Abstract Body

Background: Bedaquiline (B) and PA-824 (PA) have demonstrated potent bactericidal and sterilizing activity in combination with clofazimine (C) and pyrazinamide (Z) in murine models of TB. C is currently in use for multi-drug resistant TB, but its efficacy has not been established in controlled clinical trials. This study evaluated the 14 day early bactericidal activity (EBA) of C and Z alone and in combination with B and PA.

Methodology: This prospective, randomized, open label trial randomized subjects with newly diagnosed smear positive pulmonary DS-TB into 7 parallel treatment arms of equal size: 1) Isoniazid, rifampicin, pyrazinamide and ethambutol (H-R-Z-E); 2) C; 3) Z; 4) PA-B-Z; 5) PA-B-Z-C; 6) PA-B-C; 7) B-Z-C. C was administered as 300 mg/day x 3d, then 100 mg/d X 11d. Subjects were evaluated through daily overnight sputum collections. Sputum was cultured on solid agar (CFU) and in liquid culture (TTP). The primary endpoint was the EBACFU(0-14) as determined by the rate of change in logCFU per ml sputum over days 0 – 14 and the secondary endpoint was the rate of change in log TTP over days 0-14, both analyzed by a Joint Bayesian Non-linear Mixed Effects Regression Model (JB-NLME). Minimum inhibitory concentrations (MIC) for B, PA and C were determined by microtiter assay on isolates from baseline sputum samples. The time over MIC (TMIC) was determined using day 14 pharmacokinetic (PK) data with and without correction for plasma protein binding.

Results: 105 subjects (11 HIV+) were enrolled. H-R-Z-E had bactericidal activity similar to prior studies, with an EBACFU(0-14)/day of 0.151, 95% CI [0.070; 0.231]. The PA-B-Z combination had an EBACFU(0-14)/day of 0.167, 95% CI [0.078; 0.256]. C had no EBA by CFU or TTP and did not add to the bactericidal activity of the studied combinations. The MIC90 of the isolates from baseline were similar to prior reports, with B = 0.029 ug/mL in all arms, PA = 0.06-0.125 ug/ml and C = 0.125 – 0.25 ug/ml. The TMIC, calculated both with and without taking into account plasma protein binding for B and PA was 100% for the majority of patients. The TMIC for C, calculated without taking into account plasma protein binding was 100%, but taking into account plasma protein binding was 0%, for the majority of patients.

Conclusions: The PA-B-Z combination had robust EBA and will be taken in to a trial of longer duration. C did not have demonstrable EBA alone or in combination over 14 days of therapy. These results are consistent with the finding that the TMIC was nearly 100% for all subjects across arms for B and PA, but was 0% for most subjects for C, when plasma protein binding was accounted for.