Superior drugs with increased anti-HIV activity, decreased toxicity, and/or improved schedules of administration are needed to increase the odds that previously untreated HIV-infected patients will achieve and maintain prolonged virologic suppression. In addition, drugs with unique mechanisms of action are required to circumvent the drug resistance that frequently arises in the viruses of previously treated patients and render them more responsive to salvage therapy regimens.

Unfortunately, past clinical experience has shown that predictions about the mechanisms of resistance to newly developed drugs and of the extent of cross-resistance between newly developed drugs and standard therapies are frequently based on insufficient or misleading data. Negative outcomes in tens of thousands of patients have been necessary before it has become evident that drugs once touted as being unique with respect to resistance mutation patterns are ineffective in previously treated persons.

How can we better interpret pre-clinical and early clinical drug resistance data to prevent this type of tragic error from recurring? How can we differentiate drugs that merely imitate already available treatments from those that will truly strengthen our anti-HIV arsenal?

Currently, a great deal of significance is attached to viral mutations arising during in vitro passage experiments and phase I clinical trials of a new drug. These data identify panels of mutations that are sufficient to cause resistance to the drug under evaluation without proving that these mutations are in fact necessary for the development of drug resistance. Analysis of genotypic and phenotypic data at later stages of clinical testing may allow for more accurate prediction of a new drug's efficacy in previously treated patients. However, this critical data is unavailable for many drugs now in clinical development.

These drugs include the following PIs: ABT-378 (iopinovir), tipranavir (PNU-140690), and AG1776 (formerly JE-2147); the following NRTIs: PMPA (tenofovir), FTC (emtricitibine), F-ddA (lodenosine), and d-OTC; the following NNRTIs: AG1549 (formerly S-1153), MKC-442 (emivirine), and GW420867X; and a prototype for a new class of drugs: the fusion inhibitor T-20. Some of these drugs may end up being superior to those currently available; however, only the fusion inhibitors are guaranteed to circumvent virologic cross-resistance in patients treated with currently approved anti-HIV drugs.

ABT-378 is a PI in phase II/III clinical development [1,2]. In vitro passage experiments have shown that 340-fold ABT-378 resistance develops when the virus develops mutations at codons 10, 16, 46, 47, 84, and 91 [3]. In contrast, mutations at codon 82, the most common mutations associated with indinavir and ritonavir resistance, cause a relatively small decrease in ABT-378 susceptibility [4]. However, data on the genetic mechanisms of resistance in vivo have not been published and the subject of viral cross-resistance to ABT-378 and other PI was only skirted in a recent abstract that presented no raw genotypic or phenotypic data [5].

ABT-378, which appears headed for FDA approval, will be used in combination with low-doses of ritonavir. This combination may be as effective as ritonavir-saquinavir and ritonavir-indinavir in previously untreated patients or in patients harboring isolates with low-level PI resistance. Ideally, genotypic and phenotypic data on clinical isolates from patients failing therapy with other PIs or with ABT-378 will be made publicly available to allow assessment of the extent of cross-resistance between the therapies. These data may allow physicians to determine which PI-resistant isolates are most likely to respond to ABT-378.

Tipranavir is an experimental PI discussed by Dr. Brian Conway in this issue (see "Tipranavir: The Answer to Protease Inhibitor Resistance?"). Phase II clinical data will be presented for the first time at the upcoming 7th Conference on Retroviruses and Opportunistic Infections later this month. Tipranavir has less anti-HIV activity in vitro than ABT-378 and suffers from bioavailablility problems in vivo. The range in susceptibility (e.g., IC₅₀) between wild-type isolates and tipranavir-resistant isolates has not been reported. This makes it difficult to assess recent claims that most PI-resistant clinical isolates retain susceptibility to tipranavir [6]. Moreover, the data underlying these claims have not been made publicly available.

PMPA is an acyclic NRTI phosphonate (a.k.a. "nucleotide" RT inhibitor) that has been demonstrated to decrease plasma HIV-1 RNA levels by approximately 1 log over 7 days in previously untreated patients and over 24 weeks when added to a stable anti-HIV treatment regimen [7,8]. There are no published sequence data on HIV-1 isolates from patients receiving PMPA and the dynamic susceptibility range or fold-resistance required to abrogate PMPA activity have not been determined. Many NRTI-resistant isolates have been reported to be susceptible to PMPA [9,10], but as with tipranavir, the significance of these susceptibility results would be better understood if the dynamic susceptibility range for this drug were known.

Other NRTI under development include FTC, F-ddA, and d-OTC. Although viral isolates resistant to 3TC are also highly resistant to FTC, FTC may be a better alternative to 3TC because of its once-daily dosing and similar potency [11,12]. F-ddA is a weak NRTI that is touted as active against isolates resistant to other NRTI inhibitors. HIV-1 isolates cultured in the presence of F-ddA have developed different mutations from those developing with other NRTI [13]. However, clinical isolates resistant to other NRTI have been found to also have decreased F-ddA susceptibility [9]. In a phase I/II trial, d-OTC monotherapy decreased plasma HIV-1 RNA levels by approximately 1.3 logs at the two highest dosages [14]. Mutations at codon 184 have been reported to develop during in vitro d-OTC passage [15]. However, the dynamic susceptibility range for this drug, the activity of this drug against isolates resistant to other NRTI, and the genetic mechanism(s) of resistance to this drug in clinical settings have not been published.

AG-1549 appears to be the most promising of the experimental NNRTI. It is as potent as efavirenz and is highly active against isolates containing most of the common NNRTI mutations including K103N and Y181C [16,17]. With the exception of V106A, mutations developing during in vitro passage experiments with AG-1549 (F227L and L234I) are not known to cause resistance to the currently approved NNRTI [18]. MKC-442 is the experimental NNRTI at the most advanced stage of clinical development. The combination of MKC-442 with d4T and ddI reduced plasma HIV-1 RNA levels to <400 copies/mL in about 80% of 198 patients over a 16-week period [19]. However, the RT mutations in MKC-442-resistant clinical isolates include K103N and are usually similar to those developing during treatment with currently approved NNRTIs. There is only limited clinical data on GW420867, but in vitro studies suggest that this drug is also highly cross-resistant to the currently approved NNRTI [20].

Related HIVresistanceWeb articles:

Tipranavir: The Answer to Protease Inhibitor Resistance?

JE-2147: A New Dipeptide Protease Inhibitor Potentially Active Against PI-Resistant HIV

New Antiretroviral Drugs (Report From the 3rd International Workshop on HIV Drug Resistance).