Evidence of Endogenous Reactivation of Tuberculosis after a Long Period of Latency

To the Editor —Many clinicians will have diagnosed tuberculosis (TB) in patients who did not belong to any of the risk groups for TB (e.g., prison inmates, immigrants originating from a country with high TB rates, or persons with human immunodeficiency virus infection). The intuition that many cases in which there is no history of recent exposure to a patient with smear-positive pulmonary TB, especially those that occur in elderly persons, represent reactivation of TB acquired many years previously is probably correct. However, this cannot be proved unless both the causative strain and the presumed source strain (from previous infection) are available and genotypically identical and unless recent transmission can be excluded with some certainty In a recent issue of the Journal Lillebaek and colleagues [1] describe molecular and epidemiological evidence of reactivation TB after 33 years of latency. This interesting finding was made possible by the centralized TB notification system and microbiological testing and collection of strains in Denmark and by the fact that all strains have been genotyped since 1993. The authors believe their report to represent the first instance of molecular evidence of endogenous reactivation of TB in humans. However, we know of at least 2 previous reports with similar findings. We recently described a woman who had negative results of multiple tuberculin skin tests after she had been exposed to an individual with smear-positive TB but who developed active TB 7 years later [2]. It was made certain, by molecular epidemiologic evidence, that the infection had been contracted 7 years previously from the original source case. A similar case report has been published in which spoligotyping was used to demonstrate strain identity [3]. The Netherlands and Denmark are similar with regard to centralized TB notification and the existence of a mycobacterial reference laboratory that collects and genotypes all strains The following description further illustrates the length of the interval that can exist between infection and TB disease: A 65-year-old man presented in 1998 with motor and speech deficits. Computed tomography and magnetic resonance imaging revealed a large necrotizing mass in the left parietal lobe, which was suspected to be a malignancy. During an operation to remove the mass, the lesion was found to be an abscess, and a few gram-positive rods were seen in the pus. These were identified by acid-fast staining as Mycobacterium and a polymerase chain reaction of the pus using a probe for Mycobacterium tuberculosis complex had positive results. The culture yielded M. bovis. A chest radiograph revealed no abnormalities, and an evaluation of the abdomen was not performed, because the patient did not have abdominal complaints. The patient recovered completely during a 9-month course of chemotherapy. On questioning, he reported having drunk raw, unpasteurized milk when he stayed at a farm during World War II. At that time, bovine TB had not yet been cleared from Dutch cattle (this was first accomplished in the early 1950s). The patient had never had any symptoms compatible with TB and did not remember any previous skin-testing. He had not been vaccinated with bacille Calmette-Guérin. Between 1970 and 1998, he had made a few short business trips to Africa and Mexico, where he stayed in high-standard hotels, was never exposed to cattle, and did not consume unpasteurized dairy products. Thus, although molecular evidence is lacking in this case, we believe that this patient experienced reactivation TB, presenting with a brain abscess caused by M. bovis after a latency period of at least 53 years. Who will set the next record?