Resistance to Antiretroviral Drugs

• Plasma Trough Levels Correlate with Distinct Genetic Mechanisms During the Development of Amprenavir Resistance (Poster 465)
  Authored by R. Elston, S. Randall, R. Myers, M. Maguire, A. Rakik, M. Ait-Khaled, D. Stein, and W. Snowden
  View the original abstract

• Molecular Mechanism of I50V HIV-1 Protease Resistance and Cross-Resistance to Protease Inhibitors (Poster 466)
  Authored by R. Xu, W. Andrews, A. Spaltenstein, D. Danger, W. Dallas, L. Carter, L. Wright, and E. Furfine
  View the original abstract

Understanding the mechanisms of resistance to HIV drugs can assist with the strategic planning of treatment rounds. The presence of unique resistance mutations may allow for the sequential use of drugs of the same class. To a certain extent, this phenomenon has been exploited in initial HAART therapy with nelfinavir. The protease inhibitor amprenavir has been touted as an attractive drug for initial therapy because it can select for a unique resistance mutation at protease codon 50. High-level resistance to amprenavir usually requires other mutations in addition to codon 50. An alternative mutational pathway for amprenavir resistance exists, mediated by mutations at PR codons 84, 32/47, or 54. These mutations may be more prone to confer cross-resistance to other protease inhibitors.

A drawback to the initial use of amprenavir has been a very high pill count. One way to circumvent the pill burden issue is through the use of low-dose ritonavir in combination with amprenavir. The best dosing of boosted amprenavir remains undecided, but many physicians have had success with 600mg/100mg ritonavir twice daily. The net reduction in pill burden may make this combination more attractive to some patients. A yet unanswered question is how pharmacokinetic boosting of the parent drug (amprenavir) affects the genotypic or phenotypic patterns of resistance.

The susceptibility pattern of HIV with the I50V mutation to other protease inhibitors has been previously described. In abstract 466, Furfine and colleagues from GlaxoSmithKlein analyze the biochemical properties of HIV protease that contains the I50V mutation. They find that the enzyme has decreased binding affinity to amprenavir (140-fold) and ritonavir (50-fold) but only modest reduction in binding to nelfinavir (10-fold); compared to wild-type protease. Atomic level (crystalographic) modeling of APV, RTV, or NFV bound to the mutant enzyme provides a detailed look into why this resistance happens. When the protease is mutated at codon 50, a favorable environment for the binding of APV in the active site of protease is lost. APV is less likely to be bound to the inhibitor; hence HIV replication is less likely to be inhibited. The inhibitors RTV and NFV are effected by these spatial changes to lesser degrees, and therefore remain more active as viral drugs.

In a second related presentation, Dr. Elston from the United Kingdom presented new information about the role of drug levels and distinct pathways for the development of drug resistance to APV. This action-packed study had data derived from patients in the PRO2002, PRO2004, and PRO3006 dose-ranging studies of APV in combination with 2 nukes. Twenty patient isolates were obtained. Isolates that contained the I50V mutation had the highest degree of resistance to APV (approximately 23-fold), compared to wild type. From this small data set, it appears that there is a relationship between the dose of drug used and the likelihood of selecting the I50V mutation. Five of eight patients receiving 1,050mg bid APV and 0/6 patients taking 900mg bid selected the I50V mutation. This conclusion is corroborated by the plasma trough concentration data (available for 9 subjects). The median trough APV concentration among the 14 patients that selected the I54L/M pathway is 182.5 ng/ml whereas the median trough concentration among 13 patients who selected the I50V pathway is 421 ng/mL. The replicative capacity (a marker of viral fitness) appeared to be impaired in the presence of the I50V mutation.

These data are only of limited immediate relevance for the treatment community. This statement should not diminish from the potential importance of the studies. A mechanistic understanding of resistance is of critical importance in planning for possible sequential rounds of therapy. The suggestion of the retention of NFV susceptibility after APV failure is intriguing and needs to be confirmed by other studies. The ability to generalize the Elston study is limited because of the relatively small sample size. Taken at face value, this study suggests that drug resistance may occur through distinct pathways that depend on the concentration of inhibitor. Paradoxically, it is the higher drug concentration that may select for virus with the least amount of cross-resistance. This is in contrast to the "conventional wisdom" view of boosted protease inhibitors, such as lopinavir/ritonavir or indinavir/ritonavir: that higher degrees of resistance (and cross-resistance) are required to overcome higher drug levels. Will these data mitigate concern about the possibility of protease cross-resistance with boosted PIs? I think that the jury is still out. Stay tuned.