More treatment failure in people on TB treatment who start once daily nevirapine-based ART than efavirenz-based ART

This article is more than 15 years old. Click here for more recent articles on this topic

People on TB treatment who started a once-daily antiretroviral therapy (ART) regimen of nevirapine/ddI/3TC were significantly more likely to fail ART than those who started on a once-daily regimen of efavirenz/ddI/3TC, according to a randomised prospective study from Chennai, India. In fact, the nevirapine arm performed so poorly that the study’s Data Safety and Monitoring Board (DSMB) ended accrual to that study arm and closed the study ahead of schedule.

“The once-daily nevirapine arm was definitely inferior to the efavirenz regimen in terms of virological failure and death and therefore is not recommended to be used when patients are on rifampicin-containing anti-TB treatment,” said Dr Soumya Swaminathan of Chennai’s Tuberculosis Research Centre, who presented the findings this week at the Sixteenth Conference on Retroviruses and Opportunistic Infections in Montreal.

Drug interactions between rifampicin and antiretroviral drugs

Treating both HIV disease and active TB at the same time is complicated by drug interactions between rifampicin, the key anti-TB drug, and some antiretrovirals. Rifampicin strongly induces the cytochrome p450-enzyme system that metabolises many drugs, including the protease inhibitors and the non-nucleoside reverse transcriptase inhibitors (NNRTIs). Of the two widely available NNRTIs, efavirenz levels are less affected by rifampicin (by about 20%), so it is the preferred drug to be used in combination with anti-TB treatment.

However, nevirapine is the most widely used NNRTI worldwide because it is cheap (efavirenz generally costs twice as much), known to be safe in pregnant women and children, and is available in fixed-drug combinations. However, pharmacokinetic studies suggest that rifampicin co-administration reduces blood levels of nevirapine by 35 to 55%, and there have also been reports of more liver toxicity with the combination.



How well something works (in a research study). See also ‘effectiveness’.

pharmacokinetics (PK)

How drugs are processed and used in the body, including how they are absorbed, metabolised, distributed and eliminated.


Affecting the lungs.


prospective study

A type of longitudinal study in which people join the study and information is then collected on them for several weeks, months or years. 



It should be noted that nevirapine’s pharmacokinetics are complex – it must be given at lower doses during the first two weeks on treatment (the 'lead in') because it begins to induce its own metabolism. Regardless, it is not clear how or whether it can be used efficaciously and safely together with rifampicin.

A few clinical studies have looked at the interaction between NNRTIs and rifampicin. A study in 27 people with TB starting ART in Malawi reported good outcomes with Triomune (which contains a fixed dose of nevirapine), though the researchers noted 59% of subjects had sub-therapeutic nevirapine levels during the lead-in phase, which they recommended should be skipped in people already taking rifampicin (Van Oosterhout). In a study with 140 patients, Thai researchers reported that nevirapine-based ART had a similar virological efficacy whether or not subjects were also on TB treatment (Manosuthi, 2007). The same research team conducted a retrospective cohort study in 188 patients on TB treatment and found similar outcomes whether they had been treated with nevirapine- or efavirenz-based regimens, although they noted that adverse events tended to be higher in nevirapine-based ART (Manosuthi, 2008). However, the body weight of the Thai patients was quite low, and there may be ethnic differences in drug metabolism so these findings may not be generalisable to other populations (Avihingsanon).

Indeed, a recent large retrospective cohort report from South Africa suggests that they may not be (Boulle). This study included 2035 people who started efavirenz-based ART (1074 with concurrent tuberculosis) and 1935 with nevirapine-based ART (209 with concurrent tuberculosis). Virological outcomes were inferior when nevirapine-ART was commenced while on TB treatment (vs without concurrent tuberculosis) but concurrent TB treatment had little effect on the response to efavirenz, nor was there any effect seen on virological outcomes when people already on nevirapine or efavirenz started taking TB treatment.

It should be noted that the findings from South Africa only became available after the study in India was well underway.

The Chennai study

“The rationale behind the selection of our study regimens was we wanted to look for a good once-daily anti-retroviral regimen which could be cheap as well as widely available and that would facilitate directly observed treatment (DOTS) of ART along with anti-TB treatment,” said Dr Swaminathan.

She noted that several studies, such as the 2NN study have shown that once-daily nevirapine can be as effective as twice-daily. In particular, the combination of nevirapine/ddI/3TC once daily had been looked at and found to be effective and safe, except that it performed better in patients with CD4 >100 (Ribera). But there had been no prospective randomised clinical trials comparing once-daily nevirapine to efavirenz when given with rifampicin.

So Dr Swaminathan and colleagues at the Tuberculosis Research Centre in Chennai began the study in May 2006. The study enrolled ART-naive HIV-positive adults with CD4 cell counts below 250 and newly diagnosed pulmonary or extra-pulmonary TB, who were first treated with a standard rifampicin-containing regimen (a two-month intensive phase, followed by four months of rifampicin/isoniazid). The study excluded anyone who was pregnant, terminally ill, with a psychiatric illness, or without normal liver and renal function.

Following the intensive phase of TB treatment, the participant’s TB disease was assessed, their CD4 counts and viral loads measured and they were randomised to one of the two ART regimens: either efavirenz (600 mg) or nevirapine (400 mg, after a 14-day lead-in phase with 200 mg) along with ddI (250/400 mg) and 3TC (300 mg), all given once daily in the morning. All drugs were directly observed, three days in the week (with TB treatment), for up to 24 weeks.

Participants were followed every month with a clinical exam, weight measurements, sputum smear and mycobacterial culture. CD4 and viral loads were done at four, 16 and 24 weeks after initiation of ART, and liver and renal function tests were done every two weeks for the first eight weeks, and then every four weeks thereafter.

Unfavourable responses included: a viral load >400 copies/mL or death at 24 weeks; a serious adverse event necessitating a change of study drug; and loss to follow-up.

This was a non-inferiority study designed to see whether once-daily nevirapine was at least similar to the effectiveness of efavirenz. The sample size was based on the assumption that the control regimen (efavirenz) would have a favorable virological response (VL <400 copies/ml) of about 90%. In order to detect a difference of 15%, a sample size of 90 in each arm was required at a power of 80% and a significance level of 5%. “We were willing to accept a lower limit of 75% efficacy in the NVP [nevirapine] arm,” said Dr Swaminathan.

What they found was significantly worse.


A total of 116 individuals were enrolled (93 men, 23 women), 69% with culture-positive pulmonary TB. Baseline characteristics were well matched, with a mean weight between 42 and 43 kg; the median CD4 count was 84 cells and a median viral load of 310,000 copies/mL at baseline.

The response to TB treatment was good – with a 90% treatment response overall. However, Dr Swaminathan said that “though it was not statistically significant, there did appear to be a trend towards poorer response in the NVP arm”.

As far as the virological efficacy, in an intent-to-treat analysis, 50 of 59 (85%) of the patients in the efavirenz arm and 38 out of 57 (67%) in the nevirapine arm had viral loads below 400 at week 24 (p=0.038).

There were nine unfavourable responses on efavirenz vs 19 on nevirapine, including five deaths and eleven virologic failures in the neviraine arm, while there were no deaths and six failures in the efavirenz arm. Three were lost to follow-up in each arm. The relative risk of failure on nevirapine was 1.28 (95% CI 1.03 to 1.6).

“Doing an efficacy analysis only for those patients who had greater than 80% adherence to the study drugs, we had 55 [efavirenz] and 54 [nevirapine] patients in the two regimens. And again, the favourable response in the EFV regimen was seen in 91% compared to 70% in the NVP arm,” said Dr Swaminathan.

However, there didn’t seem to be much difference in CD4 cell response, or in the number of serious adverse events and side-effects. “We were concerned about the potential hepatotoxicity of combining NVP and rifampicin, but we didn’t have any major problems,” she said.

The second interim analysis led the DSMB to advise stopping intake into the nevirapine arm but to continue analysis of those patients already randomised. But by the third interim analysis in June 2008, the results were clearly significant and the study was halted.

Discussion and alternative approaches

It is important to note that after the Chennai study began, other studies have reported problems with once-daily nevirapine dosing even in people who aren’t on concurrent TB treatment, including early virological failure and higher rates of liver toxicity and rash – which is at least part of the reason nevirapine’s maker, Boehringer-Ingelheim, has begun studying an extended-release formulation of the drug. In the discussion after her talk, Dr Swaminathan also noted that although pharmacokinetic studies have suggested that the area under the curve concentrations are similar whether nevirapine is dosed once or twice daily, the peak levels are higher and the trough levels are actually lower when it is given once daily.

“With rifampicin having an added effect, we probably had many patients going to sub-therapeutic levels, and we do have data on pharmacokinetics on all these patients that we will be looking at,” she said.

But it is important not to give up on nevirapine, because in many settings or some populations (such as children under the age of three with TB and HIV) there are very few alternatives. In fact, described a failed attempt to deliver adequate levels of lopinavir to children on TB treatment by double dosing Kaletra.

There is also the possibility that the study results would have been different without the lead-in dose.

“I think there’s a place for more studies. If I were to design this trial today, I think I would initiate nevirapine at 400 mg once daily in patients who have already been on rifampicin because I think that’s where all the [resistance] mutations develop, during that time you have very low blood levels,” said Dr Swaminathan.

Another possibility is that doses of nevirapine should be adjusted for use in TB patients. A recently published mathematical model of nevirapine population pharmacokinetics in a South African population, when taken with and in the absence of rifampicin treatment, found that an increased dose of 300 mg twice daily would achieve adequate nevirapine concentrations in most patients during rifampicin-containing treatment for tuberculosis (Elsherbiny). But this would need to be evaluated in a clinical setting to be certain that it doesn’t result in unacceptable levels of toxicity.

Another alternative would be to increase the worldwide access to rifabutin, which is related to rifampicin but doesn’t affect blood levels of antiretroviral drugs nearly as much. In fact, no dosage adjustment is required with nevirapine. There are moves afoot to get rifabutin on WHO’s Essential Drug List – but that would be only a very early step in making it available in a programmatic setting, and rifabutin is yet to be formulated for use by children. Furthermore, more evidence may be needed to demonstrate that it is of equivalent efficacy to rifampicin in fixed-dose drug combinations for TB.


Swaminathan S et al. Once-daily nevirapine vs efavirenz in the treatment of HIV-infected patients with TB: a randomized clinical trial. Sixteenth Conference on Retrovirus and Opportunistic Infections, Montreal, abstract 35, 2009.

Boulle A et al. Outcomes of nevirapine- and efavirenz-based antiretroviral therapy when coadministered with rifampicin-based antitubercular therapy. JAMA. 300(5):530-539, 2008.

Elsherbiny D et al. Population pharmacokinetics of nevirapine in combination with rifampicin-based short course chemotherapy in HIV- and tuberculosis-infected South African patients. Journal European Journal of Clinical Pharmacology. 65(1) 71-80, 2009.

Manosuthi W et al. Nevirapine levels after discontinuation of rifampicin therapy and 60-week efficacy of nevirapine-based antiretroviral therapy in HIV-infected patients with tuberculosis. Clin Infect Dis. 2007;44 (1):141-144.

Manosuthi W et al. Standard-dose efavirenz vs standard-dose nevirapine in antiretroviral regimens among HIV-1 and tuberculosis co-infected patients who received rifampicin. HIV Med. 2008;9(5):294-299.

Ribera E et al. Efficacy and safety of once-daily combination therapy with didanosine, lamivudine and nevirapine in antiretroviral-naive HIV-infected patients. Antiviral Therapy 10:605–614, 2005.

van Oosterhout JJ et al. Nevirapine-based antiretroviral therapy started early in the course of tuberculosis treatment in adult Malawians. Antivir Ther.12 (4):515-521, 2007.