At IAS 2013, investigators from the EARNEST trial reported the primary outcome of their study at 96 weeks. Study details are as follows: 1,277 HIV-infected Africans with failure of their first NNRTI-based regimen were randomly assigned to one of three regimens:

  • lopinavir/ritonavir (Kaletra) + two or three NRTIs; 79% received tenofovir (Viread) + lamivudine (3TC, Epivir) or emtricitabine (FTC, Emtriva)
  • lopinavir/ritonavir + raltegravir (Isentress)
  • lopinavir/ritonavir monotherapy with raltegravir induction (the raltegravir was phased out after three months)

The volunteers' median CD4+ cell count was really low (71 cells/mm3), and 42% of the participants had a plasma viral load above 100,000 copies/mL.

The primary endpoint, so-called "good HIV disease control," included being alive and without new WHO4 events at 96 weeks, having a CD4+ cell count greater than 250 cells/mm3, and having a plasma viral load below 10,000 copies/mL (or greater than 10,000 copies/mL provided there were no protease inhibitor resistance mutations) -- criteria that would be considered unacceptable in any developed country.

Despite this lenient definition, the rates of "good HIV disease control" were 60% for the NRTI arm, 64% for the raltegravir arm, and 56% for the monotherapy arm, which was not enough to demonstrate non-inferiority for monotherapy against the NRTI arm (treatment difference -4.1%; -10.8, +2.6). The rates of viral load suppression to below 50 copies/mL at 96 weeks were 74% for the NRTI arm, 73% for the raltegravir arm, and a complete disaster for the monotherapy arm at 44%. Moreover, 18% of those on monotherapy had intermediate- or high-level resistance to lopinavir/ritonavir, and that was assuming that all patients with a viral load less than 1,000 copies/mL were still susceptible to the drug.

The rates of grade 3/4 adverse events were similar among the three arms: 22%, 23% and 24%, respectively.

One wonders about the rationale of even including a boosted protease inhibitor monotherapy arm in this rescue study, after the approach noisily failed in at least two previous studies (SARA and HIV STAR), and wonders as well about the design of only 12 weeks of intensification with raltegravir, when it is universally accepted that at least six months of undetectable viral load are required to avoid failures on boosted protease inhibitor monotherapy.

The authors emphasized this was a pragmatic trial that replicates typical public approach settings. Therefore, baseline resistance tests were not done in all individuals (and were not reported in the IAS 2013 presentation).

In addition to its similarities to the SARA and HIV STAR studies, this is the second study (after the SECOND LINE study presented at CROI 2013) that has been unable to demonstrate superiority of lopinavir/ritonavir + raltegravir to lopinavir/ritonavir + two recycled NRTIs (including inactive lamivudine and a second, jeopardized NRTI). This is not a great result for raltegravir, even though it was one of the hypotheses going into this study. It stresses, however, the substantial activity of recycled (often inactive) NRTIs.

Which other studies presented at IAS 2013 will have lasting impact long after memories of the conference itself have faded? Read more of Dr. Llibre and Dr. Young's top picks.