Stavudine, lamivudine and nevirapine combination therapy for treatment of HIV infection and AIDS in adults.

Abstract

Background

A favourable regimen for people infected with HIV/AIDS is one that provides optimal efficacy, durability of antiretroviral activity, tolerability, and has low adverse effects and drug‐drug interactions. The combination of the non‐nucleoside reverse transcriptase inhibitor nevirapine (NVP), and two nucleoside reverse transcriptase inhibitors, stavudine (d4T) and lamivudine (3TC), is widely used as first‐line therapy, especially in low‐resource countries. Analysis of the efficacy, durability and tolerability of the regimen is thus important to clinicians, consumers and policy‐makers living in both rich and poor countries.

Objectives

To examine the efficacy of the stavudine, lamivudine and nevirapine regimen for the treatment of HIV infection and AIDS in adults.

Search methods

We used the comprehensive search strategy developed specifically by the Cochrane HIV/AIDS Review Group to identify HIV/AIDS randomised controlled trials, and searched the following electronic databases: MEDLINE (searched July 2004); Embase (searched October 2004); and CENTRAL (July 2004). This search was supplemented with a search of AIDSearch (April 2005) to identify relevant conference abstracts, as well as searching reference lists of all eligible articles. The search was not limited by language or publication status.

Selection criteria

Randomised controlled trials of the stavudine, lamivudine and nevirapine regimen, compared with any other regimens for treating HIV/AIDS, in antiretroviral treatment‐naive or antiretroviral treatment‐experienced adults.

Data collection and analysis

Two reviewers independently assessed the methodological quality of the trials and extracted data.

Main results

Our search resulted in 1,148 records, of which two studies described trials that met our inclusion criteria. One trial was a small single‐centre Australian trial of 70 antiretroviral‐naive participants, while the other trial was a large, multicentre trial, conducted in 14 countries, of 1,216 antiretroviral‐naive participants. In both trials over 60% of participants were male. As the therapeutic combinations compared in both trials were not identical, it was not possible to conduct a meta‐analysis to increase the power of the results. The main findings, therefore, are from the much larger trial, which was of a high quality. This trial found that there was no statistically significant difference in the efficacy (measured by treatment failure) between nevirapine and efavirenz (EFZ), when used in combination with 3TC and d4T (RR = 1.16; 95%CI: 0.95, 1.41). There was no statistically significant difference between once daily or twice‐daily dosing of NVP, when used in combination with 3TC and d4T (RR = 1.00; 95%CI: 0.83; 1.21). It also showed that, compared with NVP plus EFZ, 3TC and d4T, a once‐daily dosing of NVP, in combination with 3TC and d4T, performs better in averting treatment failure (RR = 0.82; 95%CI: 0.67, 1.00) than does twice‐daily dosing of NVP with 3TC and d4T (RR = 0.82; 95%CI: 0.69; 0.97). Frequency of toxicity was higher in participants receiving NVP, compared with EFZ.

Authors' conclusions

The combination of nevirapine, 3TC and d4T is as efficacious as a combination of efavirenz, 3TC and d4T. Once‐daily NVP with twice‐daily 3TC and d4T is as efficacious as twice‐daily NVP, 3TC and d4T. However, toxicity may be increased in the once‐daily NVP regime. Additional trials of sufficient duration are required to provide better evidence for the use of this combination as a first line therapy. Ideally, trials should use standardised assessment measures especially with respect to measuring viral load, so that results can be compared and combined in meta‐analyses.

Plain language summary

Stavudine, lamivudine and nevirapine combination therapy for treatment of HIV infection and AIDS in adults

People infected with HIV/AIDS require an antiretroviral regimen that works well, has good activity against the virus, has few adverse effects (unintended negative effects of the drug) and that does not interact with other drugs. The regimen of nevirapine, stavudine and lamivudine is widely used as first‐line therapy, and is recommended as such by the World Health Organization for so‐called low‐resource countries (in other words, for poor countries). This review identified two randomised controlled trials that assessed the efficacy of this drug combination. One trial was a small single‐centre Australian trial of 70 participants, whereas the other trial was a large, multicentre trial, conducted in 14 countries, of 1,216 participants. In both trials over 60% of participants were male and none had been on previous antiretroviral treatment. As one trial was very small, we cannot be sure of its results. The main findings therefore come from the much larger trial. This trial compared the combination of nevirapine, stavudine and lamivudine with the combination of efavirenz, stavudine and lamivudine, and found that participants had similar treatment outcomes on either combination. It also found that taking nevirapine once a day with twice daily stavudine and lamivudine worked as well as taking nevirapine twice a day in combination with twice daily stavudine and lamivudine. Nevirapine did appear to cause more adverse effects compared with efavirenz, but additional assessment of this is necessary to be more certain.

It is important that more trials which follow participants for a longer time be done to provide better evidence for the use of this combination as a first‐line therapy. A trial assessing fixed‐drug (providing drugs in a single tablet) is also required, as this reduces the number of pills people must take each day. These studies should include assessment of adverse effects, as well as tracking whether resistance to the drugs develop over time.

Background

A variety of combination therapy regimens are available to treat patients with human immunodeficiency virus (HIV). There are currently three classes of antiretroviral drugs in widespread use, although newer classes such as integrase inhibitors are just entering the market. These drugs are initially given in combinations of three or four drugs. Types of combination therapies include a single protease inhibitor (PI) combined with two nucleoside reverse transcriptase inhibitors (NRTI); one non‐nucleoside reverse transcriptase inhibitor (NNRTI) combined with two NRTIs; or three NRTIs. The choice of an initial regimen (also called first‐line therapy) is determined by various considerations. These include extent and acceleration of infection, pregnancy, cost (especially in low‐resource countries), drug tolerability, presence of drug‐resistant mutations in non‐treated populations, development of drug resistance and efficacy of the regime.

A favourable initial regimen is one that provides optimal efficacy, durability of antiretroviral activity, tolerability, and has low adverse effects and drug‐drug interactions (Yeni 2002). Convenience of the regimen in terms of the dosing schedule, the need for hydration, and timing with meals, are also considerations. In resource‐poor settings all of these factors are vital, as well as the cost of both the drug and drug monitoring for efficacy, and adverse drug reactions.

Efficacy can be measured by clinical parameters, such as occurrence of death, or occurrence of a new event (an acquired immunodeficiency syndrome [AIDS]‐defining illness); and by laboratory parameters, such as CD4+ lymphocyte count, or HIV‐RNA levels in the plasma (plasma viral load). The risk of clinical AIDS and death from HIV depends on both the rate and burden of viral infection, typically as determined by plasma viral load, and the degree of immunosuppression, as determined by the number or percentage of CD4+ lymphocytes in the peripheral blood (Castro 1992). Current World Health Organization guidelines for resource‐limited settings call for initiation of antiretroviral therapy when CD4 counts fall below 200 per mm³ or when symptoms of severe immunosuppression, such as opportunistic infections, appear (WHO 2004a). In other countries and regions, treatment can start earlier, such as when counts fall below 350 CD4⁺ cells per mm³ (DHHS 2005; BHIVA 2003; Ministerio 2004; SAHIV 2005).

PIs block the enzyme that is essential for HIV assembly in the cytoplasm of the host cell (BHIVA 2003). There are concerns about long‐term adverse effects of PIs, such as body fat changes (lipodystrophy), insulin resistance, and dyslipidemia. Short‐term effects (gastro‐intestinal intolerance and neuropsychiatric manifestations), and their cost, high pill burden, dietary constraints and frequent dosing schedule are also important (BHIVA 2003). There is thus added impetus to analyse regimens that do not use PIs in initial therapy (so‐called PI‐sparing regimens). One such combination is the two NRTI, one NNRTI regimen of stavudine (d4T), lamivudine (3TC) and nevirapine (NVP). We will refer to this combination as SLN.

Both d4T and 3TC are NRTIs. NRTIs as a class have side effects such as lactic acidosis, lipodystrophy and hepatic steatosis (Weller 2001). While lactic acidosis is potentially fatal, it is also rare, occurring in 0.1% of patients (SAHIV 2005). Peripheral neuropathy, pancreatitis and myelosuppression are also described (BHIVA 2003). The substitution of d4T for ZDV in a regimen was widely practiced, in the belief that it would retain activity if virus with resistance to ZDV was present (BHIVA 2003). This is now questioned, as there is evidence that thymidine analogue mutations within the virus are associated with reduced susceptibility to both d4T and ZDV (Collaboration 2004). Side effects of d4T include peripheral neuropathy and lipo‐atrophy, and a high potential for hyperlactaemia/steatohepatitis (SAHIV 2005). There are very few reported side effects of 3TC, usually occurring in late disease. These include anaemia, gastrointestinal upset, myalgia, and rarely pancreatitis and hyperlactaemia/steatohepatitis (SAHIV 2005). However, a mutation on the M184V gene confers high‐level resistance to 3TC (Yeni 2002).

NVP is a NNRTI (BHIVA 2003). The NNRTIs have a more favourable adverse‐effect profile than PIs, as well as good pharmacokinetic properties and ease of administration. NVP has been shown to have a favourable effect on lipids (decreasing cholesterol and increasing high density lipoproteins) in patients who discontinued PIs (Yeni 2002) because of lipodystrophy or dyslipidemia. However, a major disadvantage of this group of drugs is that a single gene mutation can cause high‐level resistance to the whole class of drugs (SAHIV 2000). Cross‐resistance between NVP and efavirenz (EFZ) is a major problem. Resistance to one usually means that the other drug will be ineffective (SAHIV 2005). In addition, the major side effect of NNRTIs, especially NVP, is a rash, which may manifest in severe form as a Stevens‐Johnson syndrome. Severe or fatal hepatotoxicity is also a problem (SAHIV 2005). In a study comparing regimens containing NVP or EFZ, development of hepatoxicity occurred in 17% (66/385) of patients on a NVP‐containing regimen versus 0% (0/83) of patients in the EFZ group (Sanne 2005). This hepatotoxicity occurred early: 80% within the first three months of therapy. NVP remains the chosen NNRTI for use in pregnancy, due to the teratogenic potential of EFZ, except when initiating triple therapy in the third trimester, when the risk to the fetus has passed (WHO 2004a).

During 2001, at least one pharmaceutical company announced it was willing to supply the combination of SLN at US$350 per patient per year, to a number of countries in Africa, at one‐thirteenth of prices then charged by other companies (UNAIDS 2001). The combination is easy to take as one tablet twice daily. Given the reduction in cost, the provision of this antiretroviral regime to people in most low‐income countries has thus become a possibility (Laurent 2004). Analysis of the efficacy, durability and tolerability of the SLN regimen is thus important to clinicians, consumers and policy‐makers living in both rich and poor countries.

Objectives

To examine the efficacy of the stavudine, lamivudine and nevirapine regimen in the treatment of HIV infection and AIDS in adults.

Methods

Criteria for considering studies for this review

Types of studies

Randomised controlled trials

Types of participants

Adults with HIV infection and/or AIDS, with or without prior exposure to antiretroviral therapy.

'Adult' is defined as age over 13 years (as used in the CDC Classification, which includes adolescents) (Castro 1992).

Criteria for HIV infection as for CDC classification (1993‐revised) (Castro 1992), and AIDS definitions by CDC classification: stages A3, B3, C1‐3 (see Table 1) or by WHO classification: Clinical stage 4 (see Table 2) (Grimwood 1998).

1. Clinical Categories.

CD4+ Cell Categories	(A) Asymptomatic,	(B) Symptomatic, not	(c) AIDS, indicator
	acute	(A) or (C) conditions	conditions
	HIV* or PGL**
(1) > 500/mm3	A1	B1	C1
(2) 200 ‐ 499/MM3	A2	B2	C2
(3) <200/MM3	A3	B3	C3

2. World Health Organization: Staging System for HIV Infection and Disase.

Clinical Stage 1	Clinical Stage 2	Clinical Stage 3	Clinical Stage 4
Acute retroviral infection	Weight loss < 10 % body weight	Weight loss > 10 % body weight	Extrapulmonary Tuberculosis
Asymptomatic	Minor mucocutaneous manifestations	Unexplained chronic diarrhoea >1 month	Candida ‐ oesophagus, trachea or bronchi
Persistent generalised lymphadenopathy	Herpes Zoster, within the last 5 years	Oral candidiasis	Pneumocystis carinii pneumonia
+/‐ performance scale 1 (asymptomatic, normal activity)	Recurrent upper respiratory infections	Vulvo‐vaginal candidiasis, chronic (> 1 month)	Herpes simplex ‐ oral >1 month
	+/‐ performance scale 2 (symptomatic, normal activity)	Oral hairy leukoplakia	Herpes simplex‐ visceral, any duration
		Pulmonary Tuberculosis, within past year	HIV wasting syndrome
		Severe bacterial infections	Kaposi's sarcoma
		+/‐ performance scale 3 (bedridden < 50 % of the day)	Toxoplasmosis of the brain
			Cryptococcosis, extrapulmonary
			HIV encephalopathy
			Non‐Typhoid Salmonella bacteraemia
			Cytomegalovirus infection
			Cryptosporidiosis‐diarrhoea > 1 month
			Progressive multifocal leuko‐encephalopathy
			Lymphoma
			Atypical mycobacteriosis, disseminated
			Any disseminated endemic mycosis
			performance scale 4 (bedridden > 50 % of last month)

Types of interventions

Studies comparing the SLN regimen either to placebo or to any other antiretroviral regime that is available to clinicians.

Types of outcome measures

Primary Outcomes:

‐ Death (all cause)

‐ Occurrence of new event (death or AIDS‐defining illness)

‐ Proportion of patients achieving and maintaining undetectable viral load, e.g. less than 50 copies/ml

Secondary Outcomes:

‐ CD4+ count: mean relative change (percent) or mean absolute change, compared with baseline mean, and standard deviation.

‐ HIV‐RNA levels: mean relative change (percent) or mean absolute change, compared with baseline mean, and standard deviation. 
 
 ‐ Quality of life indicators as reported in the studies. 
 
 ‐ Any adverse events. These will be reported in two broad categories as severe or minor. Severe is such that therapy needs to be withdrawn. Minor will be grouped according to organ system e.g. skin, cardiac, respiratory, etc.

Search methods for identification of studies

See: HIV/AIDS Collaborative Review Group search strategy.

With the assistance of the HIV/AIDS Review Group Trials Search Co‐ordinator, we formulated a comprehensive and exhaustive search strategy in an attempt to identify all relevant studies regardless of language or publication status (published, unpublished, in press, and in progress). Full details of the Cochrane HIV/AIDS Review Group methods and the journals hand‐searched are published in The Cochrane Library in the section on Collaborative Review Groups (http://www.mrw.interscience.wiley.com/cochrane/clabout/articles/HIV/frame.html).

We used the RCT search strategy developed by the Cochrane Collaboration and detailed in the Cochrane Reviewers' Handbook, in combination with terms specific to SLN. We searched the following electronic databases:

(1) MEDLINE (1966‐2003) via PubMed in August 2003, updated in February 2004, and again in July 2004, using the following strategy:

#1 STAVUDINE 
 #2 LAMIVUDINE 
 #3 NEVIRAPINE 
 #4 #1 OR #2 OR #3 
 #5 HIV 
 #6 HIV INFECT* 
 #7 HIV‐1 
 #8 HIV‐2 
 #9 ACQUIRED IMMUNODEFICIENCY SYNDROME 
 #10 #5 OR #6 OR #7 OR #8 OR #9 
 #11 RANDOMIZED CONTROLLED TRIAL[PT] OR CONTROLLED CLINICAL TRIAL[PT] OR RANDOMIZED CONTROLLED TRIALS[MH] OR RANDOM ALLOCATION[MH] 
 #12 DOUBLE‐BLIND METHOD[MH] OR SINGLE‐BLIND METHOD[MH] OR TRIPLE‐BLIND METHOD[MH] OR CLINICAL TRIAL[PT] OR (CLINICA*[TW] AND TRIAL*[TW]) 
 #13 ((SINGL* OR DOUBL* OR TRIPL* OR TREBL*) AND (MASK* OR BLIND*)) 
 #14 PLACEBO[MH] OR PLACEBO* OR RANDOM* 
 #15 #11 OR #12 OR #13 OR #14 
 #16 #4 AND #10 AND #15 
 #17 ANIMAL NOT HUMAN 
 #18 #16 NOT #17 
 #19 (ADOLESCENCE OR ADULT OR MIDDLE AGED OR AGED OR (AGED 80)) AND #18 
 #20 #16 NOT #17 Field: All Fields, Limits: Human 
 #21 #16 NOT #17 Field: All Fields, Limits: Publication Date from 2000/01/01 to 2004/01/31, Human

This yielded 780 records in total of which we selected three for full article retrieval.

(2) EMBASE (2000‐2004) in February 2004 and EMBASE (1966‐2000) in October 2004, using the PubMed strategy, modified for EMBASE.

This yielded 342 records in total, of which we selected 10 for full article retrieval.

(3) AIDSearch (1995‐2005) in April 2005, which includes coverage of the following conferences: 
 ‐ International AIDS Conference (1985‐2004) 
 ‐ Conference on Retroviruses and Opportunistic Infections (1986‐2004) 
 ‐ The British HIV Association conference (1997‐2003) 
 ‐ International Congress on Drug Therapy in HIV infection (1994‐2002)

This yielded 91 records in total, of which we selected seven for full article retrieval.

(4) The Cochrane Library Controlled Trials Register in February 2004, and updated in July 2004. This contains mainly reference information to randomised controlled trials and controlled clinical trials in health care. This yielded 637 records in total, of which we selected one for full article retrieval.

There was some overlap between the references retrieved in each database. We also checked the reference lists of all the articles retrieved by the search strategy for relevant studies. Finally, we contacted research organizations and experts in the field for unpublished and ongoing studies, and have included studies for assessment that were not retrieved by our search.

Data collection and analysis

NS and PVD independently conducted the selection of potentially relevant studies by scanning the titles, abstracts, and descriptor terms of all downloaded material from the electronic searches. Irrelevant reports were discarded, and the full article was obtained for all potentially relevant or uncertain reports.

1. Selection of studies: 
 NS and PVD independently applied the inclusion criteria, using an eligibility form specific to this review. FM acted as arbiter where there was disagreement. Studies were reviewed for relevance, based on study design, types of participants, exposures and outcome measures. Finally, where resolution was not possible because further information was required, the study was allocated to the list of those awaiting assessment. Attempts to contact authors to provide further clarification of data are ongoing.

2. Data extraction: 
 NS and PVD independently extracted data using a standardised data extraction form. The following characteristics were extracted from each included study: 
 ‐ Administrative details: identification; author(s); published or unpublished; year of publication; number of studies included in paper; year in which study was conducted; details of other relevant papers cited. 
 ‐ Details of study: study design; type, duration and completeness of follow‐up; country and location of the study (developed versus developing country). 
 ‐ Characteristics of participants: prior exposure to antiretroviral therapy (type and duration); disease stage; baseline CD4 count; baseline HIV‐RNA level. 
 ‐ Details of intervention: types and doses of drugs used; duration of therapy; adverse events (method of surveillance for adverse events and type: minor or severe) and adherence measures (if reported). 
 ‐ Outcome data: death or occurrence of new event (death or AIDS‐defining event); subsequent and final CD4 counts and HIV‐RNA levels dichotomised as having reached, or as not having reached, a non‐detectable level, i.e. less than 50 copies/ml.

3. Quality assessment: 
 The quality of trial methods was qualitatively captured by recording the method of generating the randomisation sequence, the adequacy of allocation concealment, the adequacy of blinding in the participants, providers and assessors, and the differential loss‐to‐follow‐up in the comparison groups. We did not attempt to quantify trial quality, but described it in full.

4. Data synthesis: 
 As the two trials included in the review did not measure the same interventions, no meta‐analysis was conducted and the results of the trials are presented separately.

For dichotomous data we calculated the overall measure of effect as a relative risk, with 95% confidence intervals, using the random effects model. We used an intention‐to‐treat analysis (ITT) throughout. Our definition of ITT includes analysing participants in their original randomised groups, and also uses the initial number of randomised participants per group as the denominator. Where appropriate and where results were significant, we also calculated the number needed to harm (NNTH) or benefit (NNTB) and the 95% confidence intervals (Altman 1998).

We analysed continuous data, using the weighted mean difference and standard deviation. Where possible, we also used an ITT analysis, but if data were missing we used the actual number of participants included in the calculation of the mean.

As we were not able to conduct a meta‐analysis, we were not able to explore subgroups, as had been the intention described in our protocol.

Results

Description of studies

We reviewed 12 studies, of which two trials met the inclusion criteria (French 2002; Van Leth 2004).

The French 2002 trial was a small trial (N= 70) conducted in Australia, whereas the Van Leth 2004 trial was a large multicentre, multinational trial of 1216 participants. Both trials only included participants who were antiretroviral‐naive and participants were predominantly male (over 60% in both trials). In the Van Leth 2004 trial, the percentage of participants with CD4+ cell counts < 50/microL ranged from 13.4 to 20.4% in the four different treatment groups. Figure 1 shows the country location of each trial, including the country location for each centre in the Van Leth 2004 trial.

1.

Country location of Stavudine, Lamivudine and Nevirapine trials, given by central geographic co‐ordinates.

Five of the excluded studies were trials that did not assess the drugs in the combination of SLN (Garcia 2003; Lafeuillade 2001; Raboud 2002; Ruiz 2001; Sabo 2002); one study was neither randomised nor controlled (Laurent 2004); one study was a systematic review (Yazdanpanah 2004); and one study was a pharmacokinetic study of lamivudine in non‐HIV infected participants (Yuen 2004). For the remaining two trials, there was insufficient reported information to establish whether sub‐groups within the trials received SLN (Katlama 2000; Negredo 2002) We have contacted the authors of these trials and will include the data from these subgroups in future updates of the review, should they prove to be relevant.

Risk of bias in included studies

GENERATION OF ALLOCATION SEQUENCE 
 The French 2002 trial did not provide an adequate description of the sequence of generation, but was described as randomised. The equal allocation of patients suggests that a block randomisation process was used. Generation for the Van Leth 2004 trial used minimisation based on the CD4 count, which is an appropriate method of randomisation. The trial began with three comparison groups, but a fourth was added after five months. The effect of the change in randomisation after the addition of the fourth comparison group was assessed by logistic regression analysis. The authors report that there was no evidence for such an effect (OR = 1.01, p = 0.868). For 68% of the study, sample allocation could have been to one of four groups.

ALLOCATION CONCEALMENT 
 The French 2002 trial did not provide sufficient detail to describe the allocation concealment process, and, as the trial was open‐label, it may not have been adequately concealed. The Van Leth 2004 trial used a centralised process of allocating treatment groups, so allocation concealment was adequate.

BLINDING 
 There was no blinding in the French 2002 trial, as it was open‐label. There was no blinding of the participants or providers in the Van Leth 2004 trial. Assessors were also unblinded, as treatment change was a component of the composite outcome, which required clinicians to be aware of the assigned grouping.

LOSS TO FOLLOW‐UP 
 There was a greater than 20% loss‐to‐follow‐up in the French 2002 trial, with 32% not completing 52 weeks of treatment. In the Van Leth 2004 trial there was an overall 16% loss‐to‐follow‐up, with no differential loss‐to‐follow‐up between comparison groups.

Effects of interventions

As the two trials eligible for inclusion did not report on the same drug comparisons, the results are presented separately for each trial. For each trial, the primary and secondary outcomes and adverse events are reported for each of the comparison groups, where these data are available.

French 2002 
 This trial compared the SLN regimen with two other triple‐drug regimens in antiretroviral‐naive patients. There were no reported clinical outcomes, as the primary outcomes were all related to viral load. Real‐time plasma was tested using the Roche Amplicor version 1.0, with a lower limit of plasma HIV RNA of 500 copies/ml. After the study was completed, stored plasma was analysed using the Roche Amplicor Version 1.5, a more sensitive test with a lower detection limit of plasma HIV RNA of 50 copies/ml, which only became available after the trial was completed.

PRIMARY OUTCOMES

• · Twice‐daily stavudine 40mg (or adjusted if body weight < 60kg) + lamivudine 150mg + nevirapine 200mg (d4T, 3TC, NVP); compared with twice‐daily zidovudine 250mg + lamivudine 150mg + nevirapine 200mg (ZDV, 3TC, NVP):

The weighted mean difference in the time‐weighted mean reduction from baseline to week 52 of real‐time plasma HIV RNA (in log copes/ml) was 0.13 (95%CI: ‐0.52, 0.78) more in the ZDV, 3TC, NVP comparison group and was not statistically significant (p = 0.69). The weighted mean difference in the time‐weighted mean change from baseline to week 52 in stored plasma HIV RNA (in log copies/ml) was ‐0.87 (95%CI: ‐1.68, ‐0.06). This means that those taking d4T, 3TC, NVP had a greater reduction in stored plasma viral load compared with the ZDV, 3TC, NVP group. This was a statistically significant finding (p = 0.04). The weighted mean difference in the time‐weighted mean change from baseline to week 52 of CD4 count in cells/uL was ‐26 cells/uL (95%CI: ‐106.71, 54.71) which means that CD4 count increased more in the ZDV, 3TC, NVP comparison group. This was not significant (p = 0.53). The RR for undetectable real‐time plasma HIV RNA < 500 copies/ml was 1.12 (95%CI: 0.73,1.73) for those patients in the d4T, 3TC, NVP group compared with the ZDV group. These patients were slightly more likely to reach undetectable viral loads than those in the ZDV group, but this was not a statistically significant finding (p = 0.61). The RR for undetectable stored plasma HIV RNA < 50 copies/ml was 1.24 (95%CI: 0.71,2.14) for those patients in the d4T, 3TC, NVP group and was not statistically significant (p = 0.45).

• · ·Twice‐daily stavudine 40mg (or adjusted if body weight < 60kg) + lamivudine 150mg + nevirapine 200mg (d4T, 3TC, NVP); compared with twice‐daily didanosine 200mg + stavudine 40mg (or adjusted if body weight < 60kg) + nevirapine 200mg (ddI, d4T, NVP):

The weighted mean difference in the time‐weighted mean reduction from baseline to week 52 of real‐time plasma HIV RNA (in log copies/ml) was 0.56 (95%CI: 0.00, 1.12) more in the d4T, ddI, NVP comparison group, and was of borderline significance (p = 0.05). The weighted mean difference in the time‐weighted mean change from baseline to week 52 of stored plasma HIV RNA (in log copies/ml) was ‐0.38 (95%CI: ‐0.90, 0.14). This means that those taking d4T, 3TC, NVP had a greater reduction in stored plasma viral load compared with the d4T, ddI, NVP group. This was not a statistically significant finding (p = 0.16). The weighted mean difference in the time‐weighted mean change from baseline to week 52 in CD4 count in cells/uL was ‐61 cells/uL (95%CI: ‐136.40, 14.40) which means that CD4 count increased more in the d4T, ddI, NVP comparison group. This was not statistically significant (p = 0.11). The RR for undetectable real‐time plasma HIV RNA < 500 copies/ml was 0.85 (95%CI: 0.60,1.20) for those patients in the d4T, 3TC, NVP group compared with the d4T, ddI, NVP group. These patients were less likely to reach undetectable viral loads than those in the ddI group, but this was not a statistically significant finding (p = 0.37). The RR for undetectable stored plasma HIV RNA < 50 copies/ml was 0.88 (95%CI: 0.56,1.40) for those patients in the d4T, 3TC, NVP group and was also not statistically significant (p = 0.60).

SECONDARY OUTCOMES 
 Quality of life indicators measured by both patient and physician are reported as being high at the beginning of the study (greater than or equal to 80 but scale not described), and remaining high throughout the study, with no statistically significant differences reported between any of the treatment groups (data not reported).

ADVERSE EVENTS 
 The authors classified drug‐related adverse events using the Adverse Event Toxicity Scale, graded from Grade 1 to 4. Grades 1 and 2 denote mild to moderate symptoms; Grade 3 denotes serious symptoms (incapacitating, bed rest, loss of work, decrease in social activities); and Grade 4 are serious and life‐threatening adverse events requiring significant clinical intervention. Overall 19 patients experienced Grade 3 or 4 adverse events: 4/30 in the ZDV +3TC + NVP group, 8/22 in the d4T + 3TC + NVP group and 7/23 in the d4T + ddI + NVP group. Of these, 14 patients were considered to have suffered drug‐related adverse events: four in the ZDV +3TC + NVP group, five in the d4T + 3TC + NVP group and five in the d4T + ddI + NVP group. There were 26 drug‐related adverse events in total (patients could have more than one event while on treatment). In our protocol we classified severe adverse events as those requiring cessation of treatment. In this study, ten patients ceased treatment due to severe adverse events: 3/30 in the ZDV +3TC + NVP group, 4/22 in the d4T + 3TC + NVP group and 3/23 in the d4T + ddI + NVP group (no p‐value reported). Five of these adverse events were due to neuropathy, developing at or after week 12 (3/22 in the d4T = 3TC group and 2/23 in the d4T and ddI group). Three patients ceased NVP due to the development of a rash (one patient in each group), with cessation definitely attributed to NVP in only one of these patients (the group is not specified). One patient in the ZDV + 3TC + NVP group ceased therapy due to elevated liver function tests, and one patient ceased therapy in the same group for an unspecified reason. Minor complaints included gastrointestinal symptoms (nausea, anorexia, abdominal pain), neurological symptoms (headache, anxiety, and other), general symptoms (fever/chills, malaise, tiredness, hypersensitivity) and elevated amylase and liver function tests.

Van Leth 2004 
 This was initially designed as a three‐arm trial that compared d4T + 3TC + once‐daily NVP with d4T + 3TC + EFZ , and d4T + 3TC + EFZ + NVP. After five months, an additional group assigned to NVP twice daily was added, after a study had found that the efficacy of NVP was related to minimum concentration, and raised a question around the once‐daily dosing (Veldkamp 2001). The effect of the change in randomisation after the addition of the fourth comparison group was assessed by logistic regression analysis. The authors report that there was no evidence for such an effect (OR = 1.01, p = 0.868), and so conducted a pooled efficacy analysis, despite numerical differences between the two randomisation periods. We present the results here for the entire trial period. The primary outcome in this trial was a composite endpoint comprising three components: viral load, disease progression (including death) and therapy change. There were 527 treatment failures, of which 193 (36.6%) were virologic; 33 (6.3%) were disease progression; and 301 (57.1%) were treatment changes, of which 200/301 (66.4%) were permanent changes of NNRTIs. We also report on death as a primary outcome.

PRIMARY OUTCOMES

• · ·Twice‐daily stavudine 40mg (or 30mg if body weight < 60kg) and lamivudine 150mg, and once‐daily nevirapine 400mg (d4T, 3TC, NVP od); versus twice‐daily stavudine 40mg (or 30mg if body weight < 60kg) and lamivudine 150mg, and twice‐daily nevirapine 200mg (d4T, 3TC, NVP bd)

There was no statistically significant difference in treatment failure between these comparison groups (RR = 1.00; 95%CI: 0.83, 1.21). Providing NVP once or twice daily did not produce different effects. There was no statistically significant difference between the proportion of patients reaching undetectable plasma HIV RNA < 50 copies/ml at 48 weeks (RR = 1.09; 95%CI: 0.98,1.22). There was no statistically significant difference in deaths between the two groups (RR = 1.37; 95%CI: 0.52, 3.62; p = 0.53).

• ··Twice‐daily stavudine 40mg (or 30mg if body weight < 60kg), lamivudine 150mg twice‐daily, and once‐daily nevirapine 400mg (d4T, 3TC, NVP od); versus twice‐daily stavudine 40mg (or 30mg if body weight < 60kg), lamivudine 150mg twice‐daily, and once‐daily efavirenz 600mg (d4T, 3TC, EFZ)

There was no statistically significant difference in treatment failure between these comparison groups (RR = 1.16; 95%CI: 0.95, 1.41). Once‐daily NVP and EFZ provide similar efficacy profiles. There was no statistically significant difference between the proportion of patients reaching undetectable plasma HIV RNA < 50 copies/ml at 48 weeks (RR = 1.00; 95%CI: 0.90,1.11). There was no statistically significant difference in deaths between the two groups (RR = 1.82; 95%CI: 0.65, 5.12; p = 0.26).

• · ·Twice‐daily stavudine 40mg (or 30mg if body weight < 60kg) and lamivudine 150mg, and once‐daily nevirapine 400mg (d4T, 3TC, NVP od); versus twice‐daily stavudine 40mg (or 30mg if body weight < 60kg) and lamivudine 150mg, and once‐daily efavirenz 800 mg and nevirapine 400mg (d4T, 3TC, EFZ & NVP).

The RR was 0.82 (95%CI; 0.67, 1.00) which is interpreted as the d4t, 3TC, od NVP combination reducing the likelihood of treatment failure in comparison to d4T + 3TC + NVP + EFZ. The finding was marginally significant (p = 0.05). There was no statistically significant difference between the proportion of patients reaching undetectable plasma HIV RNA < 50 copies/ml at 48 weeks (RR = 1.12; 95%CI: 0.98,1.28). There was no statistically significant difference in deaths (RR = 3.33; 95%CI: 0.70; 15.82; p = 0.13).

• · ·Twice‐daily stavudine 40mg (or 30mg if body weight < 60kg) and lamivudine 150mg, and twice‐daily nevirapine 200mg (d4T + 3TC + bd NVP); versus twice‐daily stavudine 40mg (or 30mg if body weight < 60kg) and lamivudine 150mg, and once‐daily efavirenz 600mg (d4T + 3TC + EFZ)

There was no statistically significant difference in treatment failure between these comparison groups (RR = 1.16; 95%CI: 0.98, 1.37). Nevirapine and efavirenz provide similar efficacy profiles, regardless of single or twice‐daily dosing of nevirapine. There was no statistically significant difference between the proportion of patients reaching undetectable plasma HIV RNA < 50 copies/ml at 48 weeks (RR = 0.93; 95%CI: 0.85,1.03). There was no statistically significant difference in deaths (RR = 1.33; 95%CI: 0.50, 3.53; p = 0.57).

• · ·Twice‐daily stavudine 40mg (or 30mg if body weight < 60kg) and lamivudine 150mg, and twice‐daily nevirapine 200mg (d4T + 3TC + bd NVP); versus twice‐daily stavudine 40mg (or 30mg if body weight < 60kg) and lamivudine 150mg, and once‐daily efavirenz 800 mg and nevirapine 400mg (d4T + 3TC + EFZ + NVP)

The RR was 0.82 (95%CI: 0.69, 0.97) indicating that the d4t + 3TC + bd NVP combination reducing the likelihood of treatment failure by 18% in comparison to d4T + 3TC + NVP + EFZ. The finding was significant (p = 0.02). Provision of twice‐daily nevirapine dosing produces an improved efficacy profile than when combined as a daily dose with efavirenz. The number needed to treat to benefit (NNTB) one patient to avert treatment failure was 11 patients. The 95%CI showed a NNTB at best of 7 but at worst, of 63. There was no statistically significant difference between the proportion of patients reaching undetectable plasma HIV RNA < 50 copies/ml at 48 weeks (RR = 1.04; 95%CI: 0.92,1.18). There was no statistically significant difference in deaths (RR = 2.92; 95%CI: 0.64, 13.38; p = 0.17).

ADVERSE EVENTS 
 The authors also classified clinical adverse events using the Adverse Event Toxicity Scale. In the group assigned once‐daily NVP, 33 (15%) patients had at least one grade 3 or 4 clinical adverse event. In the group assigned twice‐daily NVP 79 (20.4%) patients suffered a Grade 3 or 4 adverse event. Seventy‐two (28%) of those assigned efavirenz and 51 (24.4%) of those assigned NVP plus EFZ had a Grade 3 or 4 clinical adverse event. There was a reported statistically significant difference (p = 0.014) between the frequency of Grade 3 or 4 adverse events in the NVP plus EFV group compared with the NVP once‐daily group, but not statistically significant different between the NVP plus EFZ group and the EFZ group (p = 0.062). Adverse events affected the following systems: hepatobiliary, cutaneous, central‐nervous system/psychiatric, gastrointestinal and general. Severe peripheral neuropathy occurred in 2.7% of patients in the NVP once‐daily group, 3.1% in the NVP twice‐daily group, 2.3% in the EFV group and 3.8% in the NVP plus EFV group.

There was a statistically significant higher rate of adverse events leading to temporary or permanent discontinuation of the study drugs in the NVP plus EFZ group compared with the EFZ group at the predetermined p < 0.0125 level (p < 0.0001). Rates were also significantly higher in the NVP once‐daily group compared with the EFZ group (p = 0.011). Overall, three patients died from events attributed to medication. Of two patients assigned to the twice‐daily nevirapine group, one developed fulminant hepatitis attributed to nevirapine and later died, and the other developed Stevens‐Johnson Syndrome, later developing septicaemia and dying. One patient in the d4T + 3TC + od NVP group died of lactic acidosis, which was attributed to the use of stavudine. Rash leading to temporary or permanent withdrawal of study drugs was reported to be 12.3% for NVP once‐daily, compared with 6.5% with NVP twice‐daily, 3.8% with EFV and 13.9% in NVP plus EFV.

Grade 3/4 liver associated laboratory toxicity occurred in 13.6% of patients on NVP once‐daily, compared with 8.3% on NVP twice‐daily, 4.5% on EFV, and 9.1% on NVP plus EFV. There was a higher incidence of hepatitis B co‐infection in the NVP once‐daily group, which might have contributed to the increased liver toxicity in the NVP once‐daily group. The frequency of toxicity between NVP once‐daily and EFZ was statistically significant (p < 0.0001) at the predetermined significance level of p < 0.0125. Non‐hepatobiliary laboratory toxicity occurred in 9.1% of the group assigned once‐daily NVP; in 13.7% of the group assigned twice‐daily NVP; in 10.3% of those assigned EFZ; and in 10.5% of the NVP plus EFV group. There were no significant differences between these groups.

Discussion

The most noteworthy finding of this review is that only two trials have assessed the efficacy of the SLN drug combination. One of these trials (French 2002) had a very small sample size, which considerably limits the conclusions that can be drawn from its results. As the therapeutic combinations compared in both trials were not identical, it was not possible to conduct a meta‐analysis to increase the power of these results. The main findings therefore are from the much larger Van Leth 2004 trial. The trial found that there was no statistically significant difference in the efficacy between NVP and EFZ when used in combination with 3TC and d4T. It also showed that both once‐daily and twice‐daily dosing of NVP, in combination with the NRTIs 3TC and d4T, performed significantly better than NVP plus EFZ, 3TC and d4T.

SLN is the mainstay of the World Health Organization's guidelines for antiretroviral use in so‐called resource‐constrained settings (WHO 2004a). A recent open‐label non‐controlled trial conducted in Cameroon reported that at the end of 24 weeks of treatment, 80% of the 60 participants on SLN treatment had an undetectable viral load, defined as < 400 copies/ml (Laurent 2004). The generic fixed‐dose SLN combination was well tolerated, with only one participant developing a cutaneous reaction to NVP requiring withdrawal of treatment. Despite these optimistic findings, the authors recommend that additional controlled trials with longer duration of follow‐up are required, especially with respect to use of generic SLN in resource‐constrained settings. The findings of our review support this recommendation.

As stated earlier, the French 2002 trial is limited by a small sample size. Although it is unclear from the reporting, allocation concealment may not have been adequate, and as the trial was open‐label, there was no masking. Lack of adequate allocation concealment has been shown to be significantly associated with exaggerated treatment effects (Schulz 1995). Plasma was tested during the trial, and then again after the trial, when a more sensitive test became available. This dual testing did not result in different findings. This highlights the fact that there is no standard method for presenting HIV RNA data from clinical trials. Standardisation is essential for reliable comparison of treatment effects in different trials when using meta‐analysis (Hill 1999).

The Van Leth 2004 trial was well reported, with a relatively low loss‐to‐follow‐up (16%) for a trial of this duration. However, the change of randomisation from three to four comparison groups may affect the overall estimates in this trial. The primary outcome was a composite endpoint comprising three components: viral load, disease progression and therapy change. Caution is required when interpreting composite endpoints (Freemantle 2003), as the results may only apply to the individual components, rather than to the overall composite. Van Leth and colleagues have reported the results for the individual components to overcome this (Van Leth 2003). They acknowledge that including therapy change as part of the component outcome may have introduced selection bias (Van Leth 2004).

Although the rate of clinical adverse events are similar for the groups tested by Van Leth 2004, other features, such as events leading to temporary or permanent discontinuation of the study drugs, and individual toxicities, may favour the EFV group over the others. These results are supported by those of a recent study comparing regimens containing NVP or EFZ, in which development of hepatoxicity occurred in 17% (66/385) of patients on a NVP‐containing regimen, versus 0% (0/83) of patients in the EFZ group (Sanne 2005). For a NVP‐based regimen, clinicians may favour twice‐daily over once‐daily dosing as a result of Van Leth's findings, despite the increased dosing schedule. Clearly, more studies that incorporate measuring the adverse event rates are required, to better inform the choice of NVP versus EFZ as part of a first regimen.

Neither trial reported on the emergence of viral resistance, an important consideration in choosing a first‐line regimen. The rate of development of genotypic resistance mutations in the Cameroonian open label trial of the SLN combination was found to be 7.1 per 100 person years (95% CI: 1.8; 28.4). The study was of only 24 weeks duration, however, and therefore a relatively short follow‐up period (Laurent 2004). Future trials and cohort studies to assess the efficacy and safety of SLN should include measuring viral resistance.

Lastly, many poor countries use generic versions of SLN, either as individual tablets or as fixed‐dose combinations. Concerns were raised about the quality of generic antiretroviral drugs in May 2004 when the WHO de‐listed generic tablets for failing to meet international bioequivalence standards. Provision of HIV drugs that are not only efficacious and safe, but also of high quality and affordable, remains a central challenge to clinicians working in settings where cost limits access to health care (Havlir 2005).

Authors' conclusions

Implications for practice.

The combination of SLN is as efficacious as a combination of EFV, 3TC and d4T. Nevirapine used as a once‐daily dose is as efficacious as a twice‐daily dose in the SLN combination. This will allow clinicians to reduce pill count in order to promote adherence. However, toxicity may be increased in the once‐daily NVP regime, when compared to the NVP twice‐daily, as seen in the Van Leth trial. Clinicians may prefer a lower toxicity profile to a slightly lower pill burden, but this is a decision that needs to be taken on an individual basis.

Implications for research.

Additional trials of sufficient duration are required to provide better evidence for the use of SLN as a first line therapy. Ideally, trials should use standardised assessment measures, especially with respect to measuring viral load, so that results can be compared and combined in meta‐analyses. Assessment of fixed‐drug dosing is required in the context of a trial. Clinical studies assessing the rate of adverse events, and resistance of the SLN regimen, continue to be required to better inform practice.

What's new

Date	Event	Description
5 September 2011	Review declared as stable	Had to specify this again, as new event was added.
5 September 2011	Amended	Minor amendment to plain language summary title, although review will not be updated.

History

Protocol first published: Issue 4, 2003
 Review first published: Issue 2, 2006

Date	Event	Description
16 March 2011	Review declared as stable	This review will no longer be updated.
30 October 2008	Amended	Converted to new review format.
22 February 2006	New citation required and conclusions have changed	Substantive amendment

Data and analyses

Comparison 1. d4T, 3TC, NVP versus ZDV, 3TC, NVP.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Time‐weighted mean reductions in real‐time plasma HIV RNA in lg copies/ml	1	40	Mean Difference (IV, Random, 95% CI)	0.13 [‐0.52, 0.78]
2 Time‐weighted mean reductions in stored plasma HIV RNA in lg copies/ml	1	34	Mean Difference (IV, Random, 95% CI)	‐0.87 [‐1.68, ‐0.06]
3 Time‐weighted mean change from baseline in CD4+ cell count in cells/uL	1	42	Mean Difference (IV, Random, 95% CI)	‐26.0 [‐106.71, 54.71]
4 Undetectable real‐time plasma HIV RNA < 500cop/ml	1	45	Risk Ratio (M‐H, Random, 95% CI)	1.12 [0.73, 1.73]
5 Undetectable stored plasma HIV RNA < 50 cop/ml	1	45	Risk Ratio (M‐H, Random, 95% CI)	1.24 [0.71, 2.14]

1.1. Analysis.

Comparison 1 d4T, 3TC, NVP versus ZDV, 3TC, NVP, Outcome 1 Time‐weighted mean reductions in real‐time plasma HIV RNA in lg copies/ml.

1.2. Analysis.

Comparison 1 d4T, 3TC, NVP versus ZDV, 3TC, NVP, Outcome 2 Time‐weighted mean reductions in stored plasma HIV RNA in lg copies/ml.

1.3. Analysis.

Comparison 1 d4T, 3TC, NVP versus ZDV, 3TC, NVP, Outcome 3 Time‐weighted mean change from baseline in CD4+ cell count in cells/uL.

1.4. Analysis.

Comparison 1 d4T, 3TC, NVP versus ZDV, 3TC, NVP, Outcome 4 Undetectable real‐time plasma HIV RNA < 500cop/ml.

1.5. Analysis.

Comparison 1 d4T, 3TC, NVP versus ZDV, 3TC, NVP, Outcome 5 Undetectable stored plasma HIV RNA < 50 cop/ml.

Comparison 2. d4T, 3TC, NVP versus d4T, ddI, NVP.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Time‐weighted mean reductions in real‐time plasma HIV RNA in lg copies/ml	1	44	Mean Difference (IV, Random, 95% CI)	0.56 [0.00, 1.12]
2 Time‐weighted mean reductions in stored plasma HIV RNA in lg copies/ml	1	39	Mean Difference (IV, Random, 95% CI)	‐0.38 [‐0.90, 0.14]
3 Time‐weighted mean change from baseline in CD4+ cell count in cells/uL	1	45	Mean Difference (IV, Random, 95% CI)	‐61.0 [‐136.40, 14.40]
4 Undetectable real‐time plasma HIV RNA < 500cop/ml	1	47	Risk Ratio (M‐H, Random, 95% CI)	0.85 [0.60, 1.20]
5 Undetectable stored plasma HIV RNA < 50 cop/ml	1	48	Risk Ratio (M‐H, Random, 95% CI)	0.88 [0.56, 1.40]

2.1. Analysis.

Comparison 2 d4T, 3TC, NVP versus d4T, ddI, NVP, Outcome 1 Time‐weighted mean reductions in real‐time plasma HIV RNA in lg copies/ml.

2.2. Analysis.

Comparison 2 d4T, 3TC, NVP versus d4T, ddI, NVP, Outcome 2 Time‐weighted mean reductions in stored plasma HIV RNA in lg copies/ml.

2.3. Analysis.

Comparison 2 d4T, 3TC, NVP versus d4T, ddI, NVP, Outcome 3 Time‐weighted mean change from baseline in CD4+ cell count in cells/uL.

2.4. Analysis.

Comparison 2 d4T, 3TC, NVP versus d4T, ddI, NVP, Outcome 4 Undetectable real‐time plasma HIV RNA < 500cop/ml.

2.5. Analysis.

Comparison 2 d4T, 3TC, NVP versus d4T, ddI, NVP, Outcome 5 Undetectable stored plasma HIV RNA < 50 cop/ml.

Comparison 3. Nevirapine once daily versus nevirapine twice daily.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Treatment failure on or before week 48	1	607	Risk Ratio (M‐H, Random, 95% CI)	1.00 [0.83, 1.21]
2 Undetectable plasma HIV RNA < 50 copies/ml at 48 weeks	1	607	Risk Ratio (M‐H, Random, 95% CI)	1.09 [0.98, 1.22]
3 Death	1	607	Risk Ratio (M‐H, Random, 95% CI)	1.37 [0.52, 3.62]

3.1. Analysis.

Comparison 3 Nevirapine once daily versus nevirapine twice daily, Outcome 1 Treatment failure on or before week 48.

3.2. Analysis.

Comparison 3 Nevirapine once daily versus nevirapine twice daily, Outcome 2 Undetectable plasma HIV RNA < 50 copies/ml at 48 weeks.

3.3. Analysis.

Comparison 3 Nevirapine once daily versus nevirapine twice daily, Outcome 3 Death.

Comparison 4. Nevirapine once daily versus efavirenz.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Treatment failure on or before week 48	1	620	Risk Ratio (M‐H, Random, 95% CI)	1.16 [0.95, 1.41]
2 Undetectable plasma HIV RNA < 50 copies/ml at 48 weeks	1	620	Risk Ratio (M‐H, Random, 95% CI)	1.0 [0.90, 1.11]
3 Death	1	620	Risk Ratio (M‐H, Random, 95% CI)	1.82 [0.65, 5.12]

4.1. Analysis.

Comparison 4 Nevirapine once daily versus efavirenz, Outcome 1 Treatment failure on or before week 48.

4.2. Analysis.

Comparison 4 Nevirapine once daily versus efavirenz, Outcome 2 Undetectable plasma HIV RNA < 50 copies/ml at 48 weeks.

4.3. Analysis.

Comparison 4 Nevirapine once daily versus efavirenz, Outcome 3 Death.

Comparison 5. Nevirapine once daily versus nevirapine and efavirenz.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Treatment failure on or before week 48	1	429	Risk Ratio (M‐H, Random, 95% CI)	0.82 [0.67, 1.00]
2 Undetectable plasma HIV RNA < 50 copies/ml at 48 weeks	1	429	Risk Ratio (M‐H, Random, 95% CI)	1.12 [0.98, 1.28]
3 Death	1	429	Risk Ratio (M‐H, Random, 95% CI)	3.33 [0.70, 15.82]

5.1. Analysis.

Comparison 5 Nevirapine once daily versus nevirapine and efavirenz, Outcome 1 Treatment failure on or before week 48.

5.2. Analysis.

Comparison 5 Nevirapine once daily versus nevirapine and efavirenz, Outcome 2 Undetectable plasma HIV RNA < 50 copies/ml at 48 weeks.

5.3. Analysis.

Comparison 5 Nevirapine once daily versus nevirapine and efavirenz, Outcome 3 Death.

Comparison 6. Nevirapine twice daily versus efavirenz.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Treatment failure on or before week 48	1	787	Risk Ratio (M‐H, Random, 95% CI)	1.16 [0.98, 1.37]
2 Undetectable plasma HIV RNA < 50 copies/ml at 48 weeks	1	787	Risk Ratio (M‐H, Random, 95% CI)	0.93 [0.85, 1.03]
3 Death	1	787	Risk Ratio (M‐H, Random, 95% CI)	1.33 [0.50, 3.53]

6.1. Analysis.

Comparison 6 Nevirapine twice daily versus efavirenz, Outcome 1 Treatment failure on or before week 48.

6.2. Analysis.

Comparison 6 Nevirapine twice daily versus efavirenz, Outcome 2 Undetectable plasma HIV RNA < 50 copies/ml at 48 weeks.

6.3. Analysis.

Comparison 6 Nevirapine twice daily versus efavirenz, Outcome 3 Death.

Comparison 7. Nevirapine twice daily versus nevirapine and efavirenz.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Treatment failure on or before week 48	1	596	Risk Ratio (M‐H, Random, 95% CI)	0.82 [0.69, 0.97]
2 Undetectable plasma HIV RNA < 50 copies/ml at 48 weeks	1	596	Risk Ratio (M‐H, Random, 95% CI)	1.04 [0.92, 1.18]
3 Death	1	596	Risk Ratio (M‐H, Random, 95% CI)	2.43 [0.53, 11.14]

7.1. Analysis.

Comparison 7 Nevirapine twice daily versus nevirapine and efavirenz, Outcome 1 Treatment failure on or before week 48.

7.2. Analysis.

Comparison 7 Nevirapine twice daily versus nevirapine and efavirenz, Outcome 2 Undetectable plasma HIV RNA < 50 copies/ml at 48 weeks.

7.3. Analysis.

Comparison 7 Nevirapine twice daily versus nevirapine and efavirenz, Outcome 3 Death.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

French 2002.

Methods	Sequence of generation: Described as 'randomized' but no detail provided on method of randomization. Participants allocated equally so suggests blocked randomization. Allocation concealment: Unclear ‐ patients were randomized prior to entry into trial, but open‐label so may not be adequately concealed. Blinding: Participant: No Provider: No Assessor: No Loss‐to‐follow‐up: Overall ‐ 68% (48/70) completed 52 weeks Analysis not by intention‐to‐treat. Five patients were randomized but did not commence therapy and were excluded.
Participants	Participants recruited from 18 Australian hospital and primary care sites between September 1997 and December 1999. Inclusion criteria: Documented HIV infection; no prior antiretroviral therapy; age >= 18 years; CD4+ count > 50 cells/microL; negative pregnacy test in women; no active or ongoing opportunistic infection; no current chemotherapy, radiotherapy, or immune therapy; no ongoing alcohol or substance abuse. Exclusion criteria: Liver function tests > 5 times normal N = 70 Ninety percent (N = 59) of participants were male with a mean age of 37, SD = 9 years. The mean CD4 count overall was 399 cells/microL, SD = 249 cells/microL. Six participants had prior AIDS.
Interventions	Group 1 (NVP, AZT, 3TC): ‐ Initial dose of Nevirapine (NVP) 200mg daily for 2 weeks increasing to 200mg twice daily ‐ Zidovudine (AZT) 250mg twice daily ‐ Lamivudine (3TC) 150mg twice daily Group 2 (NVP, d4T, 3TC): ‐ Initial dose of Nevirapine (NVP) 200mg daily for 2 weeks increasing to 200mg twice daily ‐ Stavudine (d4T) 40mg twice daily ‐ Lamivudine (3TC) 150mg twice daily Group 3 (NVP, d4T, ddI): ‐ Initial dose of Nevirapine (NVP) 200mg daily for 2 weeks increasing to 200mg twice daily ‐ Stavudine (d4T) 40mg twice daily ‐ Didanosine (ddI) 200mg twice daily All three interventions given for 52 weeks. If Grade 3 or 4 drug‐related adverse events occurred, drug substitution was mandatory within a class. Drug substitution was optional for persistent drug‐related Grade 1 or 2 adverse events or therapeutic failure.
Outcomes	Primary outcomes: ‐ Time weighted mean change from baseline in plasma HIV RNA at week 52 ‐ Proportion of patients with real‐time plasma HIV RNA < 500 copies/mL at week 52 ‐ Proportion of patients with stored plasma HIV RNA < 50 copies/mL at week 52 Secondary outcomes: ‐ Quality of life scores Adverse events: ‐ All events recorded. Clinical and laboratory follow‐up took place every 4 weeks for 52 weeks. Clinical assessment: ‐ physician assessment of adherence to therapy ‐ classification of CDC clinical stage ‐ quality of life questionnaire. Laboratory tests: ‐ complete blood count ‐ electrolytes ‐ liver function tests ‐ urea and creatinine ‐ creatine kinase ‐ amylase CD4+ counts and plasma HIV RNA measurements were done at week 0, 4, 8, and every 8 weeks until week 52. Real time plasma monitoring used a lower detection limit of 500 copies/mL plasma Stored plasma monitoring took place after study completion and used a lower detection limit of 50 copies/mL.
Notes	Ethics approval obtained from each site's Research Ethics Committee. Participants provided written informed consent.
Risk of bias
Bias	Authors' judgement	Support for judgement
Allocation concealment (selection bias)	Unclear risk	B ‐ Unclear

Van Leth 2004.

Methods	Sequence of generation: The sequence was generated by use of the minimisation variables CD4‐positive T‐cell count (<= 350 vs > 350 cells/microL) and study region. Treatment allocation was stratified according to baseline plasma HIV‐1 RNA concentration (<= 30,000 cop/mL vs > 30,000 cop/mL). The initial protocol started with three comparison groups with an estimated sample size of 450 and a treatment allocation of 1:1:1. However, after 5 months a fourth comparison group was added and treatment allocations were recalculated to a 1:2:2:1 ratio. For 68% of the study sample, allocation could have been to one of four groups. Allocation concealment: Adequate: treatment allocation was done at the central study coordination centre, concealed from the investigator before enrolment. Blinding: Participant: No Provider: No Assessor: Unclear Follo Loss‐to‐follow‐up (% not completing week 48 of trial): Overall ‐ 16% (200/1216) Comparison Group A (Nevirapine once daily): 17% (38/220) Comparsion Group B (Efavirenz): 16% (63/400) Comparison Group C (Nevirapine & Efavirenz): 16% (34/209) Comparison Group D (Nevirapine twice daily): 17% (65/387) Analysis by intention‐to‐treat as all patients randomised were included in the efficacy analysis.
Participants	Participants recruited from HIV treatment centres in the following countries: Argentina, Australia, Belgium, Brazil, Canada, France, Germany, Greece, Ireland, Italy, Poland, Portugal, South Africa, Switzerland, Thailand, United Kingdom, and United States of America. The trial began in February 2000 and was completed 48 weeks after the date of last enrolment (June 2001). Inclusion criteria: Chronic HIV infection; no prior antiretroviral therapy; age >= 16 years; plasma concentration of HIV‐1 RNA > 5000 copies/mL. Exclusion criteria: Pregnancy or lactation; Hb < 6.3mmol/L in men or 5.7mmol/L in women; neutrophil count less than 1 x 10(9)/L; platelet count < 75 x 10(9)/L; serum amylase > 2 times upper limit of normal in combination with serum lipase > 1.5 times upper limit of normal; history of clinical pancreatitis or neuropathy within the previous 6 months; renal failure requiring dialysis; radiotherapy, cytotoxic, or immunomodulating treatment within the month preceeding the study or anticipated need for it; infection with HIV‐2; likely non‐adherence as judged by the investigator. N = 1216 Across all four comparison groups, between 61% and 68.4% were male with median age ranging from 33.2 to 34.4 years. The median HIV RNA in log copies/mL was 4.7 (IQR: 4.4‐5.4) in all four groups with the median CD4 count ranging between 170 and 200 cells/microL. No significance levels are reported.
Interventions	All participants regardless of randomized group, received the following drug regimen: Stavudine (d4T) 40mg twice daily (30mg twice daily if weight less than 60kg) Lamivudine (3TC) 150mg twice daily Initial protocol: Comparison Group A: 3TC, d4T and Nevirapine 400mg once daily Comparsion Group B: 3TC, d4T and Efavirenz 600mg once daily Comparison Group C: 3TC, d4T and Efavirenz 800mg once daily & Nevirapine 400mg once daily After 5 months, an additional group assigned to Nevirapine twice daily was added after a study had found that the efficacy of Nevirapine was related to minimum concentration and raised a question around the once daily dosing. Comparison Group D: 3TC, d4T and Nevirapine 200mg twice daily
Outcomes	Primary outcome: Proportion of patients with treatment failure, defined as a composite endpoint with three components: 1. Virology ‐ decline of < 1 log(10) in plasma HIV‐1 RNA concentration within the first 12 weeks or two consecutive measurements >= 50 cop/mL from week 24 onwards 2. Disease progression ‐ CDC Grade C event from 8 weeks onwards diagnosed according to published guidelines, or death 3. Therapy change ‐ non‐allowable change of allocation treatment Secondary outcomes: Proportion of patients with virological failure (never having a plasma HIV‐1 RNA concentration < 50 cop/mL, or two consecutive measurements >= 50 cop/mL after having had a concentration below the cut‐off); Proportion of patients with plasma HIV‐1 RNA concentrations < 50 cop/mL at each study week; Change in CD4‐positive cells between the start of treatment and week 48. Adverse events: ‐ All clinical and laboratory events were recorded. Those clinical adverse events that reflected only asymptomatic laboratory abnormalities were not categorised as clinical‐adverse‐events, but as laboratory abnormalities. Clinical assessment and CD4+ counts and plasma HIV RNA measurements were done at week ‐6 (screening) 0, 2, 4, 8, 12, 24, 36 and week 48. Further monitoring of plasma HIV RNA concentration was done at day 3, week 1 and week 18.
Notes	Study funded by Boehringer‐Ingelheim. Trial approved by ethics review boards in participating countries. All patients gave written informed consent.
Risk of bias
Bias	Authors' judgement	Support for judgement
Allocation concealment (selection bias)	Low risk	A ‐ Adequate

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Garcia 2003	Different intervention: trial of 20 treatment‐experienced participants randomized to receive either Stavudine, Didanosine and Indinavir, or to the same combination plus Hydroxyurea.
Katlama 2000	Insufficient information from published report. In this randomized controlled trial, participants were randomized to add Abacavir or a matched placebo to their stable background antiretroviral therapy (SBG). Twenty‐one percent of the entire sample used two Nucleoside Reverse Transcriptase Inhibitors (NRTIs) plus either a Protease Inhibitor (PI) or a Non‐nucleoside Reverse Transcriptase Inhibitor (NNRTI) as their SBG, but results are not presented according to these sub‐groups (which could contain the combination of Stavudine, Lamivudine and Nevirapine. The authors have been contacted and if possible, the sub‐group data will be included in updated versions of this review.
Lafeuillade 2001	Different intervention: participants randomized to Zidovudine, Lamivudine, Didanosine, Saquinavir, Ritonavir and Interleukin‐2 (group 1) with the same HAART combination without Interleukin‐2 (group 2). This comparison was changed at week 72 to Stavudine, Didanosine, Efavirenz, Hydroxyurea (group 1) with Zidovudine, Lamivudine, and Abacavir (group 2).
Laurent 2004	Not a RCT: this open‐label trial of the combination of Stavudine, Lamivudine and Nevirapine was a follow‐up of a group of patients and was neither randomized nor controlled.
Negredo 2002	Insufficient information from published report. In this randomized controlled trial, participants who had been receiving a Protease Inhibitor (PI)I‐containing regimen were randomized to have the PI replaced with either Nevirapine or Efavirenz in addition to continuing with the same Nucleoside Reverse Transcriptase Inhibitors (NRTIs) they had received before the study. Sixty percent of the participants in the Nevirapine group, 57% in the Efavirenz group and 63% in the PI group received Stavudine and Lamivudine, but results are not presented according to these sub‐groups. The authors have been contacted and if possible, the sub‐group data will be included in updated versions of this review.
Raboud 2002	Different interventions: individual patient data analysis of three randomized trials. INCAS trial compared the combination of zidovudine and nevirapine with the combination of zidovudine and didanosine with the combination of zidovudine, didanosine and nevirapine; AVANTI‐2 compared the combination of zidovudine and lamivudine with the combination of zidovudine, stavudine and indinavir; AVANTI‐3 compared the combination of zidovudine and stavudine with the combination of of zidovudine, stavudine and nelfinavir.
Ruiz 2001	Different intervention: randomized trial comparing treatment interruption with continuous treatment. The comparison groups were not randomized according to a specific HAART combination.
Sabo 2002	Different intervention: pharmacokinetic study as part of a larger randomized, placebo‐controlled trial comparing lamivudine and nevirapine with lamivudine and a matching placebo on a background treatment of zidovudine, didanosine, or zalcitabine.
Yazdanpanah 2004	Systematic review and meta‐analysis using indirect comparisons to compare trials. None of the included trials comprised the Stavudine, Lamivudine and Nevirapine combination.
Yuen 2004	Participants not HIV‐infected: pharmacokinetic randomized cross‐over study of lamivudine conducted in non‐HIV participants.

Sources of support

Internal sources

• South African Cochrane Centre, Medical Research Council, South Africa.

• Cochrane HIV/ AIDS Mentoring Programme, South Africa.

• University of Stellenbosch, South Africa.

• United Kingdom Cochrane Centre, UK.

External sources

• No sources of support supplied

Declarations of interest

None known.

Stable (no update expected for reasons given in 'What's new')