PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of aacPermissionsJournals.ASM.orgJournalAAC ArticleJournal InfoAuthorsReviewers
 
Antimicrob Agents Chemother. 2009 October; 53(10): 4153–4158.
Published online 2009 July 13. doi:  10.1128/AAC.00041-09
PMCID: PMC2764193

Use of Different Inhibitory Quotients To Predict Early Virological Response to Tipranavir in Antiretroviral-Experienced Human Immunodeficiency Virus-Infected Patients[down-pointing small open triangle]

Abstract

Information about the relationship between pharmacological parameters and an early virological response to tipranavir (TPV) is scarce. Human immunodeficiency virus (HIV)-infected patients who had received TPV as part of a salvage regimen were analyzed retrospectively. A virological response was defined as a decline in the HIV RNA level of ≥1 log unit or to <50 copies/ml between weeks 4 and 12 of therapy. The virtual inhibitory quotient (vIQ) was calculated as the ratio of the TPV plasma trough concentration (Ctrough)/virtual change in the 50% inhibitory concentration. Three genotypic inhibitory quotients (gIQs) were calculated by using different TPV resistance mutation scores (from the International AIDS Society—USA [IAS-USA], Randomized Evaluation of Strategic Intervention in Multidrug-Resistant Patients with Tipranavir [RESIST], and Agence Nationale de Recherches sur le Sida et les Hépatites Virales [ANRS] trials). The sensitivities, specificities, positive predictive values (PPVs), negative predictive values (NPVs), and likelihood ratios for a positive result (LHR+) and a negative result (LHR) [LHR+ = sensitivity/(1 − specificity); LHR = (1 − sensitivity)/specificity] were calculated. A total of 57 HIV-infected patients were analyzed. A virological response was achieved by 77% of the patients. TPV resistance mutations, TPV Ctrough, vIQs, and gIQs were all significantly associated with a virological response. The vIQ had the best PPV and NPV (97% and 78%, respectively). The values of the LHR+ were 7.8 for vIQ, 3.4 for the RESIST gIQ, 3.3 for the IAS-USA gIQ, 3.1 for the ANRS gIQ, 2.2 for TPV Ctrough, and 1.3 for the IAS-USA and RESIST scores. The values of LHR were 0 for the RESIST score, 0.07 for vIQ, 0.09 for the IAS-USA score, 0.27 for the RESIST gIQ, 0.32 for the IAS-USA gIQ, 0.37 for the ANRS gIQ, and 0.48 for TPV Ctrough. HIV-infected patients who initiate a salvage regimen based on TPV may benefit from baseline drug resistance testing and TPV plasma concentration determination, as vIQ is the best predictor of a virological response.

Tipranavir (TPV) is a nonpeptidic human immunodeficiency virus (HIV) protease inhibitor (PI) that has been demonstrated to have activity against multidrug-resistant HIV type 1 (HIV-1) isolates (2, 4, 7, 12). On the basis of its differential biochemical structure, the resistance profile of TPV seems to differ from the resistance profiles of the rest of the PIs. Twenty-one mutations located at 17 codons have been reported to be associated with a loss of susceptibility to TPV (13). Moreover, two scores have recently weighted the virological impact of distinct TPV resistance-associated mutations (19, 24). These scores incorporate mutations associated with resistance to TPV as well as those associated with hypersusceptibility to TPV, allowing the more accurate interpretation of the genotypic changes that influence the antiviral activity of TPV.

Resistance to ritonavir-boosted PIs generally develops gradually and requires the accumulation of several mutations within the HIV protease gene. Partial PI activity can still be recognized when enough drug exposure exists and when viruses harboring only a few protease-associated resistance mutations are present. In this regard, previous studies have shown a good correlation between the TPV plasma trough concentration (Ctrough) and the viral response in the subset of patients infected with viruses carrying an intermediate number of TPV resistance-associated mutations (20, 23). When no mutations are present or when there are too many, the extent of TPV exposure is less relevant. Altogether, these data suggest that the integration of information on the values of the pharmacokinetic parameters and drug resistance changes could improve the management of most antiretroviral-experienced HIV-infected patients treated with TPV (1, 22, 26).

The inhibitory quotient (IQ) is a pharmacological parameter that combines a measure of exposure and susceptibility to a particular drug. The IQ (or phenotypic IQ) is calculated as the Ctrough divided by the serum-adjusted 50% inhibitory concentration (IC50). Alternative quotients have been developed and include the virtual IQ (vIQ), the normalized IQ (nIQ), and the genotypic IQ (gIQ). The vIQ is the ratio between the Ctrough and the virtual phenotype (the expected change in viral susceptibility associated with a particular genotypic mutation profile) multiplied by the serum-adjusted IC50 for wild-type virus. The nIQ is the ratio between a patient IQ and a population reference IQ (the quotient between the average drug Ctrough and a threshold defining resistance for the virtual phenotype in a population). Finally, the gIQ is defined as Ctrough divided by the number of mutations associated with resistance (or weighted score) to the drug.

The gIQ may show several advantages over the other IQs, as it overcomes the difficulties derived from the need to obtain phenotypic results. Moreover, the gIQ has already shown a good ability to predict the virological response to several PIs, including lopinavir, amprenavir, atazanavir, and TPV (3, 8, 10, 11, 15-18, 23). TPV has a phenotypic IQ of 76 and a gIQ of 4.7 μg/ml/mutation, which have been defined to predict a virological response (15). However, no data are so far available about the impact of vIQ on the virological response to TPV and which single mutations or weighted genotypic scores better predict the response to TPV by the use of gIQs. The lack of appreciation of the different impacts of distinct resistance-associated changes may explain the relatively poor accuracy of algorithms for determination of genotypic resistance and the discrepancies between these different algorithms in predicting the response to TPV. Here we report the results of a retrospective study in which the predictive value of the virological response to TPV was assessed by using the vIQ and three different gIQs, two of which incorporated weighted scores.

MATERIALS AND METHODS

Study design and patient population.

A retrospective analysis of all HIV-infected individuals who had initiated therapy with TPV at 500 mg twice a day plus ritonavir at 200 mg twice a day at a referral HIV clinic in Madrid, Spain, between December 2005 and May 2008 was performed. All individuals were antiretroviral experienced and had failed other PI treatment regimens. The main demographics and relevant clinical features at the baseline were recorded in a case report form specially designed for this study. The plasma viral load and CD4 counts were available at the baseline and at each visit during follow-up. An early virological response was defined when the plasma HIV RNA load declined by ≥1 log unit or to <50 copies/ml at the time of the first visit between weeks 4 and 12 of TPV therapy.

TPV plasma exposure.

TPV plasma Ctroughs were measured at the time of the first visit between weeks 4 and 12 by a modified validated high-performance liquid chromatography method, which has already been described elsewhere (9); TPV Ctroughs ≥20.5 μg/ml were considered therapeutic (15).

Genotypic resistance testing.

At the baseline, plasma HIV RNA was evaluated for protease resistance-associated mutations by using a ViroSeq HIV-1 genotyping system (version 2.7; Abbott, Chicago, IL). Three different interpretation systems were then tested. When the mutation list of the International AIDS Society—USA (IAS-USA) panel was considered, viruses were considered TPV susceptible when four mutations or less were present, TPV intermediate resistance was defined for viruses harboring between five and seven mutations, and full TPV resistance was considered for viruses with eight or more mutations (13). The second genotypic interpretation system was derived from the Randomized Evaluation of Strategic Intervention in Multidrug-Resistant Patients with Tipranavir (RESIST) trial and incorporated a weighted score for distinct mutations (24). Viruses were considered susceptible, intermediate, or fully resistant when the total scores were ≤3, between 4 and 10, and ≥11, respectively. Finally, the genotypic interpretation system recently developed by the French Agence Nationale de Recherches sur le Sida et les Hépatites Virales (ANRS) was tested. In this algorithm, changes associated with protease resistance are weighted according to their impact on the virological response to TPV obtained with a distinct clinical data set (19).

Virtual phenotype.

The virtual change in the IC50 for wild-type virus was obtained from Virco (Mechelen, Belgium). Briefly, all specimens with genotypic sequences and paired phenotypic susceptibility results for TPV were used to infer an estimate of the phenotypic susceptibility to TPV for a virus with a given protease genotypic pattern.

Genotypic sensitivity score.

The genotypic sensitivity score was calculated as the total number of drugs in the optimized background regimen to which a viral isolate was presumed to show genotypic susceptibility according to the Stanford criteria (16). In this system, antiretroviral drugs are considered to display full, intermediate, or null antiviral activity, which are scored as 1, 0.5, and 0, respectively.

IQs.

A total of four distinct IQs for TPV, the vIQ and three gIQs, were calculated on the basis of the different TPV resistance-associated mutation scores. Table Table11 provides the definitions and the means of calculation of the IQs. Given that some patients had no TPV resistance-associated mutations at the baseline or changes that produce hypersusceptibility, a correction in the final mutation score was performed to avoid negative denominators or values equal to zero. By use of the scores of the IAS-USA panel, a total of 6 patients (11%) had no TPV resistance-associated mutations. A total of 17 (30%) and 32 (56%) patients had negative scores or scores equal to zero when the RESIST and the ANRS weighted scores, respectively, were considered. The correction consisted of the addition of the following values to the denominator: 1 for the IAS-USA gIQ, 15 for the RESIST gIQ, and 2 for the ANRS gIQ.

TABLE 1.
Definition of distinct TPV IQs

Statistical analyses.

Qualitative variables were expressed as frequencies, and quantitative parameters were expressed as medians and interquartile ranges (IQR). The Mann-Whitney U test and Fisher's exact test were used to compare quantitative and categorical variables, respectively, in the virological responders and nonresponders. Receiver operating characteristic curve analyses were used to explore the best thresholds for predicting an early virological response to TPV. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), likelihood ratio (LHR) for a positive test result (LHR+), and likelihood ratio for a negative test result (LHR) were used to compare the accuracies of the different parameters for prediction of the virological response to TPV. Conceptually, LHR is the ratio of two probabilities, namely, the probability that a specific test result will be obtained for patients with a virological response divided by the probability that the same test result will be obtained for patients without a virological response. Therefore, LHR+ is defined as the probability that a positive test result will be obtained for patients with a virological response divided by the probability that a positive test result will be obtained for patients without a virological response and is calculated as the ratio sensitivity/(1 − specificity). Accordingly, LHR is defined as the probability that a negative test result will be obtained for patients with a virological response divided by the probability that a negative test result will be obtained for patients without a virological response and is calculated as the ratio (1 − sensitivity)/specificity. The best predictors were those with the highest LHR+ and the lowest LHR. Variables showing P values of <0.05 were considered significant. All statistical analyses were performed with the SPSS package (version 13.0; SPSS Inc., Chicago, IL).

RESULTS

Study population.

A total of 57 patients who initiated salvage therapy with a TPV-containing regimen were analyzed. Table Table22 summarizes the main baseline characteristics of the study population. The optimized antiretroviral background regimen included more than two drugs in 12% of the patients, between two drugs and one drug in 63% of the patients, and less than one drug in 25% of patients. By using the IAS-USA resistance interpretation system, full TPV resistance was predicted for 7% of the patients, intermediate TPV resistance was predicted for 16%, and TPV susceptibility was considered for 77%. By using the TPV weighted resistance mutation score from Scherer et al. (24), 5% of the patients were considered to be fully resistant to TPV, 32% were considered to be intermediate resistant, and 63% were considered to be susceptible. Finally, by using the TPV weighted resistance mutation score of Marcelin et al. (19), the percentages of patients with 3, 2, 1, 0, and −1 weighted mutations were 2%, 12%, 30%, 47%, and 9%, respectively.

TABLE 2.
Baseline characteristics of the study population

Early virological response.

Following the initiation of TPV-based salvage regimens, the median decline in the plasma HIV RNA load was 1.8 log copies/ml (IQR, 0.84 to 2.4 log copies/ml) between weeks 4 and 12. Overall, a virological response was achieved by 44 (77%) patients, and the viral load was undetectable (<50 HIV RNA copies/ml) in 25 (44%) patients.

The differences between the responders and the nonresponders are depicted in Table Table3.3. The values of TPV Ctrough and the scores of the IAS-USA and RESIST genotypic resistance algorithms, vIQ, and all three gIQs (the IAS-USA, RESIST, and ANRS gIQs) were significantly higher for responders than for nonresponders. In contrast, the ANRS genotypic resistance algorithm score and the genotypic sensitivity score did not significantly distinguish the responders from the nonresponders. It is interesting that when the values from the distinct TPV genotypic resistance algorithms were grouped into three categories (susceptible, intermediate, and fully resistant), significant differences between the responders and the nonresponders were seen only for the IAS-USA and RESIST algorithms and not for the ANRS system. By use of the IAS-USA algorithm, the proportions of virological responders were 82%, 78%, and 25% for patients with scores of less than or equal to four, five to seven, and greater than or equal to eight mutations, respectively (P = 0.035). By using the RESIST algorithm, these figures were 75%, 94%, and 0% for scores of ≤3, 4 to 10, and ≥11, respectively (P = 0.002).

TABLE 3.
Differences between virological responders and nonresponders to TPV salvage therapy

Predictors of virological response.

Analysis by use of the receiver operating characteristic curves defined the thresholds for the TPV IQs. A vIQ threshold of 6 μg/ml/unit fold change (area under the concentration-time curve [AUC] = 0.85, P = 0.002) showed 94% sensitivity and 88% specificity for detection of a virological response to TPV. A threshold of 4.6 μg/ml/mutation in the IAS-USA gIQ (AUC = 0.81, P = 0.001) had a 75% sensitivity and a 77% specificity. By using the RESIST gIQ, a threshold of 1.1 μg/ml/weighted mutation (AUC = 0.79, P = 0.002) had a 79% sensitivity and a 77% specificity. Finally, a threshold of 8.3 μg/ml/weighted mutation by use of the ANRS gIQ (AUC = 0.74, P = 0.01) had a 71% sensitivity and a 77% specificity (Table (Table44).

TABLE 4.
Sensitivity, specificity, PPV, NPV, LHR+, and LHR for virological response to TPV salvage therapy by use of distinct pharmacological and genotypic parameters

The accuracies of the distinct pharmacological and virological parameters for predicting an early virological response were compared. vIQ was the one with the best PPV, NPV, LHR+, and LHR. Table Table44 displays the PPV, NPV, LHR+, and LHR for the different pharmacological and genotypic parameters.

DISCUSSION

Early changes in viral load (between weeks 4 and 12) are the most sensitive for recognition of the activities of any antiretroviral agent used for rescue therapeutic interventions. Since antiviral activity in the clinical setting is a composite of drug exposure and viral susceptibility, the recognition of early predictors of viral response by the use of pharmacological and genotypic parameters may be helpful for choosing the best antiretroviral regimen for a given HIV-infected individual (4).

In this study, the baseline TPV resistance-associated mutations, TPV Ctrough, and IQs were all found to be significantly associated with a virological response to TPV when it is used in salvage regimens. However, the ability of these parameters to predict a virological response differed substantially, as reflected by their accuracies when they were expressed as the PPV, NPV, LHR+, and LHR. It should be noted, however, that an important limitation of using predictive values to compare the accuracies of different tests is that the rate of the outcome tested is a major determinant of the results (25). In our study, the virological response rate was high (77%), and accordingly, the PPV might overestimate the accuracies of the parameters tested, and, conversely, the NPV might underestimate the accuracies of the parameters tested. In order to overcome this problem, LHRs were estimated since they measure test accuracy independently of the rate of the outcome (5, 6, 14). In this way, we found that the best parameter predicting a virological response to TPV was the vIQ; it had the highest LHR+ (LHR+ = 7.8) and the lowest LHR (LHR = 0.07). That means that a vIQ of ≥6 μg/ml/unit fold change should be seen approximately eight times more often in patients experiencing a virological response than in nonresponders; conversely, a vIQ of <6 μg/ml/unit fold change should be recognized approximately 0.07 times more often in patients experiencing a virological response than in nonresponders.

It is noteworthy that in our study the performance of the distinct gIQs did not prove to be much more superior to the performance of some genotypic resistance mutation algorithms, which suggests that differences in the levels of TPV exposure have a limited impact on the virological response, at least for the population of treatment-experienced patients evaluated in the present study. While gIQs had better LHR+ values than resistance mutation scores, the latter had better LHR values than the former. From a clinical standpoint, this observation highlights the importance of particularly close monitoring of the subset of patients with a larger number of TPV resistance mutations, in whom the chances of treatment failure are higher and in whom the treatment failure will only rarely be overcome by an increased TPV Ctrough. In this regard, the better specificities of the gIQs over resistance mutation algorithms would support the use of the gIQs in patients with an intermediate number of TPV resistance mutations (20).

No major advantage of weighting over listing of the TPV resistance-associated mutations was seen in our study. There were no significant differences between the IAS-USA interpretation algorithm (13) and the most sophisticated weighted score system derived from the RESIST trial (24) for prediction of an early virological response to TPV. Since the weighted score developed by Scherer et al. (24) was based on the results of the RESIST trials (2, 7, 12), which required that the patients recruited had to be infected with viruses with a minimum number of specific primary PI resistance-associated mutations and which excluded other PI-experienced populations. This bias most likely had a negative impact on the performance of the weighting score for a broader PI-experienced population, such as the one tested in our study. Clearly, more accurate weighted scores for TPV derived from different clinical data sets must be established. The weighted algorithm proposed by Marcelin et al. (19) underperformed compared to the performance characteristics of the other resistance interpretation systems in our series, and its use cannot be recommended.

The median TPV Ctrough was considerably lower in nonresponders than in patients with a virological response. Differences in adherence to treatment could explain the low values found in nonresponders. Although adherence data were not available in this study, the complexity of salvage therapies with TPV-ritonavir due to the high pill burden, together with meal restrictions, could impede good adherence. Therefore, in this study the nonresponders could have had worse adherence than the responders. Other factors could involve drug interactions, since TPV-ritonavir is a substrate of cytochrome P-450 and P-glycoprotein. The drugs taken concomitantly were not recorded in this study, so this issue could not be evaluated. Finally, genetic factors affecting the activities of cytochrome P-450 and/or P-glycoprotein may also contribute to low TPV concentrations.

Several questions regarding the applicability of IQs and specifically the applicability of gIQs must be answered before they are used in the clinic. The first is when is the best time to measure drug levels following the initiation of therapy? Although the earliest changes in viral load may not always determine the long-term virological response, they directly reflect the antiviral activity of a drug, which may be sustained, depending on other factors, such as the genetic barrier to resistance. Accordingly, the sooner that we may predict the virological outcome, the earlier that adjustments in drug dosing may be attempted. Therefore, we propose that drug levels be monitored between weeks 4 and 12 for the calculation of the gIQs. However, the dose-limiting side effects of drugs (and TPV is no exception) may hamper the clinical usefulness of this strategy. Another question regards the consideration of hypersusceptibility in weighting resistance mutation scores. How will gIQ be calculated for patients with no specific resistance mutations for a given drug or changes causing hypersusceptibility? While some studies assigned a gIQ value equal to the Ctrough to subjects with no resistance mutations (17), others excluded them from further analyses (3). In our study, we proposed the addition of a specific value to all denominators to avoid negative values or values equal to zero. This strategy allowed the inclusion of all patients and permitted discrimination between patients with just one resistance-associated mutation and those with no resistance-associated mutations, who were a relatively large population in our study.

In summary, both resistance and pharmacokinetic parameters predict the virological response to TPV in antiretroviral-experienced patients, with the vIQ being the most accurate parameter. This information may be helpful in allowing early intervention strategies to maximize success, such as adding a new active drug to the regimen or attempting dose adjustments, in order to either maximize drug exposure in the face of very resistant viruses or, conversely, safely pursue a reduction in plasma TPV levels in order to minimize potential dose-dependent side effects, such as liver toxicity (21).

Acknowledgments

This work was supported in part by grants from the Fundación Investigación y Educación en SIDA (IES), the Red de Investigación en SIDA (grant ISCIII-FIS-Retic-RD06/006), the Fondo de Investigación Sanitaria (grants FIS-CP07/00016 and CP07/0567), and the NEAT European project.

We have no conflicts of interest to declare.

Footnotes

[down-pointing small open triangle]Published ahead of print on 13 July 2009.

REFERENCES

1. Baxter, J. D., T. C. Merigan, D. N. Wentworth, J. D. Neaton, M. L. Hoover, R. M. Hoetelmans, S. C. Piscitelli, W. H. Verbiest, and D. L. Mayers. 2002. Both baseline HIV-1 drug resistance and antiretroviral drug levels are associated with short-term virologic responses to salvage therapy. AIDS 16:1131-1138. [PubMed]
2. Cahn, P., J. Villacian, A. Lazzarin, C. Katlama, B. Grinsztejn, K. Arasteh, P. Lopez, N. Clumeck, J. Gerstoft, N. Stavrianeas, S. Moreno, F. Antunes, D. Neubacher, and D. Mayers. 2006. Ritonavir-boosted tipranavir demonstrates superior efficacy to ritonavir-boosted protease inhibitors in treatment-experienced HIV-infected patients: 24-week results of the RESIST-2 trial. Clin. Infect. Dis. 43:1347-1356. [PubMed]
3. Delaugerre, C., G. Peytavin, S. Dominguez, A. G. Marcelin, C. Duvivier, K. Gourlain, B. Amellal, M. Legrand, F. Raffi, D. Costagliola, C. Katlama, and V. Calvez. 2005. Virological and pharmacological factors associated with virological response to salvage therapy after an 8-week of treatment interruption in a context of very advanced HIV disease (GigHAART ANRS 097). J. Med. Virol. 77:345-350. [PubMed]
4. de Mendoza, C., J. Morello, P. Garcia-Gasco, S. Rodriguez-Novoa, and V. Soriano. 2007. Tipranavir: a new protease inhibitor for the treatment of antiretroviral-experienced HIV-infected patients. Expert Opin. Pharmacother. 8:839-850. [PubMed]
5. Dujardin, B., J. Van den Ende, A. Van Gompel, J. P. Unger, and P. Van der Stuyft. 1994. Likelihood ratios: a real improvement for clinical decision making? Eur. J. Epidemiol. 10:29-36. [PubMed]
6. Fischer, J. E., L. M. Bachmann, and R. Jaeschke. 2003. A readers' guide to the interpretation of diagnostic test properties: clinical example of sepsis. Intensive Care Med. 29:1043-1051. [PubMed]
7. Gathe, J., D. A. Cooper, C. Farthing, D. Jayaweera, D. Norris, G. Pierone, Jr., C. R. Steinhart, B. Trottier, S. L. Walmsley, C. Workman, G. Mukwaya, V. Kohlbrenner, C. Dohnanyi, S. McCallister, and D. Mayers. 2006. Efficacy of the protease inhibitors tipranavir plus ritonavir in treatment-experienced patients: 24-week analysis from the RESIST-1 trial. Clin. Infect. Dis. 43:1337-1346. [PubMed]
8. Gianotti, N., L. Galli, A. Danise, H. Hasson, E. Boeri, A. Lazzarin, and A. Castagna. 2006. Ability of different lopinavir genotypic inhibitory quotients to predict 48-week virological response in highly treatment-experienced HIV-infected patients receiving lopinavir/ritonavir. J. Med. Virol. 78:1537-1541. [PubMed]
9. Giraud, E., E. Rey, J. M. Treluyer, G. Pons, and V. Jullien. 2006. Quantification of tipranavir in human plasma by high-performance liquid chromatography with UV detection. J. Chromatogr. B. 830:86-90. [PubMed]
10. González de Requena, D., A. Calcagno, M. G. Milia, A. D'Avolio, M. Sciandra, S. Garazzino, M. Siccardi, A. Sinicco, and G. Di Perri. 2006. Abstr. 13th Conf. Retrovir. Opportunistic Infect., abstr. 577.
11. González de Requena, D., S. Bonora, A. Calcagno, A. D'Avolio, M. Siccardi, S. Fontana, M. G. Milia, M. Sciandra, S. Garazzino, A. Di Garbo, L. Baietto, L. Trentini, and G. Di Perri. 2008. Tipranavir (TPV) genotypic inhibitory quotient predicts virological response at 48 weeks to TPV-based salvage regimens. Antimicrob. Agents Chemother. 52:1066-1071. [PMC free article] [PubMed]
12. Hicks, C. B., P. Cahn, D. A. Cooper, S. L. Walmsley, C. Katlama, B. Clotet, A. Lazzarin, M. A. Johnson, D. Neubacher, D. Mayers, and H. Valdez. 2006. Durable efficacy of tipranavir-ritonavir in combination with an optimised background regimen of antiretroviral drugs for treatment-experienced HIV-1-infected patients at 48 weeks in the Randomized Evaluation of Strategic Intervention in multidrug reSistant patients with Tipranavir (RESIST) studies: an analysis of combined data from two randomised open-label trials. Lancet 368:466-475. [PubMed]
13. Hirsch, M. S., H. F. Gunthard, J. M. Schapiro, F. Brun-Vezinet, B. Clotet, S. M. Hammer, V. A. Johnson, D. R. Kuritzkes, J. W. Mellors, D. Pillay, P. G. Yeni, D. M. Jacobsen, and D. D. Richman. 2008. Antiretroviral drug resistance testing in adult HIV-1 infection: 2008 recommendations of an International AIDS Society—USA panel. Clin. Infect. Dis. 47:266-285. [PubMed]
14. Jaeschke, R., G. Guyatt, D. L. Sackett, et al. 1994. Users' guides to the medical literature. III. How to use an article about a diagnostic test. A. Are the results of the study valid? JAMA 271:389-391. [PubMed]
15. La Porte, C. J. L., D. J. Back, T. Blaschke, C. A. B. Boucher, C. V. Fletcher, and J. G. Flexner. 2006. Updated guideline to perform therapeutic drug monitoring for antiretroviral agents. Rev. Antivir. Ther. 3:3-14.
16. Maillard, A., J. M. Chapplain, O. Tribut, D. Bentue-Ferrer, P. Tattevin, C. Arvieux, C. Michelet, and A. Ruffault. 2007. The use of drug resistance algorithms and genotypic inhibitory quotient in prediction of lopinavir-ritonavir treatment response in human immunodeficiency virus type 1 protease inhibitor-experienced patients. J. Clin. Virol. 38:131-138. [PubMed]
17. Marcelin, A. G., I. Cohen-Codar, M. S. King, P. Colson, E. Guillevic, D. Descamps, C. Lamotte, V. Schneider, J. Ritter, M. Segondy, H. Peigue-Lafeuille, L. Morand-Joubert, A. Schmuck, A. Ruffault, P. Palmer, M. L. Chaix, V. Mackiewicz, V. Brodard, J. Izopet, J. Cottalorda, E. Kohli, J. P. Chauvin, D. J. Kempf, G. Peytavin, and V. Calvez. 2005. Virological and pharmacological parameters predicting the response to lopinavir-ritonavir in heavily protease inhibitor-experienced patients. Antimicrob. Agents Chemother. 49:1720-1726. [PMC free article] [PubMed]
18. Marcelin, A. G., C. Lamotte, C. Delaugerre, N. Ktorza, M. H. Ait, R. Cacace, M. Bonmarchand, M. Wirden, A. Simon, P. Bossi, F. Bricaire, D. Costagliola, C. Katlama, G. Peytavin, V. Calvez, and the Genophar Study Group. 2003. Genotypic inhibitory quotient as predictor of virological response to ritonavir-amprenavir in human immunodeficiency virus type 1 protease inhibitor-experienced patients. Antimicrob. Agents Chemother. 47:594-600. [PMC free article] [PubMed]
19. Marcelin, A. G., B. Masquelier, D. Descamps, J. Izopet, C. Charpentier, C. Alloui, M. Bouvier-Alias, A. Signori-Schmuck, B. Montes, M. L. Chaix, C. Amiel, G. D. Santos, A. Ruffault, F. Barin, G. Peytavin, M. Lavignon, P. Flandre, and V. Calvez. 2008. Tipranavir-ritonavir genotypic resistance score in protease inhibitor-experienced patients. Antimicrob. Agents Chemother. 52:3237-3243. [PMC free article] [PubMed]
20. Morello, J., P. P. García-Gasco, S. Rodriguez-Novoa, C. de Mendoza, F. Blanco, G. Gonzalez-Pardo, B. Sanz, I. Jimenez-Nacher, and V. Soriano. 2008. Association between tipranavir plasma levels and virological response in HIV-infected patients. AIDS Res. Hum. Retrovir. 24:389-391. [PubMed]
21. Morello, J., S. Rodríguez-Nóvoa, F. Blanco, I. Jiménez-Nácher, G. González-Pardo, A. J. Rubio, and V. Soriano. 2008. Abstr. 9th Int. Congr. Drug Ther. HIV Infect., abstr. P66.
22. Morse, G. D., L. M. Catanzaro, and E. P. Acosta. 2006. Clinical pharmacodynamics of HIV-1 protease inhibitors: use of inhibitory quotients to optimise pharmacotherapy. Lancet Infect. Dis. 6:215-225. [PubMed]
23. Naeger, L. K., J. J. Zheng, and K. A. Struble. 2006. Abstr. 13th Conf. Retrovir. Opportunistic Infect., abstr. 639a.
24. Scherer, J., C. Boucher, J. Baxter, J. Schapiro, V. Kohlbrenner, and D. Hall. 2008. Abstr. 6th Eur. HIV Drug Resist. Workshop, abstr. 94.
25. Smith, J. E., R. L. Winkler, and D. G. Fryback. 2000. The first positive: computing positive predictive value at the extremes. Ann. Intern. Med. 132:804-809. [PubMed]
26. Winston, A., G. Hales, J. Amin, E. van Schaick, D. A. Cooper, and S. Emery. 2005. The normalized inhibitory quotient of boosted protease inhibitors is predictive of viral load response in treatment-experienced HIV-1-infected individuals. AIDS 19:1393-1399. [PubMed]

Articles from Antimicrobial Agents and Chemotherapy are provided here courtesy of American Society for Microbiology (ASM)