Abstract
Tuberculosis (TB) relapse after appropriate drug treatment is poorly understood but critical to developing shorter treatment regimens. Using a cynomolgus macaque model of human TB, macaques with active TB disease were treated with a short course of isoniazid and rifampin and subsequently infected with SIV. Serial clinical, microbiologic, immunologic, and position emission and computed tomography (PET CT) assessments were performed to identify risk factors of relapse. Of the 12 animals, eight developed radiologically defined relapse, including four that had clinical and/or microbiologic signs. Greater gross pathology and bacterial burden were observed in relapse animals. PET CT characteristics before, during, and at the end of the treatment were similar among relapse and non-relapse animals. We show that complete sterilization or very low Mtb burden is protective against SIV-induced TB relapse but cannot be predicted by PET CT. Using barcoded M. tuberculosis, we found that Mtb dissemination during relapse originated from both lung and thoracic lymph nodes, underscoring the importance of lymph nodes as a reservoir. By matching barcoded Mtb and serial PET CT, we also demonstrate that not every site of persistent Mtb growth after drug treatment is capable of dissemination and relapse, underscoring the complex nature of drug treatment and relapse.