XIV International HIV Drug Resistance Workshop: Entry inhibitors

Entry inhibitors (EI) are a class of antiretrovirals that is receiving a great deal of research attention with a number of companies poised to make an impact in the field. However, despite this enthusiasm the emerging data are founded on in vitro analysis. So whilst the understanding of complex envelope (env) kinetics and co-receptor switching is of scientific interest, EIs remain a tantalising distant option for prescribing physicians.

A number of studies at the workshop examined the impact of mutations in the env glycoprotein which mediates binding of the virus to host cell receptors. These mutations can impair efficiency of viral entry, impact on replication capacity and confer resistance to entry inhibitors. Eric Arts demonstrated some elegant techniques to highlight specific variations in the V3 region of env. His group at Case Western University in Ohio identified sequences that showed up to 100-fold variations in sensitivity to various compounds including enfurvitide (Fuzeon), TAK-779, and PSC-RANTES. In particular, the researchers observed that changes at positions 318 and 319 in the V3 loop which can occur as natural polymorphisms were associated with variable sensitivity to EIs. These mutations were cultivated in an R5-NSI HIV-1 subtype A isolate, a strain that has been previously noted for increased sensitivity to EIs. These were then introduced back into a standard reference sample by a yeast recombination cloning method.

They assessed sensitivity to the compounds, fitness and what they refer to as ‘receptor avidity’ – or propensity to bind by use of cell fusion assays. Changes at 318 or 319 involving arginine replacement and certain double mutations resulted in reduced fitness and co-receptor binding and 10-fold enhanced sensitivity to the EIs compared to WT.

Studying this and other env region variable regions (V1, V2 and V3 loop), Donald Mosier from The Scripps Research Institute, La Jolla, highlighted how these different regions can impact the efficiency of co-receptor usage and are critical for R5 to CCRX4 evolution. The critical component of their study involves the examination of the transition between infection of R5 and X4 expressing target cells. The intermediates generally incurred V3 mutations with significant loss of entry fitness; these decreases in fitness were then compensated by changes in V1 and V2. In the laboratory strains examined, the viable mutational pathways between R5 and X4 were found to be rare, constituting

Glaxo SmithKline’s CCR5 antagonist, GW873140

The GSK CCR5 antagonist (873140) has been previously shown in a 10 day monotherapy trial to lead to a viral load decrease of 1.66 logs at nadir. In the study presented here, researchers performed clonal analyses on samples from a patient where a co-receptor switch had been observed during treatment with 873140, noting the impact of pre-existing dual-tropic viruses on subsequent changes in viral tropism. A population-based tropism assay had shown only R5 tropism at Day 0, the emergence of dual/mixed tropic viruses at the end of the 10 days of 873140 dosing, and a return to R5 tropic viruses by Day 24. Moreover, the Day 24 phenotype had the same drug sensitivity profile as observed at Day 1.Clones from this subject revealed that R5 tropic virus was dominant at Day 1 (96%) and Day 24 (92%) but that viruses with dual/mixed tropism made up 27% of the virus population at Day 10. The authors explain that the original coreceptor switch findings may be attributable to pre-existing dual-tropic virus that was present at low level and simply not detected by the population-based assay at Day 1. All of the functional clones remained sensitive to the compound in the PhenoSense entry assay with dual-tropic strains being slightly more sensitive. These confounding results do not answer the key questions of tropism switch and its impact on the host They do suggest that the assay has limitations in detecting minor virus populations. It may be that the focus on actual switching from R5 is X4 is ill-placed and that dual-tropic viruses exist in the host most of the time, but simply below the levels of quantification. Many issues surrounding virus tropism and interaction with the host remain to be addressed.

Schering Plough’s CCR5 antagonists

Schering-Plough presented in vitro selection resistance data relating to their two CCR5 antagonists in development: Schering C (SCH 351125) and Schering D (SCH 417690). After passage with the compounds, resistant viruses were assessed for co-receptor tropism and cross resistance to other CCR5 inhibitors and by sequence analysis. All isolates remained CCR5 tropic on primary cells but one of the isolates gained the ability to infect X4 expressing cells in a recombinant assay. Mutations in the V3 loop were detected that lead to a reduction of the CCR5 inhibitors activity. Interestingly, one clone showed resistance but not located within V3 suggesting that mutations in regions outside of V3 can also be responsible for resistance. Importantly, the authors found that the resistant viruses were also resistant to other CCR5 binding inhibitors but were sensitive to other antiretrovirals including the fusion inhibitor enfuvirtide.

These studies demonstrate the complexity and current limitation in understanding host genetics. Although these data are promising in that they confirm slow emergence of resistance and perhaps differentials in genetic barrier to achieving resistance, nevertheless they confirm that cross-resistance across the same class of antagonists is possible. As John Mellors from Pirttsburgh University noted however, the limitations of current diagnostic techniques including the difficulty of assessing changes in X4 which can only be readily determined in vivo mean that we cannot extrapolate in vitro knowledge into clinical relevance.

Maraviroc

The most advanced CCR5 inhibitor in terms of clinical development is the Pfizer agent Maraviroc (MVC, UK-427,857). Based on tests using replication-competent and recombinant virus assays assays they deduced that virus resistant to MVC would sustain sensitivity to other CCR5 antagonists including SCH-C, SCH-D and the GSK873140. This is a curious finding since all R5 bind to the same trans-membrane region. They explain this phenomenon through modeling studies based on published data that by demonstrating that although the target regions are very similar, the inhibitors hold the receptor in conformations with sufficient diversity between the compounds to allow inhibition of entry into host cells, even of MVC resistant virus. In vitro at least, the data show that cross-resistance between these compounds is unlikely and that sequencing of R5 antagonists may be possible.

Cross resistance and synergy between entry inhibitors and CCR5 antagonists

Further data from GSK suggested that their R5 antagonist (873140), currently in Phase 2/3 trials retains efficacy against enfuvirtide-resistant virus and that there may be additional synergy between these compounds. Using a range of R5-tropic strains sensitive and resistant to enfuvirtide, they assessed the inhibition properties of 873140. The research team found that both enfuvirtide resistant and sensitive strains remained highly susceptible to 873140 with a median IC50 of 0.28nM and 0.26nM for sensitive and resistant isolates respectively. In combination 873140 and enfurvitide showed a small but statistically significant dose-related additive effect suggesting synergistic interaction between these drugs.

Although cross-resistance across the attachment and fusion inhibitors is intuitively not expected to emerge because of their different mechanisms of action, nevertheless these data rule out in vitro antagonism or cross-resistance. More interesting is the data that elucidates the interaction between these two types of agents, implicating the efficiency of one in improving the action of the other. For example, synergy could be explained by the fact that R5 inhibitors could slow the fusion process potentially enabling enfuvirtide a greater window and activity.

Charles Boucher from the University of Utrecht, Netherlands, cautioned against premature excitement of any absence of cross-resistance between R5 inhibitors noting that the ultimate proof of their efficacy will come from clinical practice. Michael Greenberg from Trimeris, an expert on fusion further noted that resistance per se may not be the issue but rather that the particular sequencing of these drugs could prove to be important. Whilst slightly discrete resistance pathways may emerge for different drugs he added, the order in which they appear may be paramount. For example, viruses with resistance to enfurvitide retain susceptible to T1249. However, their own experience had shown that T1249 resistant strains could not be similarly salvaged by enfuvirtide.

Finally, further longitudinal data on enfuvirtide were presented by Trimeris and Roche from their Phase 3 TORO studies. This smaller study followed a group of 25 patients who had continued to use enfuvirtide despite virological failure at one year. Genotypic data for gp41 and phenotypic data using entry assays were monitored for these patients. They observed changes in gp41between positions 36-45 in 76% of patients with the mean number of substitutions increasing from 1.4 to 1.8. Mutations were reported in both the HR1 and HR2 regions but interestingly, also in the fusion peptide region at positions 3, 7 and 14. These results show that continued enfuvirtide therapy in the presence of virologic failure results in decreased effectiveness of enfuvirtide. The authors suggest that the continued virus evolution they observed implies residual virological activity of enfuvirtide after virological failure. These data somewhat confound presentations from last year’s workshop that showed patients who continue with a failing enfuvirtide regimen can still achieve some virologic benefit as a result of impaired replication capacity following treatment interruption.

However it seems that the residual antiviral activity and/or fitness defects conferred by enfuvirtide resistance mutations may lead to virologic and/or immunologic benefits in some patients despite the further viral evolution reported here. Additional studies will be needed to clarify the clinical benefits of continued enfuvirtide therapy after viral rebound.