This review found that virological suppression was achieved with first-line regimes of antiretroviral therapy in patients with HIV in sub-Saharan Africa, but that reduced susceptibility due to drug resistance could mean a second-line treatment was reasonable. Large losses to follow-up, potential biases in the included studies, and some methodological flaws mean that the reliability of the authors' conclusions is unclear.

To evaluate virological efficacy and outcomes related to drug resistance of antiretroviral therapy programmes in sub-Saharan Africa in adults infected with HIV.

PubMed and EMBASE were searched to June 2009 for relevant studies; search terms were reported. Abstracts from the 2007 and 2008 International AIDS Society conferences and abstracts from the Conferences on Retroviruses and Opportunistic Infections from 2007 to 2009 were handsearched. References lists from retrieved articles were screened for additional references.

Original papers and abstracts that evaluated the first-line treatment with combinations of three different antiretroviral that provided highly-active antiretroviral therapy in adult patients infected with HIV-1 in sub-Saharan Africa were eligible for inclusion. Eligible studies were controlled trials, cohort studies, cross sectional studies or case series. Additional inclusion criteria were that HIV-1 RNA had to be measured at least once after commencement of highly-active antiretroviral therapy, and follow-up of patients was required to be at least three months after the beginning of treatment. For the analyses of acquired drug resistance, reports were included that evaluated virological failure of patients on first-line retroviral therapy. Studies with key data missing for location, sample size treatment status and HIV RNA at follow-up, studies of patients younger than 15 years, studies of patients infected with HIV-2, and studies of antiretroviral monotherapy, dual therapy or second-line highly-active antiretroviral therapy were excluded.

The included studies took place in 18 sub-Saharan countries, including nine multi-site studies. Most of the included patients had advanced disease at commencement of highly-active antiretroviral therapy, with median baseline CD4 cell counts generally less than 200 cells per μL, with a range of 43 to 387 cells per μL. The highly-active antiretroviral therapy regimes were specified in 92% of the patients, 91% of whom received treatment comprising two nucleoside reverse transcriptase inhibitors and either nevirapine or efavirenz as non-nucleoside reverse transcriptase inhibitor; the remaining patients had regimes based on protease inhibitors, dual therapy, or triple nucleoside reverse transcriptase inhibitors.

Two authors performed the study selection.

The authors stated they did not assess methodological quality. However they did summarise results by the use of intention-to-treat analyses performed in the included studies.

Two reviewers independently extracted data on study characteristics and outcomes (virological success and failure). Virological suppression was defined as a viral load of fewer than 50 copies per mL and virological success as an HIV-1 RNA of fewer than 400 copies per mL at time of measurement. The definition of virological failure was a viral load of more than 1,000 copies per mL in at least one test and at least three months after the commencement of highly-active antiretroviral therapy.

Due to the clinical heterogeneity of the included studies, the results were summarised in a narrative review. The overall proportions, medians and ranges of patients with virological success and failure at six, 12, and 24 months were summarised.

Subgroup analyses were performed of studies where more than 90% of patients received non-nucleoside reverse transcriptase inhibitors.

A total of 89 studies (n=63,684 patients) in 18 countries were included in the review, including 69 full-text articles and 20 conference abstracts. Sample sizes ranged from 16 to 9060 patients; fewer than 200 patients were included in 33 studies. Length of follow-up ranged between three months and 60 months.

Virological success: For the studies using on-treatment analyses, virological success was attained by 78% of patients (median 82%, range 54 to 97; n=13,288 patients) at six months, 76% of patients (median 82%, range 51 to 97; n=9,794 patients) at 12 months, and 67% of patients (median 72%, range 46 to 90; n=5,690 patients) at 24 months follow-up. For the studies using intention-to-treat analyses, the following proportions of patients achieved virological success: 79% (median 79%, range 71 to 90; n=2,856 patients) at six months, 69% (median 74%, range 51 to 74; n=3,236 patients) at 12 months, and 63% (median 67%, range 52 to 82; n=1,332 patients) at 24 months follow-up.

Virological failure: Virological failure was experienced by a median proportion of 14% of patients (range 0 to 43; n=19 studies). The highest proportion of failure was observed in a cohort 9% of patients who received dual or monotherapy instead of highly-active antiretroviral therapy.

Subtype analysis and drug-resistance development: Subtypes of HIV-1 were reported 849 patients (22 studies) who experienced treatment failure; the most prevalent was subtype C. Analyses of drug resistance profiles showed that they were associated with the antiretroviral medications commonly used. The most common drug-resistance mutations were the lamivudine-associated M184V mutation conferring resistance to lamivudine and emtricitabine, followed by K103N associated with the use of non-nucleoside reverse transcriptase inhibitors.

First-line antiretroviral therapy regimes used in sub-Saharan Africa were effective, but drug resistance profiles in the included studies showed reduced susceptibility to lamivudine and non-nucleoside reverse
transcriptase inhibitors. This meant that a second-line treatment regime based on protease inhibitors, with a backbone of nucleoside reverse transcriptase inhibitors, was a reasonable option for patients who experienced first-line treatment failure.

The review addressed a defined question. Criteria for the inclusion of studies were clearly stipulated, but the reviewers included data from patients who received monotherapy and dual therapy, although this was specified as an exclusion criterion. It was unclear if there were any language restrictions in the literature search. Some attempts made to identify unpublished material were made. Steps were taken to minimise reviewer error and bias for study selection and the extraction of data.

There was no formal assessment of methodological quality of the included studies, although a distinction was made between data collected from studies that used intention-to-treat analyses and studies that used on-treatment analyses. The results from uncontrolled studies were associated with particular biases and confounding, so the combination of such results to calculate proportions may not have been appropriate. The authors acknowledged some of the limitations of the review including heterogeneity of study design, definitions of virological success and failure outcomes, and the paucity of studies in which intention-to-treat analyses were used.

Some methodological flaws mean that the reliability of the authors' conclusions is unclear.

Practice: The authors stated first-line therapy regimens used in sub-Saharan Africa are effective, but that profiles of drug resistance mean that a second-line treatment regimen based on protease inhibitors, with a backbone of nucleoside reverse transcriptase inhibitors, is a reasonable option for this group of patients who experience first-line treatment failure. They also stated that efforts should be made for earlier detection and and treatment of patients with HIV infection in sub-Saharan Africa, given that African patients were at a more advanced stage of infection at the beginning of treatment, resulting in decreased response to medication. In the presence of thymidine analogue drug resistance mutations, lamivudine and zidovudine containing backbones cannot be recommended.

Research: The authors did not state any particular implications for research, but remarked that that long-term follow-up data remain scarce, and in most cases are from on-treatment analyses. Caution is required when comparing the results with those from developed countries, as African cohorts have high early mortality and large losses to follow-up after the commencement of antiretroviral therapy.

Not stated

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.