First European HIV Resistance Workshop highlights non-B subtypes, replicative capacity

Thanks to Dr Deenan Pillay, University College London, for comment and review.

The First European HIV Resistance Workshop held in Luxembourg earlier this month proved to be a meeting of both interest and utility to the HIV treatment community in Europe. The aim of establishing a focused forum committed to profiling scientific development and bringing together researchers, physicians and patients advocates from across Europe is laudable and practically needed. Most notably, differences between Europe and North America in the epidemiology of HIV, population profiles and viral diversity are not only more apparent, but benefit from an opportunity to characterise public health and patient care within the European community. The first workshop focused attention on the molecular mechanisms of resistance evolution and its clinical implications. The emphasis on diagnostic technology was concentrated on interpretation systems rather than monitoring assays, whilst the evolving profile of viral subtypes was extremely well described through a range of studies from Europe and the African sub-continent.

Of particular interest, in the context of treatment experienced patients, there were a number of presentations dedicated to sequencing strategies including the perennial modelling of alternating cycles of predefined therapy. One such structured switching approach entitled ‘continuously alternating therapy’ (CAT) involved two therapies, one targeting the reverse-transcriptase (RT) region and the other, the viral protease (PR) and required a switch at 6 week intervals. Therapy was selected on the basis of genotyping. The aims of the strategy are two-fold; to influence the shift in viral population in patients with heavily resistant virus and to mitigate toxicity associated with mega-HAART regimens used in salvage. The outcome although not dramatic, nevertheless demonstrated a benefit with a mean decrease in viral load (VL) of 0.85 log copies and an increase in CD4 counts of 50 copies observed at 20 weeks. The results are based on the analysis of 8 patients so far recruited to the study.⁽¹⁾

Q151M mutation reversion

Given that many HIV patients are now heavily treatment-experienced, there was much interest in the possibility of reversion of the multi-drug resistant (MDR) mutation Q151M associated with high level resistance to AZT, ddI, ddC, d4T and ABC and lower level resistance to 3TC and tenofovir (TDF). A prospective study from Italy involving a large clinical database of patients undergoing genotypic resistance testing (GRT) found that 3.6% of the 470 treated experienced patients with both PI and NNRTI experience (83% and 51% respectively) carried the Q151M mutation.

In this study, Q151M was not associated with the following variables: age, sex, AIDS diagnosis, time on therapy, number of PI mutations, CD4 status and VL at time of GRT or exposure to NRTIs (other than 3TC) and NNRTIs before GRT was performed. Whilst Q151M has been previously observed with a background of multiple NRTI mutations, this appears to be the first study to show a significant correlation with sexual transmission of HIV. After a median follow-up of GRT at 6 months, the authors noted a full reversion of the Q151M complex to wild-type in 5 patients who had undertaken structured treatment interruption (STI). Guided therapy using GRT results led to 47.1% of patients with Q151M achieving a viral load (VL) below 500 copies and 54.3% achieving the same level in the absence of the MDR mutation. They conclude therefore, that “Q151M mutation appearance does not preclude the efficacy of salvage regimens based upon GRT.”⁽²⁾

However, there is much contention regarding the sustainability of resistant virus reverting to WT, or whether in fact embarking on treatment after an interruption in therapy will lead to a re-emergence of resistance already archived in proviral DNA.

Archived resistance mutations

A Belgian study of 10 treatment-experienced patients who had been undetectable for over 3 years demonstrated the value of monitoring drug resistant virus preserved in the cellular reservoir. Whilst measuring RNA can help determine resistance mutations at the time the drugs are being taken, proviral DNA can provide a profile of viral resistance that has evolved over time; a historical catalogue of the patient’s previous treatment history and associated mutations. The study cohort included sequential mono and dual therapy with AZT, 3TC, ddI or an NNRTI and had documented resistance to at least one of these compounds. Despite this history of suboptimal therapy, all patients had responded well to subsequent treatment containing dual PIs and 2 NRTIs. DNA sequencing of the proviral DNA revealed a mixture of both WT and resistant virus in 9 of the 10 patients studied. In 8 patients the percentage of proviral DNA was higher compared to the presence of WT whilst in 1 patient only WT virus was found. The study showed that even after several years of successful HAART therapy resistant genotypes may persist in the cellular proviral reservoir.

Measurement of proviral DNA may be particularly useful for monitoring in patients exposed to prior suboptimal therapy but for whom no plasma samples were previously stored.⁽³⁾

Replication and Viral Fitness

Although the terms viral fitness and replication competence are at times interchangeably applied, researchers distinguish between replication capacity as a phenomenon that can be measured in vitro and viral fitness as the effect of replication capacity on viral kinetics. Nick Hellmann of ViroLogic described replicative capacity as “a unique viral characteristic and not simply a surrogate for resistance”.

Dr Hellmann had previously demonstrated a clear connection between RC and treatment-outcome modelling RC values against treatment response. He clarified that a broad range of RC values may be observed with drug resistance and that commonly observed RT and PR mutations may have significantly different effects on HIV replication capacity, citing the examples of L90M as slightly less RC-compromised (RC=60%), i.e. less fit virus than the typical wild type (WT) (RC=100%), whilst a mixture of the Q151M complex with the M184V resulting in an even greater reduction (RC = 30%), i.e. an even less fit virus.

In studies involving structured treatment interruptions (STI), both in vitro measures of RC and in vivo measures of virus fitness were found to highly correlate and were associated with outcome in patients undergoing STI.

Dr Hellmann noted from the studies presented by Bob Grant from the University of California, San Francisco at the HIV Resistance Workshop in Seville last year that the lower the measure of RC, the slower the increase in VL and immune depletion, compared to the same WT virus with high RC. Even in patients with persistently detectable low VL, RC may be an important indicator that immune activity is more likely to remain stable or increase. Conversely, higher RC values are more likely to be associated with concordant failure (high VL, low CD4).⁽⁴⁾ This was also supported by Dr Hellmann’s observation that a lower RC in patients still treatment-naïve may indicate a higher CD4 cell count and a slower decline in CD4 cells over time and therefore help to facilitate discussions about when to start treatment. Although cut-off values for RC remain elusive, Richard Haubrich had previously defined RC values that may be of significance, suggesting that 35 marked a high RC value.⁽⁵⁾

Dr Hellmann proposed four potential uses for an RC assay:

• assisting in the decision to initiate early treatment or defer therapy,
• as an effective prognostic marker for disease progression,
• deciding when and if to switch therapy
• to determine the relative risk of infection in mother to child transmission (MTCT).

With the exception of data driven from ViroLogic and a novel assay described at the meeting from Charles Boucher’s laboratory at the University of Utrecht ⁽⁶⁾, assays for the measure of replication competence are not yet available. Lessons learned from resistance monitoring would suggest that whilst generating replication curves may be conceptually intriguing, measurement of replication capacity will need to be supported with effective clinical interpretation if it is to be meaningfully applied in the management of patient care.

Resistance evolution in non-B subtypes

Deenan Pillay from University College London presented a comprehensive overview of the prevalence and implications of non-B viral subtypes. He outlined the concerns normally associated with non-B HIV subtypes, in particular the relationship between subtype and disease progression noting a study by Kaleebu et al that linked subtype D with lower CD4 and a faster progression to death ⁽⁷⁾.

Dr Pillay also referred to the potential limitations in diagnostic technologies including the performance of resistance interpretation systems. This was confirmed at the workshop with data from Annemiek Vandamme’s laboratory in Leuven, Belgium, demonstrating significant discrepancies between viral clades and the reporting of resistance, especially for subtype J virus. In this study, discordance was found to be highest between the two interpretation systems assessed (Rega 5.5 and the Stanford Database) for predicting PI susceptibility and greatest concordance for response to NNRTI associated resistance. This inconsistency between PI and NNRTI resistance was explained by the greater propensity for polymorphic change in the PR region compared to evolution in the RT segment ⁽⁸⁾.

Dr Pillay however, focused his presentation on the evolution of resistance in non-B variants, confirming increasing evidence of emerging distinctions in pathways to resistance. Referring to a number of studies, he confirmed that resistance may in fact reflect unique dynamics in both the acquisition and impact of viral mutants. He commented that the predilection to stratify B and non-B virus in such a binary form was unhelpful as distinctions between variants in clades for example between C and D clades may be equally significant.

The mutation V106M that was significantly profiled at the CROI conference (2003) associated with higher prevalence in non-B and leading to loss of susceptibility to NNRTIs of 100-1000 fold may arise from a more efficient pathway then that observed in B subtype. For example, V106 in subtype C requires only one mutation to arrive at V106M (V106 GTG in subtype C followed by one mutational change to triplet codon ATG), whilst V106 in non-C may require 2 mutations to result in V106M associated resistance (V106 GTA) ⁽⁹⁾.

Dr Pillay also reiterated historical studies that had reported different mutational patterns observed in non-B subtypes including the absence of the D30N mutation in clade C patients receiving NFV-containing therapy. He confirmed that our increasing knowledge of resistance evolution in non-B variants may indeed have implications for decisions regarding second-line switch in therapy. Dr Pillay concluded by raising caution in the monitoring of resistance for non-B patients, suggesting that monitoring assays should reflect the viral heterogeneity increasingly observed in the clinic; identification of resistance-associated mutations should take into account the underlying subtype of the virus analysed.

Molecular Epidemiology of HIV – Transmission and Non-B

Of note, the meeting provided a useful update on the molecular epidemiology of HIV demonstrating a proliferating viral diversity across Europe.

The Non-B Story

Three studies highlighted the reconfiguration of HIV, profiling prevalence, subtype variations including circulating recombinant forms (CRF) and modes of transmission.⁽¹⁰⁾

Mapping Resistance

Data from a number of studies across Europe were presented that document the rates of transmission of resistance in recently infected, seroconverters and treatment-naïve patients. A summary of these include preliminary data from the SPREAD network of the first 5 countries assessed, the German seroconverter study which confirmed the first report of MDR in 2002 and a comparative cohort of experienced and naïve patients identified from 8 central infectious disease clinics in Poland. ^(11)(12)(13)