Atazanavir begins to lose its clout as protease mutations rise

Response to the new protease inhibitor atazanavir is blunted in individuals with four or more protease inhibitor-associated resistance mutations, even when drug levels are boosted with ritonavir, according to an analysis of the 045 study presented yesterday at the Ninth European AIDS Conference in Warsaw. However, lopinavir-treated patients in the same study showed a similar fall off in response, suggesting that most of the current crop of protease inhibitors begin to lose their clout in roughly the same territory.

Atazanavir is a new once daily protease inhibitor manufactured by Bristol Myers Squib. It was licensed in the United States in June, and is expected to receive a European approval in the first quarter of 2004. The drug is already available on a named patient basis in the UK for patients whose doctors believe it will be a useful component of a new regimen. However, doubts remain about the best way to use atazanavir, particularly where a patient cannot tolerate ritonavir and needs to take atazanavir alone.

The 045 study compared the response to atazanavir/ritonavir, lopinavir/ritonavir or saquinavir/atazanavir in patients who had experienced failure of at least two prior regimens (including one protease inhibitor-containing regimen). Results from the study presented at the Paris International AIDS Society conference in July showed that atazanavir/ritonavir had an equivalent antiviral effect to lopinavir/ritonavir in this group of patients after 24 weeks, but did not analyse the impact of baseline protease inhibitor resistance on response to atazanavir.

The analysis presented today looked at data from the 045 study and also at data from the 043 study, which compared atazanavir alone with lopinavir/ritonavir. In both studies patients also took two nucleoside analogues selected by resistance testing; in the 045 study one of those drugs could be tenofovir, which is known to reduce atazanavir levels.

At baseline, 74% of patients receiving atazanavir in the 045 study were judged to be fully susceptible to the drug by resistance test, compared with 72% of atazanavir recipients in the 043 study, despite the fact that the latter group were less PI-experienced (having failed no more than one prior PI). Total duration of PI exposure was identical in the two populations, but in the 045 study, patients had an average of two years more nucleoside analogue experience than in the 043 study.

In the 043 study, patients judged atazanavir-susceptible at baseline had a median viral load reduction of –1.82 log, compared to a –1.4 log reduction in the atazanavir resistant group. In the 045 study however, the difference in response was more pronounced. Atazanavir-susceptible patients had a –2.16 log reduction, whilst atazanavir-resistant patients had a –0.94 log reduction, suggesting that the degree of resistance encountered in the 043 study was lower.

The break point for a substantial response appeared to be the accumulation of four protease mutations in both studies, although the numbers with four or more mutations in the 043 study was too small (26%) to carry out a meaningful analysis. In the 045 study, patients with less than four mutations had an average viral load reduction of –2.23 log. Those with four or more mutations had an average viral load reduction of –1.00 log.

Lopinavir-treated patients also showed a poorer response once they had accumulated at least four mutations, but the data presented did not allow observers to judge whether there might be a trend towards better response at higher resistance levels in lopinavir-treated patients. When analysed according to the relationship between the number of mutations and the proportion of patients with viral load below 400 copies, the data showed a clear trend towards a higher proportion of lopinavir-treated patients with three or more mutations achieving viral load below 400 copies in the 045 study. However, this analysis did not control for the number of nucleoside analogue and NNRTI-related mutations. Patients with three protease mutations but high level NRTI resistance may have been over-represented in the atazanavir group, but that could not be determined from this analysis, and clearly more work will be needed to assess the contribution of specific mutations to a poor atazanavir response.

The findings do however reinforce the view that atazanavir will require ritonavir boosting in any patient with more than two primary protease mutations, and beg the question of when we can expect to see data that evaluate the inhibitory quotient of ritonavir-boosted atazanavir - the relationship between drug levels and the ability of the drug to overcome drug resistant virus. Such data might help study designers and clinicians think through the possibilities of using atazanavir as a double boosted PI (combining it with a small dose of ritonavir and another protease inhibitor which also has a boosting effect on atazanavir).