Chest Wall Tuberculosis After Bacillus Calmette–Guerin Vaccination

To the Editors: A 16-month-old girl was referred to our hospital with a recently enlarging chest wall tumor. She was not vulnerable to infection and neither she nor her family had contact with anyone diagnosed with tuberculosis. She had received a bacillus Calmette–Guérin (BCG) vaccine (Mycobacterium bovis BCG Tokyo-172 strain) administered percutaneously at the left arm using the multipuncture method at age 2 days in Bangkok, Thailand. On presentation, she was afebrile and had no respiratory symptoms. Physical examination revealed a soft mass (3 × 3 cm) with a clear margin and no discoloration on the right sternum. There was no lymphadenopathy. Laboratory data showed no findings suggesting infection and immunological data were within normal range. Chest magnetic resonance imaging showed a high-intensity dumbbell-like mass lesion (Fig. 1A), suggesting malignancy such as Ewing’s sarcoma. Chest computed tomography showed a thick peripherally enhancing loculated mass-like abscess with osteolysis of the sternum (Fig. 1B). Fine-needle aspiration biopsy from chest abscess revealed cytologic and histopathologic findings negative for malignancy. Microbiological tests, including polymerase chain reaction (PCR) tests for tuberculosis and nontuberculosis mycobacteria, were also negative. She was observed without treatment because she was doing well and because no malignancy or pathogen was detected. However, culture of the needle aspirate sample for tuberculosis was continued.FIGURE 1.: Fat-suppressed T2-weighted MRI showing a high-intensity dumbbell-like mass lesion (24 × 24 × 22 mm) with peripheral enhancement (A). Chest contrast CT showing a thick peripherally enhancing mass-like abscess with osteolysis of the sternum (B). CT indicates computed tomography; MRI, magnetic resonance imaging.One month later, the culture of the needle aspirate sample became positive for acid-fast bacillus and a PCR test was positive for M. tuberculosis, but negative for nontuberculosis mycobacteria. Further genetic analysis using the RD1 region identified this isolate as M. bovis BCG Tokyo-172 strain.1 As BCG vaccine used in Thailand is M. bovis BCG Tokyo-172 strain, a diagnosis of abscess with osteomyelitis caused by M. bovis via BCG vaccination was made. The isolate was susceptible to isoniazid, rifampicin, ethambutol, streptomycin and kanamycin. She was transferred to another hospital to receive treatment for tuberculosis. BCG vaccine is used worldwide for routine vaccination against tuberculosis. Our patient received a BCG vaccine containing the Tokyo-172 strain of M. bovis, and this vaccine was recently reported to show a low rate of unfavorable side effects (0.02%) in Taiwan.2 Tuberculosis of the chest wall constitutes 1%–2% of all tuberculosis cases.3 Based on the incidence of BCG vaccine complications and chest wall tuberculosis, cases of chest wall tuberculosis caused by BCG vaccination are extremely rare. The median interval between injection and onset of symptoms of BCG osteomyelitis was reported to be 11–13.9 months4,5 and the incidence of BCG-related osteomyelitis is reported to be higher when children are vaccinated earlier,2 as in our patient. We speculate that the number of M. bovis cells in our patient was too low to detect the tuberculosis by PCR testing at biopsy, but the 1-month culture enabled M. bovis to proliferate and become detectable. The negative results of cytopathological and microbiological tests presented a diagnostic dilemma, but the culture results ultimately led to an accurate diagnosis. In conclusion, pediatricians should be aware that M. bovis BCG vaccination can cause chest wall tumors as a potential complication. Culture for tuberculosis should be performed when tuberculosis is suspected. ACKNOWLEDGMENTS The authors thank Drs. Saori Oguri and Tomoo Miyakawa (Department of Respiratory Disease, Tokyo Metropolitan Children’s Medical Center) for providing tuberculosis treatment to the patient.