This site is intended for health professionals only
Frank van Leth
International Antiviral Therapy Evaluation Center
Academic Medical Center
Department of Internal medicine
University of Amsterdam
The HIV genome encodes for an enzyme called reverse transcriptase (RT). This enzyme initiates the transcription of viral RNA into viral DNA, which is an essential step in the lifecycle of the HIV virus. Thus, the HIV-RT is a pivotal molecular target for antiretroviral therapy.
Non-nucleoside reverse transcriptase inhibitors (NNRTIs) bind to a specific hydrophobic pocket on the HIV-RT. This binding inhibits reverse transcription, due to slowing down of the chemical process needed to bind the next nucleotide or to premature dissociation of the template–primer complex.
The first NNRTIs, identified in 1989, were 1-[2- (hydroxyethoxy)methyl]-6-(phenylthio)thymine (HEPT) and tetrahydroimidazobenzodiazepinone (TIBO) derivatives. Today, the NNRTIs form a diverse class of antiretroviral drugs, with over 50 different structural compounds. All NNRTIs specifically interfere with the function of HIV-1 RT. NNRTIs do not show any inhibitory effect on the RT function of other viruses (eg, HIV-2, simian or feline immunodeficiency virus or murine leukaemia virus) or other (viral or cellular) polymerases.
Several NNRTIs, such as atevirdine and loviride, reached the early stages of clinical studies, but further development was stopped due to limited clinical efficacy. To date, only three NNRTIs are licensed for clinical use: nevirapine (NVP; 1996), efavirenz (EFV; 1998) and delaviridine (DLV; 1997, USA only).
The antiretroviral efficacy of NNRTIs
At the time of the introduction of NNRTIs, treatment of HIV-1 consisted of a combination containing two nucleoside analogues and a protease inhibitor (PI). These combinations proved very effective in numerous cohort studies and clinical trials. However, the pill burden of most PI-based regimens is high, and side-effects are considerable; in addition, the use of PIs imposes strict dietary rules on the patient. These drawbacks are less marked with NNRTI-based regimens.
It was shown that a combination of stavudine (d4T), didanosine (ddI) and NVP in therapy-naive patients was comparable, with respect to antiretroviral potency, to a regimen of stavudine (d4T), didanosine (ddI) and indinavir (IDV).(1) Similarly, the efficacy of NVP combined with zidovudine (AZT) and lamivudine (3TC) was comparable to that of nelfinavir (NFV) in combination with these nucleoside analogues.(2) A study comparing EFV and IDV in combination with AZT and 3TC showed superior efficacy of the EFV-based regimen in antiretroviral-naive patients.(3) An EFV-based regimen was superior to an NFV-based one in patients who had failed a regimen based on nucleoside analogues only.(4) No study comparing the efficacy of DLV- and PI-based highly active antiretroviral therapy (HAART) regimens has yet been conducted.
Switch studies provided further evidence that NNRTI- and PI-based regimens have comparable efficacies. In these studies, patients on a stable PI-based regimen with adequate viral suppression were switched to an NNRTI. Virological efficacy in patients who switched to an NNRTI-based regimen was not compromised, compared with those who continued on the PI-based regimen.(5,6) Similar conclusions were reached in studies in which all patients were switched from a PI- to an NNRTI-based regimen.(7,8) In a meta-analysis of switch studies, relative risk (95% confidence interval) of virological failure after switching from a PI- to an NVP-based regimen was 0.54 (0.29–1.02), and 0.83 (0.36–1.91) after switching to an EFV-based regimen.(9)
In a recent large randomised trial, it was demonstrated that the efficacy of NVP and EFV were similar, although cohort studies showed more efficacy for EFV.(10)
Side-effects of NNRTIs
The side-effects of the three licensed NNRTIs are well described. For NVP and DLV, the main side-effect is a drug-induced skin rash. A severe rash is seen in 3–7% of patients using NVP as part of their HAART regimen, prompting discontinuation of NVP.(11) This rash seldom develops into Stevens–Johnson syndrome (SJS) or toxic epidermal necrolysis (TEN), with possible fatal outcome. When a rash develops as a consequence of the use of NVP, the drug should be stopped (and not reintroduced) when rash is severe or accompanied with constitutional symptoms or increased liver transaminases. In patients using DLV, the rash is generally of mild-to-moderate severity and often does not require discontinuation of the drug; SJS and TEN have not been reported. The drug-induced rash mostly occurs in the early weeks of treatment (first six weeks with NVP, and first four weeks with DLV).
NVP is also associated with hepatotoxicity. In an extensive analysis of patients treated with NVP in 16 clinical studies (nine uncontrolled and eight controlled trials), 5% of patients developed symptomatic hepatic events, of which half were associated with rash.(12) Although hepatotoxicity has been reported with most antiretroviral drugs, it is more frequent in patients using NVP. Fatal hepatic failure has been reported in patients using NVP.
Current guidelines emphasise the need for rigorous monitoring of liver transaminase levels during the first 6–18 weeks of treatment. If liver transaminases levels are elevated and clinical or laboratory signs of hepatitis are detected, NVP should be permanently discontinued. However, sudden liver failure might occur without prior warning by elevated liver transaminases.
The main side-effects associated with EFV are neurological and/or psychiatric in nature. Sleep disturbances, vivid dreams or nightmares and dizziness have been reported in more than 50% of patients. Although severe effects are less common, they can have a marked negative influence on quality of life.(13) It is generally believed that these side-effects occur for a short period of time, although some authors disagree with this finding. Dividing the once-daily dose in two or changing the timing of drug intake can have a positive outcome on side-effects.
In the last few years, it has become clear that antiretroviral therapy is associated with metabolic changes. PIs are associated with increased insulin resistance and increases in plasma concentrations of cholesterol and triglycerides. Such lipid changes might have some effects on cardiovascular risks. NNRTI-based regimens show a clearly different, potentially less atherogenic, metabolic profile.(14) Concern about increased cardiovascular risks due to increased levels of lipids and lipoproteins is one of the reasons to choose NNRTI-based regimens or switch from a PI-based regimen to one including an NNRTI.(15)
NNRTIs, however, present a low genetic barrier for resistance, and a single mutation in the genome can increase IC(50) values by 40–100-fold. The two main resistance-associated mutations are found in codons 103 and 181 of the HIV-1 RT. Due to the narrow binding site in the hydrophobic pocket, there is a marked amount of cross-resistance among all three NNRTIs.
Optimal adherence, which results in adequate plasma drug concentrations, is essential to avoid evolution and selection of resistant virus.
NNRTIs are potent antiretroviral drugs for the treatment of HIV-1 infection.
The safety profile of these drugs includes possible fatal side-effects, requiring monitoring and timely drug cessation. The metabolic changes associated with NNRTIs are less detrimental than those linked to the use of PIs, making these drugs a valuable choice in patients with high risk of coronary heart disease.
Resistance to these drugs develops rapidly, making adequate adherence essential.