Genotypic and phenotypic cross-resistance patterns to lopinavir and amprenavir in protease inhibitor-experienced patients with HIV viremia

The Effect of Treatment-Associated Mutations on HIV Replication and Transmission Cycles

HIV/AIDS mortality has been decreasing over the last decade. While promising, this decrease correlated directly with increased use of antiretroviral drugs. As a natural consequence of its high mutation rate, treatments provide selection pressure that promotes the natural selection of escape mutants. Individuals may acquire drug-naive strains, or those that have already mutated due to treatment. Even within a host, mutation affects HIV tropism, where initial infection begins with R5-tropic virus, but the clinical transition to AIDS correlates with mutations that lead to an X4-tropic switch. Furthermore, the high mutation rate of HIV has spelled failure for all attempts at an effective vaccine. Pre-exposure drugs are currently the most effective drug-based preventatives, but their effectiveness is also threatened by viral mutation. From attachment and entry to assembly and release, the steps in the replication cycle are also discussed to describe the drug mechanisms and mutations that arise due to those drugs. Revealing the patterns of HIV-1 mutations, their effects, and the coordinated attempt to understand and control them will lead to effective use of current preventative measures and treatment options, as well as the development of new ones.

Salvage Therapy with Ritonavir-Boosted Amprenavir/Fosamprenavir: Virological and Immunological Response in Two Years Follow-up

OBJECTIVE

To evaluate the efficacy of salvage regimens containing ritonavir-boosted amprenavir (APV/r) or fosamprenavir (FPV/r) in heavily pretreated protease inhibitor (PI)-experienced HIV-1 patients.

METHOD

Evaluation of APV/r- or FPV/r-containing antiretroviral regimens in PI-experienced HIV-1 patients with 2 or more antiretroviral failures. Follow-up continued to 96 weeks with prospective collection of data.

RESULTS

54 episodes (48 on APV/r and 6 on FPV/r) were considered in 45 patients who had received a median of 5 prior antiretroviral regimens (range, 2-13) including a median of 3 PIs (range, 2-4). Median time of treatment at analysis was 72 weeks (range, 12-210). At baseline, plasma viral load (pVL) and CD4 cell count was 67,000 copies/mL and 167 cell/mm(3), respectively. At week 96, the median pVL was < 50 copies/mL and CD4 cell count was 519 cells/mm(3). Proportion of patients with pVL below detection was 62% at week 48 and 61% at week 96. Fifteen patients stopped treatment because of virologic failure; one presented a full resistance profile to APV/r, based on the ANRS 2003 resistance algorithm. Median trough APV plasma concentration 4 weeks after treatment initiation was 1406 ng/mL (range, 452-4321); dose adaptation was required in only 7 patients.

CONCLUSION

This study provides long-term follow-up of APV/r and FPV/r in the setting of salvage therapy, showing a high and sustained rate of virologic and immunologic response.

Cooperative effects of drug-resistance mutations in the flap region of HIV-1 protease

Understanding the interdependence of multiple mutations in conferring drug resistance is crucial to the development of novel and robust inhibitors. As HIV-1 protease continues to adapt and evade inhibitors while still maintaining the ability to specifically recognize and efficiently cleave its substrates, the problem of drug resistance has become more complicated. Under the selective pressure of therapy, correlated mutations accumulate throughout the enzyme to compromise inhibitor binding, but characterizing their energetic interdependency is not straightforward. A particular drug resistant variant (L10I/G48V/I54V/V82A) displays extreme entropy-enthalpy compensation relative to wild-type enzyme but a similar variant (L10I/G48V/I54A/V82A) does not. Individual mutations of sites in the flaps (residues 48 and 54) of the enzyme reveal that the thermodynamic effects are not additive. Rather, the thermodynamic profile of the variants is interdependent on the cooperative effects exerted by a particular combination of mutations simultaneously present.

Frequency and patterns of protease gene resistance mutations in HIV-infected patients treated with lopinavir/ritonavir as their first protease inhibitor

BACKGROUND

Selection of protease mutations on antiretroviral therapy (ART) including a ritonavir-boosted protease inhibitor (PI) has been reported infrequently. Scarce data exist from long-term cohorts on resistance incidence or mutational patterns emerging to different PIs.

METHODS

We studied UK patients receiving lopinavir/ritonavir as their first PI, either while naive to ART or having previously received non-PI-based ART. Virological failure was defined as viral load ≥ 400 copies/mL after previous suppression <400 copies/mL, or failure to achieve <400 copies/mL during the first 6 months. pol sequences whilst failing lopinavir or within 30 days after stopping were analysed. Major and minor mutations (IAS-USA 2008-after exclusion of polymorphisms) were considered. Predicted susceptibility was determined using the Stanford HIVdb algorithm.

RESULTS

Three thousand and fifty-six patients were followed for a median (IQR) of 14 (6-30) months, of whom 811 (27%) experienced virological failure. Of these, resistance test results were available on 291 (36%). One or more protease mutations were detected in 32 (11%) patients; the most frequent were I54V (n = 12), M46I (n = 11), V82A (n = 7) and L76V (n = 3). No association with viral subtype was evident. Many patients retained virus predicted to be susceptible to lopinavir (14, 44%), tipranavir (26, 81%) and darunavir (27, 84%).

CONCLUSIONS

This study reflects the experience of patients in routine care. Selection of protease gene mutations by lopinavir/ritonavir occurred at a much higher rate than in clinical trials. The mutations observed showed only partial overlap with those previously identified by structural chemistry models, serial cell culture passage and genotype-phenotype analyses. There remained a low degree of predicted cross-resistance to other widely used PIs.

Lopinavir/ritonavir in the treatment of HIV-1 infection: a review

Lopinavir/ritonavir is the first and only coformulated HIV-1 protease inhibitor (PI). Large clinical trials have demonstrated lopinavir/ritonavir's clinical efficacy in both antiretroviral-naïve and -experienced patients. The immunologic and virologic benefits of treatment with this agent have been proven in HIV-infected adults, adolescents, and children. Smaller studies support the use of lopinavir/ritonavir monotherapy as a therapeutic option in certain patients. The drug is characterized by a high genetic barrier to resistance, and appears to be more forgiving of non-adherence than earlier, unboosted PIs. The most frequent side effects observed are diarrhea, nausea, and vomiting. These gastrointestinal adverse effects are generally mild to moderate. Metabolic derangements, including hyperlipidemia and glucose intolerance, have also been observed in lopinavir/ritonavir recipients. As the menu of available antiretroviral agents continues to expand, lopinavir/ritonavir remains a proven and effective drug for the treatment of HIV infection.

Randomized study of dual versus single ritonavir-enhanced protease inhibitors for protease inhibitor-experienced patients with HIV

PURPOSE

To compare activity and safety of a regimen containing lopinavir/ritonavir (LPV/r) + fosamprenavir (FPV) to regimens with LPV/r or FPV + r and to test the hypothesis that a ritonavir-enhanced dual protease inhibitor (PI) regimen has better antiviral activity.

METHOD

This study was a multicenter, open-label, randomized study. HIV-infected adults with prior PI failure were selectively randomized based on prior PI experience to either LPV/r, FPV + r, or LPV/r + FPV. All patients received tenofovir DF and 1 to 2 nucleoside reverse transcriptase inhibitors.

RESULTS

Baseline characteristics were similar across arms. Study enrollment and follow-up were stopped early (N = 56) because pharmacokinetic analyses showed significantly lower LPV and FPV exposures in the dual-PI arm. At Week 24, proportions achieving >1 log10 decline in HIV RNA or <50 copies/mL in the dual-PI versus single-PI arms combined were 75% vs. 61% in intent-to-treat (ITT, p = .17) and 100% vs. 64% in as-treated (AT) analyses (p = .02), respectively. Median CD4+ T cell/mm3 increases were 81 vs. 41 (ITT, p = .4) and 114 vs. 43 (AT, p = .08), respectively. Clinical events and toxicity rates were not different between arms.

CONCLUSION

The trial was unable to show a difference between dual versus single PIs in ITT analyses but favored dual PIs in AT analyses.

Mutational patterns and correlated amino acid substitutions in the HIV-1 protease after virological failure to nelfinavir- and lopinavir/ritonavir-based treatments

Human immunodeficiency virus type 1 (HIV-1) antiviral drug resistance is a major consequence of therapy failure and compromises future therapeutic options. Nelfinavir and lopinavir/ritonavir-based therapies have been widely used in the treatment of HIV-infected patients, in combination with reverse transcriptase inhibitors. The aim of this observational study was the identification and characterization of mutations or combinations of mutations associated with resistance to nelfinavir and lopinavir/ritonavir in treated patients. Nucleotide sequences of 1,515 subtype B HIV-1 isolates from 1,313 persons with different treatment histories (including naïve and treated patients) were collected in 31 Spanish hospitals over the years 2002-2005. Chi-square contingency tests were performed to detect mutations associated with failure to protease inhibitor-based therapies, and correlated mutations were identified using statistical methods. Virological failure to nelfinavir was associated with two different mutational pathways. D30N and N88D appeared mostly in patients without previous exposure to protease inhibitors, while K20T was identified as a secondary resistance mutation in those patients. On the other hand, L90M together with L10I, I54V, A71V, G73S, and V82A were selected in protease inhibitor-experienced patients. A series of correlated mutations including L10I, M46I, I54V, A71V, G73S, and L90M appeared as a common cluster of amino acid substitutions, associated with failure to lopinavir/ritonavir-based treatments. Despite the relatively high genetic barrier of some protease inhibitors, a relatively small cluster of mutations, previously selected under drug pressure, can seriously compromise the efficiency of nelfinavir- and lopinavir/ritonavir-based therapies.

Amprenavir and Lopinavir Pharmacokinetics following Coadministration of Amprenavir or Fosamprenavir with Lopinavir/Ritonavir, with or without Efavirenz

Background

Amprenavir (APV), fosamprenavir (FPV), lopinavir (LPV), ritonavir (RTV) and efavirenz (EFV) are to varying degrees substrates, inducers and inhibitors of CYP3A4. Coadministration of these drugs might result in complex pharmacokinetic drug-drug interactions.

Methods

Two prospective, open-label, non-randomized studies evaluated APV and LPV steady-state pharmacokinetics in HIV-infected patients on APV 750 mg twice daily + LPV/RTV 533/133 mg twice daily with EFV ( n=7) or without EFV ( n=12) + background nucleosides (Study 1) and after switching FPV 1,400 mg twice daily for APV ( n=10) (Study 2).

Results

In Study 1 EFV and non-EFV groups did not differ in APV minimum plasma concentration (Cmin; 1.10 versus 1.06μg/ml, P=0.89), area under the concentration-time curve (AUC; 17.46 versus 24.34μg•h/ml, P=0.22) or maximum concentration (Cmax; 2.61 versus 4.33 μg/ml, P=0.08); for LPV there was no difference in Cmin (median: 3.66 versus 6.18μg/ml, P=0.20), AUC (81.84 versus 93.75 μg•h/ml, P=0.37) or Cmax (10.36 versus 10.93 μg/ml, P=0.61). In Study 2, after switching from APV to FPV, APV Cmin increased by 58% (0.83 versus 1.30 μg/ml, P=0.0001), AUC by 76% (19.41 versus 34.24 μg•h/ml, P=0.0001), and Cmax by 75% (3.50 versus 6.14, P=0.001). Compared with historical controls, LPV and RTV pharmacokinetics were not changed. All treatment regimens were well tolerated. Seven of eight completers (88%) maintained HIV-1 RNA <400 copies/ml 12 weeks after the switch (1 lost to follow up).

Conclusions

EFV did not appear to significantly alter APV and LPV pharmacokinetic parameters in HIV-infected patients taking APV 750 mg twice daily + LPV 533/133 mg twice daily. Switching FPV 1,400 mg twice daily for APV 750 mg twice daily resulted in an increase in APV Cmin, AUC, and Cmax without changing LPV or RTV pharmacokinetics or overall tolerability.

Structural analysis of an HIV-1 protease I47A mutant resistant to the protease inhibitor lopinavir

We have identified a rare HIV-1 protease (PR) mutation, I47A, associated with a high level of resistance to the protease inhibitor lopinavir (LPV) and with hypersusceptibility to the protease inhibitor saquinavir (SQV). The I47A mutation was found in 99 of 112,198 clinical specimens genotyped after LPV became available in late 2000, but in none of 24,426 clinical samples genotyped from 1998 to October 2000. Phenotypic data obtained for five I47A mutants showed unexpected resistance to LPV (86- to >110-fold) and hypersusceptibility to SQV (0.1- to 0.7-fold). Molecular modeling and energy calculations for these mutants using our structural phenotyping methodology showed an increase in the binding energy of LPV by 1.9-3.1 kcal/mol with respect to the wild type complex, corresponding to a 20- to >100-fold decrease in binding affinity, consistent with the observed high levels of LPV resistance. In the WT PR-LPV complex, the Ile 47 side chain is positioned close to the phenoxyacetyl moiety of LPV and its van der Waals interactions contribute significantly to the ligand binding. These interactions are lost for the smaller Ala 47 residue. Calculated binding energy changes for SQV ranged from -0.4 to -1.2 kcal/mol. In the mutant I47A PR-SQV complexes, the PR flaps are packed more tightly around SQV than in the WT complex, resulting in the formation of additional hydrogen bonds that increase binding affinity of SQV consistent with phenotypic hypersusceptibility. The emergence of mutations at PR residue 47 strongly correlates with increasing prescriptions of LPV (Spearman correlation r(s) = 0.96, P < .0001).

Confronting the emergence of drug-resistant HIV type 1: impact of antiretroviral therapy on individual and population resistance

Resistance to antiretroviral agents, and in particular the increasing levels of transmitted resistant virus could offset the substantial gains won with potent antiretroviral therapy. Primary and acquired antiretroviral resistance rates reflect the relative usage of different antiretroviral drugs in the population, as well as the inherent genetic barrier to the development of resistance associated with individual drugs. Data on antiretroviral resistance rates, gleaned from the growing HIV-1-infected population treated with a continuously increasing number of antiretroviral drugs and drug combinations, provide insights into patient management approaches for delaying the emergence of resistance and minimizing the degree of resistance. Evolving data suggest that the relative ease by which HIV-1 escapes the selective pressure of chronic drug exposure varies for the different antiretroviral drug classes and individual antiretroviral drugs. The development of resistance in vivo can be anticipated based on these data, in conjunction with the individuals treatment history and resistance testing results. These in turn can guide the judicious use of antiretroviral drugs to attain optimal treatment responses and to preserve therapeutic options for the time when antiretroviral-resistant strains emerge. The recent developments of new antiretroviral drugs, including the use of boosted protease inhibitors, suggest that treatment strategies can limit the development of resistance.

Treatment exhaustion of highly active antiretroviral therapy (HAART) among individuals infected with HIV in the United Kingdom: multicentre cohort study

OBJECTIVES

To investigate whether there is evidence that an increasing proportion of HIV infected patients is starting to experience increases in viral load and decreases in CD4 cell count that are consistent with exhaustion of available treatment options.

DESIGN

Multicentre cohort study.

SETTING

Six large HIV treatment centres in southeast England.

PARTICIPANTS

All individuals seen for care between 1 January 1996 and 31 December 2002.

MAIN OUTCOME MEASURES

Exposure to individual antiretroviral drugs and drug classes, CD4 count, plasma HIV RNA burden.

RESULTS

Information is available on 16,593 individuals (13,378 (80.6%) male patients, 10,340 (62.3%) infected via homosexual or bisexual sex, 4426 (26.7%) infected via heterosexual sex, median age 34 years). Overall, 10,207 of the 16,593 patients (61.5%) have been exposed to any antiretroviral therapy. This proportion increased from 41.2% of patients under follow up at the end of 1996 to 71.3% of those under follow up in 2002. The median CD4 count and HIV RNA burden of patients under follow up in each year changed from 270 cells/mm3 and 4.34 log10 copies/ml in 1996 to 408 cells/mm3 and 1.89 log10 copies/ml, respectively, in 2002. By 2002, 3060 (38%) of patients who had ever been treated with antiretroviral therapy had experienced all three main classes. Of these, around one quarter had evidence of "viral load failure" with all these three classes. Patients with three class failure were more likely to have an HIV RNA burden > 2.7 log10 copies/ml and a CD4 count < 200 cells/mm3.

CONCLUSIONS

The proportion of individuals with HIV infection in the United Kingdom who have been treated has increased gradually over time. A substantial proportion of these patients seem to be in danger of exhausting their options for antiretroviral treatment. New drugs with low toxicity, which are not associated with cross resistance to existing drugs, are urgently needed for such patients.

HIV protease mutations associated with amprenavir resistance during salvage therapy: importance of I54M

OBJECTIVE

To examine the clinical significance of HIV protease mutations detected before and after therapy with amprenavir.

STUDY DESIGN

Serial plasma HIV loads and protease gene mutations were monitored in 31 patients who received amprenavir, including 19 who had been exposed to other protease inhibitors (salvage therapy). Recombinant phenotyping was used to assess the significance of new mutations appearing after amprenavir therapy.

RESULTS

After 6-8 months of amprenavir, 4 treatment-naïve and 5 salvage patients had an undetectable plasma HIV load, while 12 other salvage patients showed less than 10-fold HIV load reduction. HIV protease mutations associated with amprenavir resistance included I84V, I50V, I47V, V32I, and I54M. Among mutations newly detected after amprenavir treatment, I54M occurred in six cases, I54L in two cases, M46I in two cases, I47V in one case and I50V in one case. When compared with pretreatment plasma without the mutation, recombinant phenotyping showed that I54M increased the amprenavir resistance by at least 6-fold, resulting in up to 48-fold resistance over a drug-sensitive control.

CONCLUSIONS

Protease I54M frequently appears after amprenavir therapy, and when combined with pre-existing mutations, leads to high-level amprenavir resistance and treatment failure.

Simple Linear Model Provides Highly Accurate Genotypic Predictions of HIV-1 Drug Resistance

Drug resistance is a major obstacle to the successful treatment of HIV-1 infection. Genotypic assays are used widely to provide indirect evidence of drug resistance, but the performance of these assays has been mixed. We used standard stepwise linear regression to construct drug resistance models for seven protease inhibitors and 10 reverse transcriptase inhibitors using data obtained from the Stanford HIV drug resistance database. We evaluated these models by hold-one-out experiments and by tests on an independent dataset. Our linear model out-performed other publicly available genotypic interpretation algorithms, including decision tree, support vector machine and four rules-based algorithms (HIVdb, VGI, ANRS and Rega) under both tests. Interestingly, our model did well despite the absence of any terms for interactions between different residues in protease or reverse transcriptase. The resulting linear models are easy to understand and can potentially assist in choosing combination therapy regimens.

Cross-resistance patterns among HIV protease inhibitors

Acquired resistance to antiretroviral agents is an established sequela of HIV pharmacotherapy. Viral mutations can confer reduced susceptibility to antiretroviral medications, resulting in virologic and clinical failure in more than half of treated patients. Cross-resistance that can develop within each drug class leads to the progressive loss of future therapeutic options for individual patients. Although protease inhibitors (PIs) are a potent class of antiretrovirals, resistance can still develop rapidly, and multiple-PI resistance has become a serious, growing clinical problem. Development of rational treatment strategies that recognize specific patterns of cross-resistance among PIs are needed to help clinicians choose the most appropriate PI. Rational sequencing of PI use should be based on genotypic and phenotypic resistance testing. Maintaining higher drug plasma levels or using specific PI combinations may also diminish PI cross-resistance. New agents that are less likely to induce or be susceptible to cross-resistance will be of value in HIV treatment. This article reviews the acquisition of resistance to currently available PIs, discussing the drug-specific mutational patterns and evidence of clinical cross-resistance. The resistance profiles of two newer PIs, atazanavir and tipranavir, are also presented.

The Influence of Protease Inhibitor Resistance Profiles on Selection of HIV Therapy in Treatment-Naive Patients

Although protease inhibitors (PIs) have dramatically improved outcomes in HIV-infected patients, half still fail treatment with PI-based combination therapy. Genetic pressure from incomplete viral suppression rapidly selects for HIV variants with protease gene mutations that confer reduced susceptibility to PI drugs. A number of specific amino acid substitutions have been associated with PI resistance. However, high-level resistance to individual PIs requires the accumulation of several primary and secondary mutations, developing along drug-specific, step-wise pathways. HIV variants resistant to saquinavir and ritonavir usually contain L90M and V82A substitutions, respectively. Indinavir resistance may be linked to substitutions at positions 46 or 82. Resistance to nelfinavir is primarily associated with D30N but may alternatively be found with L90M. Resistance during exposure to amprenavir can follow development of I50V, which also may confer resistance to lopinavir. Failure during treatment with atazanavir is closely linked to I50L. The overlapping of these pathways can lead to multiple-PI resistance, limiting therapeutic options in antiretroviral-experienced patients. Reduced susceptibility to more than one PI is most likely to be associated with amino acid substitutions at six positions: 10, 46, 54, 82, 84 and 90. Other mutations (D30N, G48V, I50V or I50L) are relatively specific for particular PIs and are less likely to produce cross resistance. Certain resistance mutations selected by exposure to one PI may actually increase susceptibility to others. Patients newly diagnosed with HIV infection are increasingly found to harbour virus that is resistant to the more commonly used drugs. Newer PIs may select for mutations that result in less cross resistance with older agents.

HIV type 1 genotypic resistance in a clinical database correlates with antiretroviral utilization

We established a database of HIV-1 reverse transcriptase (RT) and protease (PR) sequences and mutations to monitor the prevalence of antiretroviral drug resistance and mutational patterns in clinical samples submitted for testing to a major U.S. reference laboratory. At the end of 1998, 80% of the clinical samples tested harbored HIV strains with genotypically predicted resistance to at least one antiretroviral (ARV) drug. By the third quarter of 2002, the frequency of genotypically predicted resistance declined to 65% of samples tested. The prevalence of both PR and nucleoside RT inhibitor resistance declined over this period, while an increase in resistance to nonnucleoside RT inhibitors was found. These genotypic results strongly correlated with a nationwide decrease in the prescription of PR and nucleoside RT inhibitors, and an increase in the prescription of nonnucleoside RT inhibitors over the time period. The increased number of strains that were genotypically sensitive to all classes of ARV probably indicates an increase in genotypic assay use in ARV-naive individuals, however, the trends and correlations in this data set were similar when evaluated after removal of genotypically sensitive strains. Continued monitoring of ARV resistance prevalence, patterns, and utilization trends in clinical databases provides insight into the evolving relationship between clinical practice and ARV resistance.

Drug resistance mutations and newly recognized treatment-related substitutions in the HIV-1 protease gene: Prevalence and associations with drug exposure and real or virtual phenotypic resistance to protease inhibitors in two clinical cohorts of antiretroviral experienced patients

This study aimed at identifying HIV-1 protease amino acid changes associated with protease inhibitor (PI) exposure and susceptibility. New amino acid substitutions were correlated with the number of experienced PIs, reaching statistical significance only for those at positions 3, 44, and 74. The correspondence multivariate model demonstrated that > or =3 experienced PIs and substitutions or mutations at positions 3, 46, 54, 73, 74, and 84 were correlated with PI cross-resistance, including resistance for lopinavir and amprenavir in this cohort of patients who were naive for these drugs.

New anti-HIV protease inhibitors provide more treatment options

For several years, protease inhibitor (PI)-containing antiretroviral treatment (ART) regimens have demonstrated long-term virologic and immunologic benefits and good durability of response. However, first-generation PIs have been associated with high pill burdens, gastrointestinal side effects, perturbation of lipid levels and glucose metabolism, and, in some cases, food and hydration requirements. Coadministration of low-dose ritonavir with PIs has enhanced their pharmacokinetic profile (lower doses, fewer pills, less frequent dosing schedules) and pharmacodynamics (increased potency, especially against resistant viruses) but has also been associated with increases in lipid levels. Two new PIs, atazanavir and 908 (fosamprenavir), may offer salvageable PI treatment options and may also address issues of potency, tolerability, and convenience by requiring fewer pills and causing fewer lipid and glucose perturbations than current PI options. The availability of these novel PIs may improve longterm treatment options for many patients.

Drug Resistance Genotypes Predict Response to Amprenavir-Containing Regimens in Highly Drug-Experienced HIV-1-Infected Patients

We have undertaken a study of virological responses to amprenavir-containing antiretroviral regimens, during the expanded access programme within the UK. Ninety-five HIV-1-infected patients were included for which virological and immunological follow-up was available for 75, and baseline drug resistance data available for 51. These were highly drug-experienced patients, having previously received a median of nine antiviral drugs, within all available classes. Eighty-eight percent of patients had a virological response to the new regimen, with a median maximal decline of 1.45 log10 copies/ml, and 34% of patients reached <400 copies/ml on treatment. Although 68% of patients with resistance data had protease inhibitor resistance mutations, only 10% patients had key amprenavir resistance mutations, and virological response was predicted by the number of active drugs utilized in the amprenavir-containing regimen, as determined by the baseline genotypic resistance test. Other independent predictors of viral load decline were a higher baseline viral load and fewer previous antiviral drugs. We conclude that amprenavir can contribute to antiviral efficacy in salvage regimens, and that resistance testing may help to optimize its use in this scenario. New formulations of amprenavir, together with boosted regimens, may enhance the activity in the presence of protease inhibitor-resistant virus.