Tenofovir/FTC more effective than abacavir/3TC in first-line boosted-PI regimens: meta-analysis

A systematic review of controlled clinical trials has suggested that the dual-nucleoside backbone of tenofovir/FTC results in higher efficacy than abacavir/3TC. The results, drawn from 4896 patients in 21 treatment arms of 12 clinical trials, were presented in a poster by Dr Andrew Hill of Liverpool University at the 48^th Interscience Conference on Antimicrobial Agents and Chemotherapy in Washington DC.

A combination of two nucleoside reverse transcriptase inhibitors (NRTIs) forms the "backbone" of most current antiretroviral combinations. Two of the most widely used dual-nucleoside backbones are tenofovir/emtricitabine (TDF/FTC, Truvada) and abacavir/lamivudine (ABC/3TC, Kivexa).

These two nucleoside backbones have been evaluated, in combination with boosted protease inhibitors (PIs), in numerous clinical trials. Results have tended to favour Truvada but have not been entirely consistent: while Truvada suppressed viral load more successfully than Kivexa in ACTG 5202, the HEAT trial did not find any significant difference, and Truvada's superiority in the BICOMBO trial seemed largely driven by drop-outs due to abacavir hypersensitivity, which can now be avoided by screening.

When the ACTG 5202 and HEAT findings were presented back-to-back at the recent World AIDS Conference in Mexico City, there was some speculation as to what a more complete meta-analysis might show.

Now, in this review, investigators have performed such an analysis – searching MEDLINE to identify clinical trials which used a boosted PI with a Truvada and/or Kivexa backbone. They identified a total of 21 treatment arms in 12 clinical trials, including HEAT, KLEAN, ARTEMIS, CASTLE, and eight others. ACTG5202 was not included in this meta-analysis, because data were only available for participants with baseline viral loads over 100,000 copies/mL.

The 12 trials included a total of 4896 antiretroviral-naïve patients, of whom 3340 received Truvada and 1556 received Kivexa. The most commonly-used study protease inhibitor was lopinavir/ritonavir (Kaletra); ritonavir-boosted darunavir, atazanavir, fosamprenavir and (in one study) saquinavir were also used.

Data from these studies were then pooled and analysed for one primary outcome: the percentage of people, on either nucleoside pair, who reached viral suppression (HIV RNA

There were no significant differences between the pooled Truvada and Kivexa groups in median baseline CD4 cell count (204 and 195 cells/mm³, respectively) and mean HIV RNA level (4.9 log₁₀ and 5.0 log₁₀ respectively).

When data were grouped by protease inhibitor, comparable or higher rates of HIV suppression (Truvada than with Kivexa.

• For lopinavir/r (Kaletra), Truvada was clearly superior, with 74% suppression (95% confidence interval [CI], 72 to 76%, n = 2285) compared to 66% (95% CI, 63 to 70%) with Kivexa (n = 722).
• For atazanavir/r, the two were comparable, with 79% suppression (95% CI, 75 to 82%) with Truvada (n = 493) and 77% (95% CI, 69 to 85%) with Kivexa (n = 112).
• Saquinavir/r and darunavir/r did not have Kivexa arms. Truvada performed well paired with darunavir/r (84% suppression, 95% CI, 80 to 88%, n = 343) but poorly with the weaker saquinavir/r (65% suppression, 95% CI, 58 to 72%, n = 166).
• There were no protease inhibitors alongside which Kivexa outperformed Truvada in terms of viral suppression rates.

Across all trials, suppression rates were significantly higher for people with baseline viral load below 100,000 copies/ml versus those above (75.1% vs 68.7%, p = .0019). For the three PIs which were paired with both NRTI backbones (lopinavir/r, atazanavir/r and fosamprenavir/r), Truvada significantly outperformed Kivexa in people with baseline viral load below 100,000 copies/ml (79.1% vs 69.5%, p = .0001) and tended toward superiority, but not significantly so, for viral loads above 100,000 copies/ml (70.6% vs 65.9%, p = .0995).

Two key issues should be considered in interpreting this analysis. Firstly, as the investigators note, this meta-analysis "does not have the strength of randomised comparative trials". Nevertheless, the analysis did show better, or at worst comparable, virological outcomes with Truvada versus Kivexa across a range of different PIs and viral load levels.

The other point is that "failures" here were not purely virological failures, but (due to the intent-to-treat analysis) included all people who did not have documented viral suppression at 48 weeks for whatever reason. This may mean that some of the apparent advantage of Truvada is due to tolerability rather than potency, as more people may have remained on the drug. This could still reflect a real-world advantage, but the distinction should be explored further, especially as abacavir hypersensitivity screening is now likely to rule out many people who would have had to discontinue abacavir in the past.

While noting these caveats, the researchers concluded that this systematic review of first-line clinical trials of dual nucleosides plus boosted PIs "suggests higher efficacy for first-line use of a TDF/FTC NRTI backbone, relative to use of ABC/3TC."