The majority of virological studies with anti-HIV agents have used subtype B virus, the predominant HIV-1 subtype in developed countries. However other HIV-1 subtypes and recombinant forms are emerging or even predominant in other parts of the world.

Since variations in susceptibility to antiretroviral drugs have been reported among the different HIV-1 subtypes, it is important to monitor a drug's antiviral activity against isolates belonging to subtypes other than B.

Tipranavir (Aptivus) is a non-peptidic HIV protease inhibitor (PI) with potent antiviral activity against a broad range of PI-resistant mutant viruses. Tipranavir maintains a level of antiviral activity against non-subtype B forms that is similar overall to subtype B.

However, detailed analyses have revealed subtle changes in tipranavir susceptibility among the different HIV-1 subtypes, such as evidence of hypersusceptibility of subtype G. Mutation score algorithms have been proposed to predict virological response to tipranavir. These scores are based mainly on subtype B clinical isolates, and it is still unclear whether they can be applied to other non-B subtypes.

To further explore this issue, in the current study researchers evaluated tipranavir susceptibility of a broad panel of prevalent and emerging HIV-1 subtypes and recombinant forms. These results were used to determine the applicability of different tipranavir mutation scores to adequately predict susceptibility of HIV-1 isolates belonging to subtypes other than B.

Tipranavir has previously been shown to have similar antiviral activity against treatment-naive viruses from HIV-1 subtypes B, C, F, G, and CRF02_AG (a recombinant subtype). To confirm and extend these data, the in vitro antiviral activity (50% effective concentration, or EC50) of tipranavir was determined against a broad panel of wild-type HIV-1 Group M clinical isolates belonging to prevalent and emerging subtypes and circulating recombinant forms.

The antiviral activity of tipranavir and other protease inhibitors was measured using the PhenoSense assay for a panel of 57 HIV-1 Group M wild-type clinical isolates from antiretroviral treatment-naive patients (subtypes A, B, C, D, F, G, and H as well as recombinant forms CRF01_AE, CRF02_AG, and CRF12_BF). EC50 values were compared to a drug-susceptible reference subtype B HIV-1 NL4-3 strain.

Results

• Tipranavir had the following mean EC50 values (fold-change range):

- 180 nM (0.6-4.3) for subtype A;
- 99 nM (0.8-1.7) for B;
- 124 nM (0.8-2.1) for C;
- 121 nM (0.6-1.8) for D;
- 195 nM (1.2-2.9) for F;
- 103 nM (0.5-1.6) for G;
- 24 nM (0.9-2.0) for H;
- 109 nM (0.6-2.1) for CRF01_AE;
- 116 nM (0.9-1.8) for CRF02_AG;
- 133 nM (0.5-2.8) for CRF12_BF.

• All subtypes had a mean fold-change at or below the PhenoSense lower clinical cutoff value of 2.0.

Discussion

• Tipranavir maintains a level of activity against wild-type HIV-1 non-subtype B isolates that is similar overall to that observed with subtype B isolates (mean EC50 of 134 nM across all 51 non-B isolates vs 99 nM for all 6 B isolates).

• All non-subtype B isolates have a mean phenotypic fold-change lower than 2 for tipranavir, i.e., below the PhenoSense lower clinical cutoff value for tipranavir (determined mainly from subtype B viruses). Only subtype F isolates have a mean fold-change of 2.0, i.e., exactly at the cutoff value.

• As reported by Abecasis et al, some subtype F isolates show evidence of lower susceptibility to tipranavir. However, unlike that group, the authors of the current study could not confirm any clear trends toward hypersusceptibility for subtype G or CRF02 AG isolates.

• Tipranavir mutation scores have been developed mainly using data from HIV-1 subtype B. This study shows that these do not correlate well when applied to wild-type non-subtype B isolates. For example:

- All scoring methods (except Calvez score) predict subtype F to be well within the susceptibility area defined by each algorithm, contrasting with the evidence of lower susceptibility to tipranavir among some subtype F isolates.

- Conversely, the TUW, TW, and Calvez scoring methods predict CRF02_AG to be at or above the susceptibility threshold defined by each algorithm, yet all individual CRF02_AG isolates in this study are below the Phenosense lower clinical cut-off for tipranavir.

Based on these findings, the study investigators concluded:

All HIV-1 non-B subtypes included in this study had a mean fold-change at or below the tipranavir lower clinical cutoff value of 2 as determined by the PhenoSense assay.

Tipranavir mutation scores have been developed mainly using data from subtype B isolates. This study shows that these do not correlate well when applied to wild-type non-subtype B isolates, as evidenced by the lack of any discernable trend between mean phenotypic fold change and mean tipranavir mutation scores across the subtype/CRF panel.

By extension, it is unlikely that tipranavir mutation scores will accurately predict the susceptibility of tipranavir to non-subtype B isolates from PI-experienced patients having developed PI-resistant mutations.

There is a need to develop better susceptibility scoring algorithms targeted more specifically toward prevalent non-subtype B HIV-1 strains.

For more details on this study, see the full poster with tables and figures presented at the recent 11th European AIDS Conference (EACS) in Madrid.

Boehringer Ingelheim (Canada) Ltd/Research & Development, Laval, Québec,
Canada; Monogram Biosciences, South San Francisco, CA.

11/09/07

Reference
PR Bonneau, Y Lie, and R Bethell. Tipranavir displays similar in vitro antiviral activity against non-subtype B clinical isolates of HIV-1 compared with isolates from subtype B. 11th European AIDS Conference (EACS). Madrid, Spain. October 24-27, 2007. Abstract (poster) P3.3/02.