Ritonavir-boosted atazanavir as protease-only maintenance therapy: 48-week results

In a small pilot study, 30 out of 34 people who had undetectable viral loads on protease inhibitor-based triple therapy were still virally suppressed 48 weeks after switching to a maintenance regimen of atazanavir/ritonavir only. No major protease inhibitor resistance mutations were found in any of the people whose viral load rebounded on the simplified treatment. The findings were published in the Journal of Infectious Diseases in March.

In several studies, Kaletra (lopinavir boosted by ritonavir) has proven surprisingly effective thus far as a maintenance regimen – i.e. at keeping HIV viral load suppressed in people who switch to Kaletra monotherapy after first suppressing their viral loads on a 'standard' combination regimen.

Investigators are now examining other maintenance regimens of ritonavir-boosted protease inhibitors (PIs), such as in this prospective, 48-week, single-arm open-label study of atazanavir/ritonavir (Reyataz/Norvir). This trial enrolled 36 adults who had maintained an HIV viral load below 50 copies/ml for at least 48 weeks on their first antiretroviral regimen – in all cases, a protease inhibitor (PI) plus two nucleoside reverse transcriptase inhibitors (NRTIs). (In fact, most had been suppressed for a considerable time – a median of 6.8 years.)

At study entry, all participants (except three, who were already taking boosted atazanavir) switched from their current PI to 300mg atazanavir plus 100mg ritonavir, once daily. Six weeks after this switch, two had dropped out of the study (one due to a viral load increase to 50 copies/ml); the remaining 34 then discontinued their NRTIs.

The main end point was time to virologic failure, defined as two consecutive HIV RNA measurements ≥ 200 copies/ml. Previously reported data had shown that 31 out of 34 patients (91%) maintained viral suppression at 24 weeks after the simplification to atazanavir/ritonavir. (This data was presented at the 13th CROI and subsequently published in JAMA).

Now, after 48 weeks, four participants failed by the stated definition: two at 12 weeks after simplifying, one at 20 weeks and one at 28 weeks. A fifth participant had a detectable plasma viral load of 508 copies/ml at the last study visit, unconfirmed by a second measurement.

Counting only the four repeated failures, the probability of virologic success at week 48 was 88%, with a lower one-sided 90% confidence interval [CI] limit of 81%. Counting the fifth participant as a treatment failure, the success rate was 84% (lower 90% CI, 76%). Finally, using a stricter definition of failure (repeated – although not single – HIV RNA measurements of ≥ 50 copies/ml), the success rate was 82% (lower 90% CI, 73%). (None of these figures include the single participant who withdrew due to viral rebound after switching to atazanavir, but before dropping the NRTIs.)

The study also examined blood plasma levels of atazanavir, residual (low-level) viraemia, and drug resistance mutations. No major PI-associated resistance mutations were found by standard population genotyping in any of the five participants with virologic rebound, although several 'minor' mutations (I64V and G73S) were identified. In eight virally suppressed participants who received single-copy assay (SCA) viral load testing, there were no significant differences in HIV RNA levels throughout the course of the study; median levels were less than 1.1 copies/ml overall at 48 weeks.

Atazanavir blood plasma concentrations were measured monthly, and were linked to likelihood of treatment failure. Three participants had undetectable atazanavir concentrations at least once during the study. Two of these three (67%) went on to virologic failure, which occurred in only in two of the 31 (6%) who had consistently detectable atazanavir levels, "strongly suggest[ing] that suboptimal adherence was an important factor in the development of virologic failure". Median plasma concentrations of atazanavir were also lower, although not statistically significantly, in participants with virologic failure (380 ng/ml vs 660 ng/ml, p=.18).

While noting that this was a non-randomised pilot study of patients with a very stable treatment history, the researchers concluded that it "adds to a growing body of data that simplified maintenance therapy with a boosted PI alone is effective in maintaining virologic control after initial suppression with a 3-drug regimen." Adherence and adequate drug concentrations may be particularly important in the continued success of such simplified treatments.