Evidence of endogenous reactivation of tuberculosis after a long period of latency.