PER 5/30/95 AMENDMENT: In HIV-negative patients, intermittent anti-TB therapy has been shown to be as effective as daily therapy, but the optimal duration of therapy in HIV-infected patients has not been established.

ORIGINAL: In some areas of the country, resistance to one or more of the drugs commonly used to treat TB has emerged. Thus, the need to test regimens containing a new drug exists. Furthermore, the optimal duration of anti-TB therapy for HIV-infected patients with TB needs to be determined.

PER 5/30/95 AMENDMENT: Patients who have received an acceptable induction regimen prior to study entry and have been found to be susceptible to isoniazid and rifampin with no pyrazinamide resistance are randomized to receive either isoniazid or rifampin plus vitamin B6 biweekly for 18 or 31 weeks. Patients are evaluated at months 1, 2, 4, 6, 8, and 10, and every 4 months thereafter. Minimum follow-up is 1.5 years.

ORIGINAL: In the induction phase, patients enrolled in "drug-susceptible" areas (defined as metropolitan areas with a resistance rate for isoniazid therapy of less than 10 percent) receive four drugs: isoniazid (plus pyridoxine), rifampin, pyrazinamide, and ethambutol. Patients enrolled in "drug-resistant" areas (resistance rate for isoniazid of 10 percent or higher) receive the four-drug regimen with or without a fifth drug, levofloxacin. After 8 weeks of induction, patients with multi-drug resistance are removed from study regimens; all other patients enter a continuation phase. Pansusceptible patients (showing susceptibility to all first-line anti-TB drugs) receive two study drugs for an additional 18 or 31 weeks; patients with isoniazid-resistant (or intolerant) TB receive two or three study drugs for an additional 44 weeks, while those with rifampin-resistant TB receive two or three study drugs for an additional 70 weeks. Patients are evaluated every 2 weeks in the induction phase and every 12 weeks in the continuation phase. Minimum follow-up is 2 years.