Pathogenesis of tuberculosis: the 1930 Lübeck disaster revisited.

During the 1930 Lübeck Mycobacterium bovis bacille Calmette-Guérin (BCG) disaster, 251 neonates received three oral BCG doses accidentally contaminated by virulent Mycobacterium tuberculosis; 67 (26.7%) infants died of tuberculosis. BCG reversion to pathogenicity did not occur. Detailed post mortem examinations clarified contested aspects of tuberculosis pathogenesis. Gastrointestinal infection was seldom "silent" and did not cause typical primary pulmonary lesions. In 15 infants, primary pulmonary foci were found but these resulted from vaccine ingestion and aspiration and were not caused by gastrointestinal infection spreading to the lungs without trace of its journey, as claimed by prominent researchers such as Calmette and von Behring. Further, among 60 infants in whom post mortem evaluation was completed, a "silent" gastrointestinal infection without an intestinal primary focus was found in only one. Lymphohaematogenous-disseminated tuberculosis caused death in 24/67 (35.8%) infants and tuberculous meningitis in a further 17/67 (25.4%). Gastrointestinal tuberculosis complications caused death in 26/67 (38.8%) infants. Half of the tuberculosis-attributed deaths had occurred by 3 months, 93% by 6 months and 100% by 12 months; remarkably no further deaths or tuberculosis recurrences occurred within 5 years post-vaccination/infection. These findings provide graphic confirmation that the early introduction of chemoprophylaxis in recently M. tuberculosis-infected young children is critical and urgent.