Article Text


New drugs for treating drug resistant HIV-1
  1. A Isaac1,
  2. D Pillay2
  1. 1Whittal Street Clinic, Birmingham, UK
  2. 2PHLS Antiviral Susceptibility Reference Unit, Birmingham, UK
  1. Correspondence to:
 Dr Deenan Pillay, Department of Virology, Windeyer Institute, University College Hospital, 46 Cleveland Street, London W1T 4JF, UK; 

Statistics from

Clinical management of virological failure remains an important and difficult issue for HIV physicians

One of the major barriers to successful treatment of HIV-1 infection is the emergence of drug resistant virus.1 The greatest impact of resistance is that it limits the effectiveness of subsequent antiretroviral combinations following initial drug failure. At a population level, more than 50% of patients who fail therapy do so with viruses resistant to drugs within at least one class of drug, with 15–20% with resistance to drugs within all three currently available classes (Health Protection Agency, unpublished data). Therefore, there is an urgent requirement for new drugs with activity against such resistant species. Over the past year or so, there has been a welcome upsurge in data presented on new drugs, both within existing classes and new classes, with the promise of more effective therapies for HIV resistant viruses (see table 1).

Table 1

Some new drugs recently approved and in clinical development with possible activity against resistant viruses


Nucleoside analogue drugs have been the mainstay of HIV therapy since zidovudine was first licensed in 1988,2 and it is not surprising that resistance to this class of drugs is most common at a population level (Health Protection Agency, unpublished data). Despite some specific signature mutations for individual nucleoside analogues, there is increasing evidence for cross resistance between certain drugs, such as ZDV and d4T, as well as the emergence of mutations conferring broad cross resistance, such as the 69 insertions, and the Q151M constellation of mutations within reverse transcriptase. Interesting data have been presented for alovudine, a thymidine analogue previously shown to have considerable toxicity in the clinic. Now reassessed at lower doses, activity is observed in patients with ZDV/d4T resistance (up to five thymidine analogue resistance associated mutations, TAMS) although antagonism between these thymidine analogues is observed when used in combination.3,4 More data are awaited for this rejuvenated compound. Amdoxovir (DAPD) is a new nucleoside analogue prodrug whose oral administration leads to a rapid in vivo conversion to (−)-α-d-dioxalane guanosine (DXG). Resistance to this drug in the laboratory appears to involve the K65R and L74V mutations, similar to those observed for abacavir (although ABC failure is rarely associated with these mutations in the clinic).5 Phase I/II studies demonstrate a reasonable activity of this drug against nucleoside analogue resistant viruses, although more data are needed before clarifying its potential role. However, activity in vitro is compromised by the multinucleoside resistance mutation Q151M together with changes at amino acid 69 of reverse transcriptase, which may limit its role in higher nucleoside analogue experienced patients.6,7 The drug attracting most excitement at present is the recently approved nucleotide analogue tenofovir, which appears to be unencumbered by the toxicity problems of its cousin, adefovir. As for many other drugs, the HIV mutations in reverse transcriptase associated with reduced activity in the clinic are not necessarily those selected by tenofovir in the laboratory (K65R). This is because the drug has been most widely tested in drug experienced patients in whom resistant virus already exists and predictors of poor response can be identified. Thus, common nucleoside analogue resistance mutations such as M41L, L210W (possibly a key marker in this respect), and T215Y appear to reduce, although not negate, clinical efficacy; nevertheless, the widespread use of tenofovir in salvage therapy and promising first line treatment trial data suggest that it represents an important addition to our antiviral armoury.8–10


The phenomenon of extensive cross resistance between nNRTIs is one of the more widely accepted truths of HIV drug resistance, owing to the small binding site for this group of drugs within the viral reverse transcriptase.11 The key mutations in this regard are K103N, T181C, and G190A/E, all of which compromise nevirapine, efavirenz, and delavirdine responses, and this cross resistance represents a major limitation of the class as a whole. However, two new compounds, TMC 125 and TMC120, appear to have activity against such resistance viruses, both in vitro and in vivo.12,13 Another compound (capravirine) demonstrated activity against a virus bearing the K103N or V106A or L100I single mutation, although high level resistance to this drug was reported in the presence of mutations at codon 181.14 It appears not so much that different patterns of resistance mutations are observed with these new nNRTI drugs, but rather that emergence of resistance is much slower than existing nNRTIs—note that single dose nevirapine in pregnancy is sufficient to select for resistant mutants—and that the well recognised nNRTI mutations have a marginal, and possibly clinically irrelevant, impact on fold susceptibility. It is argued that these properties are a function of the unique structures of these second generation nNRTIs, in the context of binding to the RT enzyme. We look forward to more extensive clinical trial data for both these drugs.


Issues of resistance and cross resistance are particularly pertinent to the protease inhibitor class of drugs. Many claims have been made on the apparent uniqueness of resistance patterns for specific drugs, based on in vitro data, which do not then translate into clinical benefit for that drug in PI experienced patients. Two new PIs have now undergone initial clinical evaluation. Atazanavir (ATZ), soon to be available within an expanded access programme, demonstrated different resistance profiles when used in PI naive or PI experienced patients. In the former group, resistance emerges with the I50L and A71V mutations.15 This is a unique combination since amprenavir resistance mutations include a different amino acid change at position 50 (namely, I50V), although the A71V mutation is a polymorphism (not infrequently observed in the absence of PI therapy). By contrast, in PI experienced patients, some level of cross resistance between atazanavir and other PIs was apparent.16 and therefore the utility of this drug as a second line PI may be limited. Clinical data for this scenario are awaited at the time of writing. Clinical data have also been presented for tipranavir, which shows potency against viruses containing a large variety of PI resistant mutants in vitro.17 Clinical activity was indeed observed in PI experienced patients, with a suggestion that a very large number of PI resistant mutations were required to compromise activity.18 More work is required to further clarify such “clinical cut offs” whereby clinicians can be guided on the likely effect of this new drug in a patient with existing PI resistant virus.


Data are now emerging from the trials of T-20 (enfuvirtide), the first fusion inhibitor to enter the clinic. Since the phase III trials were undertaken in heavily pretreated patients it is not surprising that failure rates (lack of full suppression) were relatively high overall; however, this affords the opportunity to characterise the emergence of resistance.19,20 Data from phase II studies demonstrate that the majority of such failure patients had mutations in the gp41 region targeted by the drug—namely, between amino acids 36–45, which indeed confirms that activity of the drug is mediated through the proposed mechanism.21 Since variation in this region is very rare in T-20 naive patients, including those infected with non-subtype B viruses, it can be assumed that previous RT inhibitor and PI therapy will not compromise T-20 activity itself.22 The key issue with use of T-20 in salvage therapy will therefore be the choice of other active drugs to combine with it. Of interest, the second generation fusion inhibitor T-1249 appears to be active against most T-20 resistance mutants, although this is based on in vitro evidence alone.23


Despite the undoubted success of antiretroviral therapy clinical management of virological failure remains an important and difficult issue for HIV physicians. Since such patients often have drug resistant virus, the choice of new combinations is often based, at least to some extent, on our knowledge of resistance characteristics of available drugs. We have summarised the data on a whole series of new drugs within existing and new classes. After some years of promising in vitro data, these drugs have demonstrated promise in clinical trials, with particular interest focused on unique resistance patterns, or the slow development of resistance. As further clinical trial data are presented for new drugs, it is important for HIV physicians to ask two specific questions. Firstly, what are the resistance patterns at baseline, which define success or failure of this new drug in antiretroviral experienced patients and, secondly, what are the resistance correlates of failure when used as a first line drug? It is answers to these questions that will contribute to identifying the optimal role of these promising new drugs in routine clinical practice.

Clinical management of virological failure remains an important and difficult issue for HIV physicians


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.