Conference Coverage

Prednisolone, immunotherapy ineffective for most tuberculous pericarditis

View on the News

Stop routine glucocorticoids

These findings clearly suggest that adjunctive glucocorticoids should not be used routinely in patients with tuberculous pericarditis, which may surprise many clinicians. But because they prevent constrictive pericarditis and reduce hospitalizations, the drugs may be appropriate for patients at highest risk for complications, such as those with large effusions, high levels of inflammatory cells or markers in the pericardial fluid, and early signs of constriction.

The use of glucocorticoids should be curtailed in patients coinfected with HIV unless the risk of constrictive pericarditis is high, since these drugs increase the risk of cancer in this patient population.

Dr. Richard E. Chaisson and Dr. Wendy S. Post of Johns Hopkins University, Baltimore, made these remarks in an editorial accompanying Dr. Mayosi’s report (N. Engl. J. Med. 2014 Sept. 2 [doi:10.1056/NEJMe1409356]. Dr. Chaisson reported receiving support from Merck.




Neither standard anti-inflammatory therapy using prednisolone nor an experimental immunotherapy using Mycobacterium indicus pranii injections reduced the composite outcome of death, cardiac tamponade requiring pericardiocentesis, or constrictive pericarditis in an international clinical trial of adults with tuberculous pericarditis, a study showed.

However, prednisolone was beneficial for one component of this composite outcome – lowering the rate of constrictive pericarditis – compared with placebo, Dr. Bongani M. Mayosi reported at the annual congress of the European Society of Cardiology in Barcelona. His report was simultaneously presented at the meeting and published online Sept. 2 (N. Engl. J. Med. 2014 [doi:10.1056/NEJMoa1407380]).

Glucocorticoids are thought to attenuate the inflammatory response in patients with tuberculous pericarditis, and are recommended as adjunctive therapy in current American and World Health Organization treatment guidelines. But the studies on which these recommendations are based had "very small" numbers of events and patients, and the treatment effect was minimal, leading European expert groups to advise against using the drugs in this patient population. In addition, glucocorticoids may raise the risk of cancer in patients who are coinfected with HIV, which is common in the regions of sub-Saharan Africa and Asia where tuberculous pericarditis is most frequent, said Dr. Mayosi of the department of medicine, Old Groote Schuur Hospital, Cape Town, South Africa.

The 1,400 participants in the Investigation of the Management of Pericarditis (IMPI) trial had pericardial effusion confirmed by echocardiography and evidence of definite or probable tuberculous pericarditis; two-thirds also had concomitant HIV infection. They were treated at 19 hospitals across 8 countries in Africa during a 5-year period. All received background antimicrobial therapy for tuberculosis, and all patients with HIV received antiretrovirals according to WHO guidelines.

The participants first were randomly assigned to receive prednisolone (706 patients) or placebo (694 patients) in tapering doses for 6 weeks. In the second phase of the study, 1,250 of these participants were then randomly assigned to receive five intradermal injections of either heat-killed M. indicus pranii or placebo at intervals over the course of 3 months. This nonpathogenic, rapidly growing mycobacterium species has been shown to suppress inflammation in patients with leprosy.

The primary efficacy outcome was a composite of death, cardiac tamponade requiring pericardiocentesis, or constrictive pericarditis. After a median follow-up of 636 days, there were 14.3 such events per 100 patient-years in the prednisolone group and 14.8 in the placebo group, a nonsignificant difference.

When each component of this composite outcome was considered individually, prednisolone did not improve the death rate or the rate of cardiac tamponade, compared with placebo, but did reduce the rate of constrictive pericarditis (4.4%, vs 7.8% with placebo), and thus the rate of hospitalization (20.7%, vs 25.2% with placebo). "This finding is important because pericardiectomy, the definitive treatment for chronic pericardial constriction, is associated with high perioperative mortality and morbidity, and cardiac surgery is not widely available in Africa," Dr. Mayosi said.

Prednisolone also raised the rate of opportunistic infections, chiefly candidiasis, compared with placebo. And it markedly increased the rate of cancer, primarily Kaposi’s sarcoma, in patients coinfected with HIV (1.05 cases per 100 person-years, vs. 0.32 cases with placebo).

In the M. indicus pranii comparison, the active treatment was no different from placebo with regard to the composite outcome or any secondary outcomes, and that portion of the trial was halted early for futility. Like prednisolone, this experimental agent raised the rate of cancer in HIV-positive patients (0.92 cases per 100 person-years, vs. 0.24 cases with placebo) – an adverse event that has not been reported previously with M. indicus pranii.

In addition, significantly more patients who received M. indicus pranii (41.4%) than placebo (2.9%) developed injection-site reactions. Fifteen percent of patients given the active injections developed abscesses at the injection site, compared with only 1% of those given placebo injections.

The IMPI trial 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). Cadila donated the study drugs but had no role in the design or conduct of the study or in data analysis. The authors had no relevant financial conflicts of interest to disclose.

Next Article: