Efavirenz and Nevirapine Interactions with Rifampicin: Resolving the Dilemmas?

Tuberculosis (TB) is the most common opportunistic infection in persons with human immunodeficiency virus (HIV) infection, and the 2 diseases together represent a deadly duo 1. The TB epidemic has been aggravated by the HIV epidemic, particularly in developing and resource-limited regions of the world. According to the World Health Organization (WHO), there were 9.2 million new cases and 14.2 million prevalent cases of TB globally in 2006 2. It is estimated that 1.2 million deaths in 2006 were attributable to TB. Globally, there were an estimated 709,000 new cases of TB among HIV-infected persons in 2006. Africa accounts for 85% of these cases, India accounts for 3.3%, Europe accounts for 1.8%, and other countries account for 9.4% 2. The burden of TB is increasing in countries such as South Africa, where the rate of HIV infection is still very high. Currently, the incidence of TB is close to 1000 cases per 100,000 persons in South Africa 2. It is comforting that emerging data support the hypothesis that coadministration of highly active antiretroviral treatment (HAART) to patients who are receiving treatment for active TB improves morbidity and mortality 3. The question of early versus delayed introduction of HAART should be answered by the A5221 Study of the Adult AIDS Clinical Trials Group and the Cambodian Early vs. Late Introduction of Antiretrovirals (CAMELIA) Study 4. Concurrent treatment of HIV infection in patients who are receiving anti-TB treatment is a top priority in countries with a high burden of HIV infection and TB burden. Therefore, it is important that HAART be prescribed to eligible patients receiving anti-TB treatment with confidence and safety. Rifampicin forms the backbone of first line anti-TB treatment and has allowed for the use of short-course chemotherapy for TB even in HIV-infected patients. However, the drawback of rifampicin is that it is a potent inducer of cytochrome P450 CYP3A4, which, in turn, interferes with the metabolism of a number of antiretroviral drugs—notably, efavirenz, nevirapine and the protease inhibitors, and the recently developed CCR5-receptor antagonist maraviroc. Rifampicin also reduces the concentration of the integrase inhibitor raltegravir, although the recommended dosage of 400 mg twice per day appears to be adequate when raltegravir and rifampicin are coadministered; however, clinical experience with the combination of raltegravir and rifampicin is very limited 5. The second-generation nonnucleoside reverse-transcriptase inhibitor (NNRTI) etravirine is a diarylpyrimidine derivative that was developed to overcome NNRTI-resistant variants that encode L100I, K103N, Y181C, Y188L, and G190A/S mutations 6. Etravirine cannot be recommended for use with rifampicin, because rifampicin has been shown to significantly reduce the concentration of etravirine 6. The NNRTI delavirdine is less potent than efavirenz and nevirapine and inhibits cytochrome P450; use of this agent is not recommended in patients receiving rifampicin 7. The data for administration of protease inhibitors with rifampicin are also disappointing. Recent evidence has shown that administration of increased doses of lopinavir-ritonavir to healthy volunteers who were also taking rifampicin resulted in a high frequency of nausea, vomiting, and hepatitis 8. Whether patients with HIV infection and active TB would experience similar levels of hepatotoxicity is unclear, but the data call for extreme caution. Ritonavir-boosted lopinavir may still be useful for young children who are taking rifampicin 9, and this strategy remains an option for adults in resource-limited settings. The clinical vigilance required when administering protease inhibitors is least available in these settings. Because of the lack of adequate data, the general recommendation is that protease inhibitors should be avoided for patients who are receiving rifampicin. Nucleoside reverse-transcriptase inhibitors (NRTIs) are generally well tolerated, but HAART regimens that consist of NRTIs alone are inferior to NNRTI-based regimens 10, although NRTI-based regimens are currently recommended by the WHO in patients who cannot tolerate efavirenz or nevirapine. A controlled trial of a NRTI-based HAART regimen given concurrently with TB treatment is essential. The clinical dilemma also relates to patients who develop active TB while receiving protease inhibitor–based HAART regimens. The only options are the first-generation NNRTIs efavirenz and nevirapine. The jury is still out regarding whether higher doses of efavirenz are needed when it is administered with rifampicin and whether nevirapine is actually safe, given the risk of hepatotoxicity. Although higher doses of efavirenz could be used to overcome the reduced therapeutic concentrations that occur when it is given with rifampicin, potential increases in neuropsychiatric adverse effects and hepatotoxicity are a concern. In the event that efavirenz use is contraindicated for women who have been exposed to single-dose nevirapine for prevention of mother-to-child transmission of HIV, the options for HAART are limited. This is because nevirapine-based HAART regimens are contraindicated because of a high risk of resistance 11. In the event of resistance to NNRTIs, the options are severely restricted. The alternative is to look for substitutes for rifampicin. The replacement candidates are rifapentine and rifabutin, which are rifamycins and which are less likely to cause induction of cytochrome P450. These drugs are prohibitively expensive, especially in resource-limited regions, where the need is greatest. Rifabutin is contraindicated in persons who are receiving ritonavir, which is a potent CYP3A inhibitor, because ritonavir may increase the levels of rifabutin up to 4 times, with attendant toxicity 12. Alternatives to current NNRTIs (efavirenz and nevirapine) are also limited. There are no data to support replacement of rifampicin with moxifloxacin in anti-TB regimens. The Global Alliance for TB Drug Development has only elected to investigate replacement of ethambutol and isoniazid with moxifloxacin in its clinical drug development program, although there are data from short-term studies of humans that moxifloxacin is at least as potent as rifampicin in its bacteriocidal activity against Mycobacterium tuberculosis 13, 14 . This remains an option that warrants attention by the Global Alliance for TB Drug Development. The N2R study by Manosuthi et al. 15 in this issue of Clinical Infectious Diseases is the first prospective, randomized clinical trial to have compared serum concentrations and the efficacy of efavirenz and nevirapine in patients receiving a standard anti-TB treatment regimen that contains rifampicin. This study has 3 important conclusions: (1) efavirenz, given at a standard dosage of 600 mg per day, is adequate for suppression of HIV, despite variation in blood concentrations of the drug; (2) nevirapine is effective in standard dosages of 400 mg per day, although efavirenz is superior; and (3) low drug concentrations and body weight <55 kg were predictive of HAART failure. The higher frequency of lower drug concentrations explained the trend toward a higher failure rate for nevirapine, compared with efavirenz. Presumably, lower body weight predicted more-advanced HIV infection and/or TB, as well as a higher mortality rate, but would not explain HAART failure. Although the objective of the N2R study was to assess virologic outcomes, it would have been informative if the TB outcomes had been reported. This may well be the topic of a separate publication. The N2R study involved Thai patients; the question is whether these findings may be extrapolated to other population groups. The authors note that Asians have certain allotypes of CYP2B6 that predict higher concentrations of NNRTIs and account for the good outcomes without increase in dosage in persons who receive concomitant rifampicin treatment. A small study from South Africa also reported a good virologic response with a 600-mg daily dose of efavirenz 16. A recent study found polymorphisms in CYP2B6 in African American persons 17. This indicates that studies that involve different populations are required. Two large, ongoing studies of early versus deferred HAART in patients with TB–AIDS Clinical Trials Group study A5221, which is a multinational, and the CAMELIA study in Cambodia, which is a National Institutes of Health Comprehensive International Program for Research on AIDS (CIPRA) study—should resolve the issue of efficacy of efavirenz 4. It is difficult to predict how many patients in these 2 studies will receive nevirapine. In conclusion, when given at a standard dosage of 600 mg per day, efavirenz appears to be the NNRTI of choice for HAART for patients who are receiving rifampicin-based TB treatment. If efavirenz is contraindicated, then nevirapine (at a standard dosage of 400 mg per day) appears to suffice. We must await the results of the A5221 and CAMELIA trials for a definitive position on efavirenz. If NNRTIs cannot be used, then the options are severely restricted and may involve the cautious use of a ritonavir-boosted lopinavir based regimen. Substitution of rifabutin is an option where this drug is available. Currently, there is no suitable substitute for the rifamycins, although moxifloxacin is a potential candidate. In the absence of these options, one may have to opt to defer HAART until TB treatment has been completed, but there is a high mortality rate in the absence of HAART. The dilemmas are unresolved, and clinicians and scientists will continue to be challenged by the complexities of TB treatment and HAART. In the midst of all this, we should not lose sight of the humble goal of prevention as the most cost-effective strategy to combat the deadly duo of TB and HIV infection. Potential conflicts of interest. U.G.L.: no conflicts.