Despite the success of antiretroviral therapy in treating HIV-1 infection [1–3], significant numbers of patients experience virologic failure [4,5]. Although not all viruses recovered from patients failing protease inhibitor (PI)-based regimens exhibit decreased susceptibility to PI [6,7], substantial numbers do so, especially after failing multiple PI-based regimens [5,8–12]. Variable responses to subsequent single and dual PI-based therapy [8,13–18] may partly be due to increasing numbers of protease amino acid substitutions associated with a more pronounced and broader loss of PI susceptibility [8,9,19–21]. Plasma drug levels achieved with individual PI may not be adequate to overcome this loss of PI susceptibility .
However, ritonavir (RTV) is known to substantially increase plasma levels of other co-administered PI through inhibition of the cytochrome P450 system [8,22–27], and recent studies suggest that indinavir (IDV) exposures achieved with co-administration of IDV 800 mg with RTV 200 mg twice daily  may be high enough to suppress the replication of viruses resistant to IDV and to other PI .
Here, we analyze the associations between response to IDV–RTV-based regimens, reported adherence to therapy, and baseline genotypic and phenotypic resistance among 28 individuals with prior virologic failures of multiple PI-containing regimens.
This study was conducted at Jackson Memorial Hospital in Miami, Florida, after approval by the Office for the Protection of Human Subjects of the University of Miami School of Medicine. Patients with prior failure [viral RNA (vRNA) rebound to ≥ 1000 copies/ml after suppression to < 400 copies/ml or failure to achieve vRNA < 400 copies/ml after 6 months on treatment] of one or more combination regimens containing one or more PI who had subsequently received IDV (800 mg twice daily) and RTV (200 mg twice daily) in combination with one or more nucleoside analogue reverse transcriptase inhibitor (NRTI) and/ or a non-nucleoside reverse transcriptase inhibitor (NNRTI) were identified and their charts reviewed for pertinent demographic, clinical, and laboratory data.
Responders were defined as individuals achieving nadir vRNA ≤ 400 copies/ml at least once within 6 months of starting IDV–RTV-based regimens. Non-responders were defined as individuals with vRNA > 400 copies/ml for all measurements within 6 months of starting IDV–RTV-based regimens.
Adherence to therapy over the preceding week was routinely assessed through patient self-report at each clinic visit. Adequate adherence was defined as taking ≥ 85% of the doses of the IDV–RTV-based regimen at the latest time point for which information was available for each patient.
Population-based sequencing of the protease and reverse transcriptase genes  was done on the plasma sample obtained closest to initiation of the IDV–RTV-based regimen while still on a PI-containing regimen. Genotypic resistance to IDV was defined as the presence of substitutions at protease positions 82, 84, and/or 90, plus one or more of the other substitutions implicated in IDV resistance [9,19]. Initially, phenotypic resistance [≥ 2.5-fold increase in the 95% inhibitory concentration (IC95) of the tested virus relative to wild-type HIV-1] was determined with the PhenoSense assay (ViroLogic, Inc., South San Francisco, California, USA)  on all viral samples regardless of genotypic changes. Subsequently, however, phenotypic testing was performed only on those samples with genotypic resistance as it became evident that genotypically susceptible variants were invariably phenotypically susceptible.
The Mann–Whitney test was used to compare differences between the means of independent samples, Wilcoxon's test was used to compare differences between the means of paired samples, and Fisher's exact test was used to compare associations in 2 × 2 tables. A two-tailed P value ≤ 0.05 was considered significant.
Sixteen responders and 12 non-responders were identified with similar sex and age distributions and similar mean ± SEM CD4 cell counts (268 ± 46 × 106 and 150 ± 41 × 106 cells/l; P = 0.1) and vRNA values (165 810 ± 68 210 and 228 231 ± 75 839 copies/ml; P = 0.1), respectively, prior to initiation of the IDV–RTV-based regimens.
Both responders and non-responders had extensive, yet similar, prior treatment histories. All 28 patients had previously received either IDV or RTV. Among responders, 14 out of 16 had received IDV and 9 out of 16 had received RTV. Among non-responders, 12 out of 12 had received IDV and 6 out of 12 had received RTV. For both responders and non-responders, the mean (± SEM) number of antiretroviral agents received [6.3 (± 0.5) versus 7 (± 0.3); P = 0.3], number of HAART regimens used [2.8 (± 0.3) versus 2.8 (± 0.2); P = 0.6], duration (weeks) of all prior antiretroviral therapies [131 (± 24) versus 117 (± 14); P = 0.7], number of PI received [2.5 (± 0.3) versus 2.3 (± 0.2); P = 0.8], number of PI regimens used [2.4 (± 0.3) versus 2.3 (± 0.2); P = 0.8], and duration (weeks) of PI therapy [97 (± 12) versus 86 (± 10); P = 0.6], respectively, were all similar.
The numbers of IDV resistance-associated substitutions were similar among responders and non-responders (P = 0.09), and there were no differences with regard to the presence of genotypic and/or phenotypic resistance to NRTI or NNRTI (Table 1). Among six patients receiving two or more new NRTI and/or NNRTI, five responded and one did not (P = 0.2); among nine NNRTI-naive patients receiving an NNRTI, six responded and three did not (P = 0.5).
Excluding one responder lost to follow-up after week 16, mean follow-up of the other 27 patients was 69 ± 5.6 weeks (Table 1). Of the remaining 15 responders, 10 had vRNA ≤ 400 copies/ml at their last visit (nine had < 50 copies/ml) and a higher mean CD4 cell count (449 ± 101 × 106 cells/l) than that at baseline (291 ± 65 × 106 cells/l) (P = 0.02). Seven responders had transient increases in vRNA to > 400 copies/ml associated with inadequate adherence, but all vRNA values had returned to ≤ 400 copies/ml (six to < 50 copies/ml) by the latest follow-up.
Five other initial responders subsequently had sustained virologic rebounds to > 400 copies/ml, also associated with non-adherence. Nonetheless, there was a non-significant trend for the last observed mean vRNA (28 987 ± 24 902 copies/ml) to be lower than at baseline (150 996 ± 127 592 copies/ml) (P = 0.08). The last observed mean CD4 cell count (294 ± 82 × 106 cells/l) was higher than at baseline (216 ± 70 × 106 cells/l) (P = 0.04).
Among non-responders, mean vRNA (171 667 ± 43 006 copies/ml) and mean CD4 cell count (152 ± 36 × 106cells/l) were similar to those at initiation of IDV–RTV-based therapy (P = 0.6 and P = 1.0, respectively).
Relationships between adherence, response to therapy within the first 6 months, and baseline genotypic and phenotypic IDV resistance are shown in Table 2. Among 27 patients followed beyond 6 months of initiating therapy, we observed similar findings. Of 16 adherent patients, 10 (63%) showed favorable responses, whereas none of 11 inadequately adherent patients did so (P = 0.001). Strikingly, genotypic and/or phenotypic resistance to IDV did not adversely affect long-term response to therapy, and, as with shorter-term responses, there was a non-significant trend for both types of resistance to be associated with favorable virologic outcomes rather than with therapeutic failure.
Decreased susceptibility to PI selected during prior therapy decreases the efficacy of subsequent PI-containing regimens [13,15,17,30–36]. Resistance testing prior to initiating a new regimen can lead to improved virologic outcomes of subsequent therapies [37–40]. However, interpretation of drug susceptibility data requires simultaneous consideration of drug potency and in vivo exposure . For some agents, particularly those with low potencies and/or those that achieve low drug exposures, a small loss of susceptibility may preclude successful salvage if free drug levels fail to achieve an inhibitory concentration. For all individual PI in current use, trough (Cmin) protein binding-adjusted drug exposures are calculated to be near (and in some cases, below) the wild-type IC95, such that a less than fourfold reduction in drug susceptibility would raise the inhibitory concentration above the adjusted Cmin .
However, co-administration of RTV can significantly increase exposures of saquinavir (SQV), amprenavir (APV), lopinavir (LPV), and IDV [22–27]. For IDV–RTV, LPV–RTV, and APV–RTV, these increased drug exposures are predicted to exceed inhibitory concentrations of many ‘resistant’ viruses, permitting their virologic control [8,20,21].
In this study, IDV ‘resistance’ was not associated with diminished virologic responses. In normal volunteers receiving the same dosage of IDV–RTV, trough exposures of IDV were nearly 80 times the protein binding-adjusted IC95 for wild-type HIV-1 [20,26], substantially exceeding the phenotypic shifts seen here. Therefore, our patients’ responses are consistent with previous predictions .
Not surprisingly, adherence to therapy was strongly associated with favorable responses. However, although not statistically significant, the trend toward more favorable responses among patients with resistant viruses was surprising and suggested a confounding effect of non-virologic factors in the clinical responses.
The association between baseline resistance (selected during prior therapy) and adequate adherence to subsequent therapy provides a plausible explanation for this otherwise counter-intuitive finding. Baseline resistance implies that prior adherence to therapy had been adequate to select it; conversely, no resistance would be expected among patients with very poor drug exposures (i.e., poor adherence). Most patients with IDV-resistant viruses (and presumably partially adherent to prior PI therapies) were ‘adequately’ adherent to subsequent salvage therapy. Because of the strong relationship between adherence and response, these patients derived the greatest benefit from IDV–RTV-based therapy despite pre-existing IDV resistance. Thus, IDV resistance was a strong positive predictor of adherence but was not associated with virologic failure.
Although the definition of virologic ‘response’ used here (nadir vRNA < 400 copies/ml at least once during the first 6 months) could have been made more stringent with a longer period of follow-up, the chosen endpoint was most appropriate to address this fundamentally virologic question: Can PI suppress viral replication in patients harboring viruses resistant to them? Because adherence to therapy was more strongly associated with response than was baseline resistance, an immediate response is perhaps the most relevant one because the longer the follow-up, the greater the chance that therapeutic efficacy (i.e., virologic suppression) will be confounded by non-adherence. Having achieved initial suppression, the most critical issue then becomes the persistence of adherence adequate to maintain long-term suppression.
Several limitations of this study deserve mention. The collection of adherence information by patient self-reporting may be less reliable than through pill counts or electronic monitoring. However, the observed agreement between our measures of adherence and response to therapy is consistent with previous observations showing similar associations between adherence and viral suppression .
The number of patients in this report is relatively small, and some associations suggested by the data may not have achieved statistical significance due to small sample sizes. Moreover, our exploration of the interaction between adherence and measures of susceptibility was not pre-specified. It is known that such analyses may lead to an inflated rate of false positives. In order to confirm the putative associations we observed and more fully investigate the relationships between adherence, susceptibility, and response, a larger study with pre-specified hypotheses is needed.
Finally, the observed responses were likely influenced by concomitant medications. The use of agents, particularly NNRTI, to which patients are naive, is associated with better outcomes among patients failing PI-based therapies . Although non-significant, there were trends that suggested that patients receiving two or more new agents, or those receiving an NNRTI for the first time, were more likely to respond than patients who did not. These agents may have contributed to the efficacy of IDV–RTV-based regimens, but numbers of patients in both categories are too small to draw definitive conclusions.
Although the antiviral activity of combination IDV–RTV therapy is believed to be driven primarily by IDV exposure , the measurement of plasma IDV levels in our patient plasma samples was judged unlikely to yield interpretable data, for several reasons. It would have been impossible to know the exact time at which patients took the last dose relative to the time when blood samples were collected. Even if the last dose had been witnessed, there can be no certainty that the patients’ IDV concentration would have been at steady-state or whether that level was representative of the blood level over the preceding (and following) days to weeks of therapy. Therefore, blood drug levels obtained in this way can be misleading and are not appropriate surrogates of long-term drug exposure or adherence .
The strong influence of adherence on virologic responses suggests that resistance test results should be interpreted with caution. When virologic failure is due to poor adherence, measures to improve adherence may be more appropriate than changes in therapy. Further, a finding of genotypic or phenotypic ‘resistance’ to a given drug may not rule out future use of that drug if its concentration can be increased sufficiently in vivo. Our data are consistent with previous findings [20,21] that drug resistance can be overcome if sufficient antiviral potency and drug exposure can be achieved.
We are grateful to K. Holmes, Merck Research Laboratories, for logistical assistance and C. Petropoulos and R. Ziermann, ViroLogic, Inc., for the performance of genotypic and phenotypic assays.
Sponsorship: REC is a consultant and investigator for Merck & Co., Inc., and DJH, MS, DMD, JLW, KH, WAS, EAE, and JHC are employees of Merck & Co., Inc.
1. Palella FJ, Jr., Delaney KM, Moorman AC, Loveless MO, Fuhrer J, Satten GA, et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. N Engl J Med
2. Hogg RS, Heath KV, Yip B, Craib K, O'Shaughnessy MV, Schechter MT, et al. Improved survival among HIV-infected individuals following initiation of antiretroviral therapy. JAMA
3. Hogg RS, O'Shaughnessy MV, Gataric N, Yip B, Craib KJP, Schechter MT, et al. Decline in deaths from AIDS due to new antiretrovirals. Lancet
4. Lucas GM, Chaisson RE, Moore RD. Highly active antiretroviral therapy in a large urban clinic: risk factors for virologic failure and adverse drug reactions. Ann Intern Med
5. Hirsch MS, Brun-Vezinet F, D'Aquila RT, Hammer SM, Johnson VA, Kuritzkes DR, et al. Antiretroviral drug resistance testing in adult HIV-1 infection: recommendations of an International AIDS Society-USA Panel. JAMA
6. Descamps D, Flandre P, Calvez V, Peytavin G, Meiffredy V, Collin G, et al. Mechanisms of virological failure in previously untreated HIV-infected patients from a trial of induction-maintenance therapy. JAMA
7. Havlir DV, Hellman NS, Petropoulos CJ, Whitcomb JM, Collier AC, Hirsch MS, et al. Drug susceptibility in HIV infection after viral rebound in patients receiving indinavir-containing regimens. JAMA
8. Duval X, Lamotte C, Race E, Descamps D, Damond F, Clavel F, et al. Amprenavir inhibitory quotient and virological response in human immunodeficiency virus-infected patients on an amprenavir-containing salvage regimen without or with ritonavir. Antimicrob Agents Chemother
9. Condra JH, Gabryelski LJ, Graham DJ, Quintero JC, Schlief WA, Blahy OM, et al. In vivo emergence of HIV-1 variants resistant to multiple protease inhibitors. Nature
10. Molla A, Korneyeva M, Gao Q, Vasavanonda S, Schipper PJ, Mo HM, et al. Ordered accumulation of mutations in HIV protease confers resistance to ritonavir. Nature Med
11. Boden D, Markowitz M. Resistance to human immunodeficiency virus type 1 protease inhibitors. Antimicrob Agents Chemother
12. Cabana M, Clotet B, Martinez MA. Emergence and genetic evolution of HIV-1 variants with mutations conferring resistance to multiple reverse transcriptase and protease inhibitors. J Med Virol
13. Casado JL, Dronda F, Hertogs K, Antela A, Moreno S. Failure of a ritonavir plus saquinavir-based rescue regimen precludes the use of protease inhibitors. AIDS
14. Deeks SG, Grant RM, Beatty GW, Hortoan C, Detmer J, Eastman S. Activity of a ritonavir plus saquinavir-containing regimen in patients with virologic evidence of indinavir or ritonavir failure. AIDS
15. Deeks SG, Hellmann NS, Grant RM, Parkin NT, Petropoulos CJ, Becker M, et al. Novel four-drug salvage treatment regimens after failure of a human immunodeficiency virus type 1 protease inhibitor-containing regimen: antiviral activity and correlation of baseline phenotypic drug susceptibility with virologic outcome. J Infect Dis
16. Fätkenheuer G, Hoetelmans RMW, Hunn N, Schwenk A, Franzen C, Reiser M, et al. Salvage therapy with regimens containing ritonavir and saquinavir in extensively pretreated HIV-infected patients. AIDS
17. Piketty C, Race E, Castiel P, Belec L, Peytavin G, Si-Mohamed A, et al. Efficacy of a five-drug combination including ritonavir, saquinavir and efavirenz in patients who failed on a conventional triple-drug regimen: phenotypic resistance to protease inhibitors predicts outcome of therapy. AIDS
18. Harrigan PR, Hertogs K, Verbiest W, Pauwels B, Larder S, Kemp S, et al. Baseline HIV drug resistance profile predicts response to ritonavir-saquinavir protease inhibitor therapy in a community setting. AIDS
19. Condra JH, Holder DJ, Schleif WA, Blahy OM, Danovich RM, Gabryelski LJ, et al. Genetic correlates of in vivo viral resistance to indinavir, a human immunodeficiency virus type 1 protease inhibitor. J Virol
20. Condra JH, Petropoulos CJ, Ziermann R, Schleif WA, Shivaprakash R, Emini EA. Drug resistance and predicted virologic responses to human immunodeficiency virus type 1 protease inhibitor therapy. J Infect Dis
21. Kempf DJ, Isaacson JD, King MS, Brun SC, Xu Y, Real K, et al. Identification of genotypic changes in human immunodeficiency virus protease that correlate with reduced susceptibility to the protease inhibitor lopinavir among viral isolates from protease inhibitor-experienced patients. J Virol
22. Buss N, Snell P, Bock J, Hsu A, Jorga K. Saquinavir and ritonavir pharmacokinetics following combined ritonavir and saquinavir (soft gelatin capsules) administration. Br J Clin Pharmacol
23. Hsu A, Granneman GR, Cao G, Carothers L, Japour A, El-Shourbagy T, et al. Pharmacokinetic interaction between ritonavir and indinavir in healthy volunteers. Antimicrob Agents Chemother
24. Merry C, Barry MG, Mulcahy F, Ryan M, Heavey J, Tija JF, et al. Saquinavir pharmacokinetics alone and in combination with ritonavir in HIV-infected patients. AIDS
25. Piscitelli S, Bechtel C, Sadler B, Falloon J. The addition of a second protease inhibitor eliminates amprenavir-efavirenz drug interactions and increases plasma amprenavir concentrations. Seventh Conference on Retroviruses and Opportunistic Infections.
San Francisco, January 2000 [abstract 78].
26. Saah AJ, Winchell GA, Nessly ML, Seniuk MA, Rhodes RR, Deutsch PJ. Pharmacokinetic profile and tolerability of indinavir-ritonavir combinations in healthy volunteers. Antimicrob Agents Chemother
27. Sham HL, Kempf DJ, Molla A, Marsh KC, Kumar GN, Chen CM, et al. ABT-378, a highly potent inhibitor of the human immunodeficiency virus protease. Antimicrob Agents Chemother
28. Parkin NT, Lie YS, Hellmann NS, et al. Phenotypic changes in drug susceptibility associated with failure of human immunodeficiency virus type 1 (HIV-1) triple combination therapy. J Infect Dis
29. Petropoulos CJ, Parkin NT, Limoli KL, Lie YS, Wrin T, Huang W, et al. A novel phenotypic drug susceptibility assay for human immunodeficiency virus type 1. Antimicrob Agents Chemother
30. Karmochkine M, Mohamed AS, Piketty C, Ginsburg C, Raguin G, Schneider-Fauveau V, et al. The cumulative occurrence of resistance mutations in the HIV-1 protease gene is associated with failure of salvage therapy with ritonavir and saquinavir in protease inhibitor-experienced patients. Antiviral Res
31. Walter H, Schmidt B, Rascu A, Helm M, Moschik B, Paatz C, et al. Phenotypic HIV-1 resistance correlates with treatment outcome of nelfinavir salvage therapy. Antiviral Ther
32. Perez-Elias MJ, Lanier R, Munoz V, Garcia-Arata I, Cassado JL, Marti-Belda P, et al. Phenotypic testing predicts virological response in successive protease inhibitor-based regimens. AIDS
33. Casado JL, Dronda F, Hertogs K, Sabido R, Antela A, Marti-Belda P, et al. Efficacy, tolerance, and pharmacokinetics of the combination of stavudine, nevirapine, nelfinavir, and saquinavir as salvage regimen after ritonavir or indinavir failure. AIDS Res Human Retroviruses
34. Zolopa AR, Shafer RW, Warford A, Montoya JG, Hsu P, Katzenstein DA, et al. HIV-1 genotypic resistance patterns predict response to saquinavir-ritonavir therapy in patients in whom previous protease inhibitor therapy had failed. Ann Intern Med
35. Lawrence J, Schapiro J, Winters M, Montoya J, Zolopa A, Pesano R, et al. Clinical resistance patterns and response to two sequential protease inhibitor regimens in saquinavir and reverse transcriptase inhibitor-experienced patients. J Infect Dis
36. Para MF, Glidden DV, Coombs RW, Collier AC, Condra JH, Craig C, et al. Baseline human immunodeficiency virus type 1 phenotype, genotype, and RNA response after switching from long-term hard-capsule saquinavir to indinavir or soft-gel-capsule saquinavir in AIDS Clinical Trials Group protocol 333. J Infect Dis
37. Durant J, Clevenbergh P, Halfon P, Delgiudice P, Porsin S, Simonet P, et al. Drug-resistance genotyping in HIV-1 therapy: the VIRADAPT randomised controlled trial. Lancet
38. Baxter JD, Mayers DL, Wentworth DN, Neaton JD, Hoover ML, Winters MA, et al. A randomized study of antiretroviral management based on plasma genotypic antiretroviral resistance testing in patients failing therapy. CPCRA 046 Study Team for the Terry Beirn Community Programs for Clinical Research on AIDS. AIDS
39. Clevenbergh P, Durant J, Halfon P, del Giudice P, Mondain V, Montagne N, et al. Persisting long-term benefit of genotype-guided treatment for HIV-infected patients failing HAART. The Viradapt Study: week 48 follow-up. Antiviral Ther
40. Cohen CJ, Hunt S, Sension M, Farthing C, Conant M, Jacobson S, et al. A randomized trial assessing the impact of phenotypic resistance testing on antiretroviral therapy. AIDS
41. Bangsberg DR, Hecht FM, Charlebois ED, Zolopa AR, Holodniy M, Sheiner L, et al. Adherence to protease inhibitors, HIV-1 viral load, and development of drug resistance in an indigent population. AIDS
42. Benson CA, Deeks SG, Brun SC, Gulick RM, Eron JJ, Kessler HA, et al. Safety and antiviral activity at 48 weeks of lopinavir/ritonavir plus nevirapine and 2 nucleoside reverse-transcriptase inhibitors in human immunodeficiency virus type 1-infected protease inhibitor-experienced patients. J Infect Dis
43. Liechty C, Alexander C, Harrigan R, et al. Are random antiretroviral drug levels associated with objectively measured adherence behavior? Tenth Conference on Retroviruses and Opportunistic Infections.
Boston, MA 2003. [abstract 529].