli (AFB). He was given rifampicin (RFP), ethambutol (EB), and clarythromycin (CAM), and then his symptoms and radiographic findings improved. [Case 2] A 74-year-old man with multiple myeloma as an underlying disease was admitted because of a cavitation found on chest radiography and a positive result for AFB in his sputum. Standard antituberculous drug therapy with isoniazid (INH), RFP, EB, and pyradinamide (PZA) was initiated and then the chest radiographic findings improved. As M. abscessus was isolated two weeks after the induction of therapy, the therapeutic regimen was changed to another combination therapy consisting of EB, clarithromycin (CAM) and ciprofloxacin (CPFX), and then his symptoms and radiographic findings were further improved. In both cases, the bacilli found in their sputum were identified as M. abscessus by DNA hybridization. They were completely resistant to all anti-tuberculosis agents and many antibiotics with a high value of MIC. However, their symptoms, radiographic abnormalities and the results of sputum examination improved following chemotherapy. The results obtained by MIC measurement were inconsistent with the clinical outcomes. The measurement of the MIC value of antibiotics do not necessarily predict its therapeutic effect.

[A case of successful desensitization therapy for isoniazid-induced pneumonitis].

A 65-year-old man was admitted to hospital for treatment of pulmonary tuberculosis. He was treated with isoniazid (INH), rifampicin (RFP), ethambutol (EB), and pyrazinamide (PZA). On the 14th day, he developed a fever and interstitial pneumonia, which improved promptly after discontinuation of the antituberculous drugs. Drug lymphocyte stimulation tests against INH, RFP and PZA were negative. However, the provocation test on INH (only) was positive, leading to a diagnosis of pneumonitis caused by INH. We then tried desensitization of INH over a period of two weeks, which was successful and occurred without any clinical event. In the past, five cases of INH-induced pneumonitis were reported, but desensitization of INH did not occur in any. We conclude that physicians should be aware not only of paradoxical reactions but also of drug-induced pneumonitis when a new pulmonary infiltrate develops in the course of tuberculosis treatment. Furthermore, drug desensitization may be possible in some cases of drug-induced pneumonitis.