Short course in latent TB matches standard treatment

Also, treatment for tuberculosis pericarditis advances

Short courses of rifamycin-containing treatment regimens prevented active tuberculosis (TB) as well as standard monotherapy did, researchers reported.

In a Bayesian network meta-analysis, some of the rifamycin-containing regimens also were less likely to cause liver damage than standard therapy with isoniazid (Nydrazid), according to Helen Stagg, PhD, of University College London, and colleagues.

Such regimens might become "effective alternatives to isoniazid monotherapy," Stagg and colleagues reported online in Annals of Internal Medicine.

While 9 months of isoniazid -- the standard in the U.S. -- is efficacious at preventing latent TB from becoming active, the regimen's length means many patients do not complete therapy, the researchers noted.

In addition, isoniazid is associated with a risk of hepatotoxicity that can be fatal.

For those reasons, there has been increased interest in finding shorter regimens containing rifamycin derivatives, such as rifampin (Rifadin), rifapentine (Priftin), and rifabutin (Mycobutin).

"Effective treatment of latent tuberculosis infection is an important component of tuberculosis elimination programs," Stagg and colleagues noted.

While new regimens are promising, few can be directly compared, they reported. Most studies have involved conventional meta-analyses, which are "severely limited in the inferences they could make about relative efficacy and toxicity," they wrote.

To help fill the gap, they turned to Bayesian hierarchical models, whose network approach allows indirect comparisons "and thus produces inferences of relative efficacy that would not otherwise be possible," Stagg and colleagues reported.

The raw material for the analysis was 53 randomized, controlled trials that evaluated treatment for latent TB and reported at least one of two endpoints -- preventing active TB or hepatotoxicity.

Fifteen regimens were available, leading to 105 possible comparisons of which 42 were compared directly in the studies.

The analysis showed that, compared with placebo, several regimens were efficacious:

  • Isoniazid for 6 months or 12 months and longer had odds ratios for active TB of 0.64 and 0.52, respectively.
  • Rifampin for 3 to 4 months had an OR for active TB of 0.41.
  • Rifapentine and isoniazid and rifampin-isoniazid were also efficacious with ORs of 0.61 and 0.52, respectively.

When the researchers ranked the regimens, rifampin alone or with isoniazid for 3 to 4 months seemed to be particularly efficacious, although they cautioned the finding is based on limited data.

Data on hepatotoxicity was sparse, they reported, but analysis based on direct comparisons suggested that rifampin alone or with isoniazid was less toxic than isoniazid alone for 6, 9, or 12 to 72 months.

The analysis contains "two useful and important messages," commented Dick Menzies, MD, of the Montreal Chest Institute in Montreal and Timothy Sterling, MD, of Vanderbilt University School of Medicine in Nashville, Tenn.

The first is that evidence accumulating over the past 20 years shows clearly that short courses of rifamycin-containing regimens work as well and are safer than those based on isoniazid, they argued in an accompanying editorial.

"Surely, it is time to get a move on -- away from (isoniazid) as our primary therapy and toward regimens containing rifamycin," they wrote.

For patients, the "considerable" advantages include greater safety and better protection against TB, while for TB programs, shorter treatment courses potentially could mean lower workloads and costs.

The second message -- which rifamycin-containing regimen physicians should choose -- is less clear, they noted.

For instance, the regimen that had the best combined rankings (efficacy and lack of hepatotoxicity) was rifampin for 3 or 4 months, but the efficacy result was based on a single 3-month trial.

"Further trials are needed to clarify which of these short-course regimens offer the greatest advantages for patients and programs and under what conditions," Menzies and Sterling concluded.

Stagg and colleagues cautioned that the risk of bias was unclear in many of the studies, and there was limited data for many comparisons, especially those concerning liver toxicity.

Meanwhile, researchers reported that adding steroids to anti-tuberculosis treatment for patients with tuberculous pericarditis can reduce the risk of an important complication and the resulting hospital admissions.

TB and Mortality

But in a large randomized trial, adding prednisolone to TB treatment did not reduce the risk of death, according to Bongani Mayosi, MBChB, DPhil, of the University of Cape Town in South Africa, and colleagues.

And for patients with HIV -- a majority of trial participants -- the therapy was associated with about a threefold increase in the risk of malignancy, Mayosi reported at the European Society of Cardiology meeting.

Study results were published simultaneously in the New England Journal of Medicine.

Tuberculous pericarditis "is the most important and the most serious form of pericardial disease in the world," Mayosi said, occurring in about 10% of the 10 million new TB patients every year.

Earlier research had suggested that steroids could reduce the mortality of the disease, which is about 26% by 6 months after diagnosis, his group.

The bottom line of the Investigation of the Management of Pericarditis trial was that "we need to be selective in the use of steroids," Mayosi said, because despite some benefits, there remains the risk of cancers in people with concomitant HIV.

Much of the disease burden of tuberculous pericarditis is in developing countries, where many TB patients are also co-infected with HIV, Mayosi noted.

The study also investigated the effect of injections of a nonpathogenic bacterium, heat-killed Mycobacterium indicus pranii (M. indicus pranii), also known as Mycobacterium w, which had been hypothesized to reduce the inflammation associated with TB.

But that part of the study was stopped early because of futility, he reported.

The investigators used a factorial design to investigate the effects of the two substances on a composite endpoint of death, cardiac tamponade requiring pericardiocentesis, or constrictive pericarditis.

Some 1,400 patients with probable or confirmed tuberculous pericarditis were randomly assigned to get prednisolone or placebo for 6 weeks. Within each group they were re-randomized to get five injections of M. indicus pranii or placebo over 3 months.

The researchers also looked at the elements of the composite endpoint separately as secondary efficacy endpoints, as well as such things as opportunistic infections, hospital admissions, and the occurrence of new malignancies to judge safety.

For the steroid/placebo comparison, Mayosi reported, there was no significant difference in the primary outcome: 23.8% of those getting prednisolone and 24.5% on placebo reached that endpoint.

In particular, when the investigators looked at the separate elements, there was no significant difference in mortality -- 18.8% of those getting the steroid died from any cause compared with 16.6% of those on placebo.

The risk of cardiac tamponade was also not significantly different, with 3.1% of those on the steroid and 4.0% of those on placebo having the complication.

On the other hand, 4.4% of those getting prednisolone had constrictive pericarditis, compared with 7.8% of placebo patients, yielding a hazard ratio of 0.56 that was significant (P=0.009).

Hospital admissions were 20.7% versus 25.2% for prednisolone and placebo patients, respectively, leading to an HR of 0.79 (P=0.04).

The difference was mainly driven by constrictive pericarditis, Mayosi said.

For the other comparison -- M. indicus pranii or placebo -- there was no significant difference either in the composite endpoint or the separate components, he reported.

On the other hand, both interventions were associated with a significant increase in the risk of malignancy compared with placebo:

  • The incidence of new cancers was 1.8% for prednisolone versus 0.6% for placebo (HR 3.27, 95% CI 1.07-10.03, P=0.03)
  • For the mycobacterium and placebo, the rates were 1.8% and 0.5%, respectively (HR 3.69, 95% CI 1.03-13.24, P=0.03)

The increase in HIV-related cancer with prednisolone therapy was mainly due to Kaposi's sarcoma and non-Hodgkin lymphoma, Mayosi said.

The reduction in constrictive pericarditis and hospital admission with prednisolone treatment is "clinically meaningful," commented Richard Chaisson, MD, and Wendy Post, MD, both of the Johns Hopkins University School of Medicine.

In an accompanying editorial, they said, "subgroup analyses to determine which patients benefited most will be informative."

But HIV-positive patients have a clearly greater risk of cancer, suggesting the "use of glucocorticoids should be curtailed in this population unless the risk of constrictive pericarditis is high," Chaisson and Post argued.

On the other hand, the lack of benefit from M. indicus pranii is no surprise, they added, "since the scientific rationale for this therapy in patients with tuberculous pericarditis is underwhelming."

Stagg and some co-authors disclosed financial support from the U.K. National Institute for Health Research.

Stagg disclosed relevant relationships with Otsuka Pharmaceutical and Sanofi. Other co-authors disclosed no relevant relationships with industry.

A co-author disclosed a relevant relationship with the World Health Organization (WHO). Another co-author disclosed serving on the WHO Latent Tuberculosis Guideline Development Group, serving as chair of the U.K. National Institute for Health and Care Excellence Tuberculosis Guideline Development Group, and serving as head of the Tuberculosis Section at Public Health England.

The study by Mayosi's group was supported by the Canadian Institutes of Health Research, the Canadian Network and Centre for Trials Internationally, the Population Health Research Institute, the South African Medical Research Council, the Lily and Ernst Hausmann Research Trust, and Cadila Pharma, India.

Mayosi and one co-author disclosed relevant relationships with Cadila Pharma.

Chaisson disclosed a relevant relationships with Merck. Post disclosed no relevant relationships with industry.

Primary Source: Annals of Internal Medicine
Source Reference: Stagg HR, et al "Treatment of latent tuberculosis: A network meta-analysis" Ann Intern Med 2014; DOI: 10.7326/M14-1019

Secondary Source: New England Journal of Medicine
Source Reference: Mayosi BM, et al "Prednisolone and Mycobacterium indicus pranii in tuberculous pericarditis" N Eng J Med 2014; DOI: 10.1056/NEJMoa1407380

Additional Source: New England Journal of Medicine
Source Reference: Chaisson RE, Post WS "Immunotherapy for tuberculous pericarditis" N Eng J Med 2014; DOI: 10.1056/NEJMe1409356

Source: MedPage Today

To subscribe to the Weekly Newsletter of new posts, enter your email here:

By Michael Smith

Published: Dec. 19, 2014, 10:10 p.m.

Last updated: Dec. 20, 2014, 3:36 a.m.

Print Share