FDA Defends ‘Paradoxical’ Approval Of J&J’s TB Drug Sirturo In NEJM

Imbalance of deaths for bedaquiline patients despite efficacy on surrogate marker in Phase II study supporting accelerated approval is still unexplained with longer-term data. But with efficacy also still looking good, FDA defends approval for select population with high unmet need.

Longer-term Phase II data for Johnson & Johnson’s TB drug Sirturo (bedaquiline) continue to support accelerated approval and use in the very select population for whom the drug is approved, despite still unanswered questions about excess deaths in a key study, FDA officials suggest in an Aug. 21 editorial in the New England Journal of Medicine.

Sirturo received accelerated approval at the end of 2012 for multi-drug resistant pulmonary tuberculosis in patients for whom an effective treatment regimen is not available unless bedaquiline is included. Labeling specifies that it is not indicated for the treatment of latent, extra-pulmonary or drug-sensitive tuberculosis or for the treatment of infections caused by non-tuberculous mycobacteria.

Sirturo was the first new TB drug with a new mechanism of action in 40 years, a clearance supported primarily on a Phase II study that tested the drug against placebo on top of a multi-drug background regimen. At the time of the regulatory review, the agency had data for up to 72 weeks of treatment in about 160 patients (Also see "Janssen’s Sirturo: TB Milestone Awaits After Advisory Committee Hurdle" - Pink Sheet, 26 Nov, 2012.).

A higher number of deaths for those on bedaquiline had been a concern (10 for treatment arm versus two for placebo), but the approval came through based on the much better efficacy in the surrogate marker of conversion of sputum cultures from positive to negative (Also see "J&J Bedaquline and the Evolving Role for Accelerated Approval: Confirming Safety, Not Efficacy?" - Pink Sheet, 4 Dec, 2012.). The drug also has a unique mechanism of action, which has translated into efficacy even in strains resistant to other therapies.

Andreas Diacon of the Stellenbosch University in Cape Town, South Africa, and colleagues reported 120-week results for Sirturo in the NEJM on Aug. 21. The longer-term results uphold efficacy based on surrogate markers, as well as less risk for developing resistance. Patients treated with bedaquiline had a faster time to culture conversion compared to placebo – 125 days versus 83 days – and a higher rate of culture conversion at 24 weeks and 120 weeks – 62% compared to 44%, the investigators reported.

The cure rate at 120 weeks was 58% for the treatment arm versus 32% for placebo. Those on the test drug were less prone to developing resistance.

Diacon, et al., note that the fact that the “significant difference in cure rates was maintained at 120 weeks … provides evidence of the potential usefulness of this surrogate endpoint.”

Safety was similar overall, but the difference in the number of deaths remained (10 for Sirturo vs. two for placebo). The rate in the placebo arm was “surprisingly low” compared to a meta-analysis of similar populations in TB studies, the investigators pointed out.

The mortality imbalance is still unexplained and the authors acknowledged that the relatively small size of the trial is a limitation.

A Paradoxical Approval

In an editorial in the NEJM accompanying the results, FDA Office of Antimicrobial Products’ Edward Cox and Katherine Laessig noted the imbalance in deaths in one Phase II study as some of the normal complexity that regulators must grapple with.

“Given this imbalance in mortality, the approval of bedaquiline has appeared paradoxical to some. But marketing applications that are reviewed by the FDA often rely on complex risk–benefit evaluations,” the officials explained.

Results from a confirmatory study will be helpful in delineating the efficacy and safety, but it will take time to get these data, they noted.

“Nonsurrogate endpoints such as patient survival, clinical resolution of tuberculosis, and rate of relapse will be included in the confirmatory trial. Although these clinical endpoints may be regarded as more rigorous and ‘traditional’ than a microbiologic endpoint of sputum-culture conversion, their use will prolong the study,” Cox and Laessig wrote.

The officials stressed the tremendous need for new TB treatments, noting that the “global health burden of tuberculosis remains enormous,” with an estimated 8.6 million new cases in 2012. And they note that the drug was approved for a very select population and that labeling has prominent warnings about the mortality results.

In conclusion, the editorial urges that “it is crucial that physicians and patients with multidrug-resistant tuberculosis carefully consider this information as well as the potential ramifications of inadequate treatment and increasing resistance.”

Diacon, et. al., advised “caution” with respect to development of the drug in new indications where a safe alternative exists until there is more data to explain the “excess deaths.”