15th HIV Resistance Workshop: resistance to integrase inhibitors

Integrase inhibitors, a new class of antiretrovirals currently in Phase II and III development target the process of integrase strand transfer. Whilst a putative resistance pathway has been previously described, the genotypic profile remains to be fully characterised. Two studies served to further elucidate the susceptibility of integrase inhibitors (INI) in particular highlighting the potential contribution of polymorphic variants in the integrase gene.

Max Lataillade from Yale reported on the prevalence and patterns of polymorphisms and the frequency of naturally-occurring amino-acid substitutions correlated with clinical resistance to INI. Using real-time PCR, the complete integrase gene was analysed covering 867 bases for 237 patients who were all INI-naive (61 patients from Yale and 176 sequences obtained from the Los Alamos database). The purpose of the study was to determine natural polymorphisms in integrase enzyme in INI-naïve patients to ascertain integrase enzyme consensus.

Analysis from the combined sequences revealed that 62% of the amino-acid positions were polymorphic. Two areas on the integrase gene appear to be well conserved: catalytic triads known as the DDE and HHCC motifs. Of the mutations associated with INI resistance, 12 of the 23 substitutions were natural polymorphisms occurring at positions V72I, A128T, E138K, V151I, S153Y/A, M154I, N155H, V165I, V201I, T206S, S230N. Of these, four were found to occur with greater frequency: V72I, V165I, V201I and T206S.

A different sequence of mutations was reported to be associated with high level INI resistance and not occurring as polymorphic variants, namely: T66I, L74M, F121Y, T125K, G140S, S230R, V249I and C280Y. The authors conclude that INI mutations associated with INI resistance appear frequently as natural polymorphisms. Furthermore, the research found that mutations conferring high-level resistance to INI appear later and with less frequency explaining the success of virologic response demonstrated in INI clinical trials. Studies such as this that characterise naturally-occurring variants will enable future analysis of new signature mutations to be identified that drive resistance to INI.

Whilst the Yale study focused on clade B virus, a Monogram Biosciences study profiled the impact on a range of B and non-B clades of the Merck integrase strand transfer inhibitor L-870,810. An adapted version of the Monogram PhenoSense assay tested 130 INI naïve samples with vectors containing 3’-end of RT, amino acids 317-440, RNaseH and integrase. Of these samples 100 were clade B and the remaining represented subtypes A, C, D, CRF01-AE, G. Interestingly, in five out of 12 clade C samples, four additional amino acid changes were observed at the C terminus of integrase but this did not impact on the susceptibility of the compound tested. Two positions were of note in the analysis:

• T97A = ~2.5 fold reduced susceptibility (present in 2.3% of samples)
• K156N = 2 fold increased susceptibility (found in 11.5% of samples)

This particular analysis showed that naturally occurring polymorphisms such as T97A and K156N can affect virologic response. Mike Westby from Trimeris in a review of the studies asked given our limited experience with INI, what range of fold-change may be significant for INI and whether the potential impact of polymorphic variations in integrase should be reconsidered. John Mellors from Pittsburgh countered that perhaps the polymorphic profile may not be remarkable compared for example, to the diversity found in the reverse transcriptase (RT) gene. A closer examination of escape mutants could help clarify the frequency and patterns of polymorphisms. Dr Mellors also postulated that polymorphisms may only become significant at a specific frequency which as yet has not been defined for INI. Neil Parkin from Monogram supported this, noting that the focus should clearly be on determining cut-off values for INI that reflect a range distinguishing biologic from clinical cut-offs.