Lucas, Gregory M.; Chaisson, Richard E.; Moore, Richard D.
The introduction of protease inhibitors (PI) and the advent of highly active antiretroviral therapy (HAART) produced a quantum leap in HIV treatment [1–4]. However, initial enthusiasm for PI-based therapy has been tempered by high adherence demands [5–7], low rates of durable viral suppression in unselected clinical cohorts , short- and long-term adverse effects [9,10], and the frequent development of antiretroviral resistance [11,12].
The number of options in, and the complexity of HAART have increased markedly in the last 5 years. Additionally, many new antiretroviral agents and classes of drugs are in clinical development . There are theoretical advantages and disadvantages to combination therapy that is based on a single PI, ritonavir (RTV) plus saquinavir (SQV), or efavirenz (EFV). According to consensus guidelines, these three regimen types are considered first-line choices in HIV-infected individuals in whom antiretroviral therapy is indicated [14,15]. Aggregate experience and data supporting clinical effectiveness are greatest for regimens based on a single PI [15,16]. The use of low-dose RTV in combination with SQV or another PI has been proposed as a strategy to reduce dosing frequency, and substantially improve the pharmacokinetic profile of the second PI [17–20]. EFV-based regimens are appealing because of high potency , a long half-life with once daily dosing [22,23], and the potential avoidance of long-term adverse effects associated with PI. However the genetic threshold to drug resistance is lower with EFV than with PI: a single mutation in the reverse transcriptase gene of HIV-1 produces high-level phenotypic resistance to the entire non-nucleoside reverse transcriptase inhibitor (NNRTI) class of agents [11,22,24], while clinically important PI resistance generally requires multiple mutations .
Recent data from a clinical trial suggested that EFV-based therapy is more effective in suppressing HIV-1 RNA than PI-based therapy in minimally pretreated HIV-1-infected patients . Community-based cohort studies reflect actual use of various treatment strategies and provide important complementary information to clinical trials. The effectiveness of therapy outside of the clinical trial setting may be substantially different than the efficacy documented in clinical trials . In the setting of a university-based, urban HIV clinic, we compared the virologic and immunologic effectiveness of initial combination therapy with nucleoside reverse transcriptase inhibitors (NRTI) plus either: a single PI; RTV/SQV; or EFV.
Setting and database
The Johns Hopkins HIV Clinic currently provides care for approximately 2500 patients in the Baltimore area. Comprehensive baseline assessments of all patients enrolling in the clinic are conducted by clinicians and social workers through structured interviews and recorded on standardized forms. The clinic-based medical record maintains a section for each visit to document prescribed therapy by treatment name, dose and number of refills. The record is also updated when prescriptions are filled over the telephone or mailed to patients.
An observational clinical database comprising all patients attending the Clinic was established in 1990 . The database contains information on more than 4000 patients who have been treated in the Clinic. Information from clinical records is reviewed and abstracted by trained technicians onto structured data collection forms, then entered into an automated database. The Clinic medical records, the main hospital medical record and various institutional automated databases (e.g., laboratory, radiology, pathology, discharge summaries) are abstracted. Comprehensive demographic, clinical, laboratory, pharmaceutical, and psychosocial data are collected at times corresponding to enrollment in the clinic and at 6 month intervals thereafter. The clinical cohort study has been approved by the Johns Hopkins Institutional Review Board.
Design and measurements
We conducted a retrospective analysis of an observational cohort to compare the effectiveness HAART regimens based on a single PI, RTV/SQV, or EFV. HIV-1-infected patients were included in the analysis if they: were ≥ 18 years of age; had started combination therapy with ≥ 3 antiretroviral drugs and ≥ 1 NRTI to which they had not been exposed previously; were PI and NNRTI naive; and had prior exposure to ≤ 2 NRTI. Patients in whom SQV was used as the sole PI or in whom SQV had been used previously were excluded from the analysis because the hard-gelatin preparation in use for much of the study period had poor bioavailability  and was associated with lower rates of viral suppression than other PI [28,29]. When used in combination, RTV and SQV were each dosed at 400 mg twice daily in our clinic.
Race was categorized as African–American or White. Asian and Hispanic patients were categorized as white, because they comprised < 2% of the study population, precluding meaningful subgroup comparisons. History of injecting drug use was determined at the time of enrollment into the Clinic. Injecting drug use activity or abstinence at the time HAART was initiated was not documented.
HIV-1 RNA levels and CD4 cell counts were obtained as part of routine clinical practice. In patients beginning HAART in the clinic, these values are typically obtained at baseline, at 4–6 weeks after starting therapy, 3 months after starting therapy, and at 3 month intervals thereafter, or more frequently if clinically indicated. HIV-1 RNA levels were measured using the Roche Amplicor assay (Roche Molecular Systems, Branchburg, New Jersey, USA) and CD4 cell counts were measured by flow cytometry. Ultrasensitive HIV-1 RNA assays (limit of detection 50 copies/ml) were only used in the latter part of the study period. Therefore, achieving viral suppression at this level was not included as an outcome. The baseline HIV-1 RNA level and CD4 cell count were defined as the values obtained closest to, but not after the date on which HAART was initiated.
The baseline demographic and clinical characteristics were determined in patients starting HAART regimens with a single PI, RTV/SQV, or EFV. Categorical variables were compared with contingency tables and Fisher's exact test. Continuous variables were compared with the Wilcoxon rank-sum test. All tests were two-sided, and P values < 0.05 were considered to be statistically significant.
An intent-to-treat analysis (missing = failure) was used to identify factors associated with achieving viral suppression (HIV-1 RNA < 400 copies/ml) within 8 months of starting combination therapy (primary outcome). The proportion of patients who did not have HIV-1 RNA levels measured within 8 months of starting therapy were compared in the three groups. In clinical trials, most patients who achieve viral suppression do so within 6 months of starting HAART . Because this study was conducted in a cohort of patients followed in routine clinical practice, we chose 8 months as the outer boundary so as not to exclude patients who achieved viral suppression by their third visit after starting HAART if that visit was slightly more than 6 months after the start date. Variables included in the analysis were regimen type (single PI, RTV/SQV, or EFV), sex, race, age (< 40 versus ≥ 40 years), history of injecting drug use, men having sex with men exposure, baseline CD4 cell count (< 50, 50–200, and > 200 × 106 cells/l), baseline HIV-1 RNA level (< 30 000, 30 000–100 000, and > 100 000 copies/ml), prior NRTI exposure (naive versus 1–2 drugs), number of NRTI included in the regimen to which the patient had not been previously exposed (1 versus ≥ 2), and calendar period in which HAART was initiated (1996–2000). Multivariate analysis was performed using logistic regression to calculate odds ratios (OR), and 95% confidence intervals (CI). Variables associated with viral suppression (P < 0.1) were included in the multivariate model. Potential interactions between variables were assessed by stratification, and combining terms in logistic regression analyses.
The proportions of patients who discontinued HAART within 2 months of starting were compared between the three groups. Two-months was chosen as the decision point because discontinuing therapy prior to this time would more likely reflect intolerance of the regimen or non-adherence than virologic failure per se. An as-treated analysis of viral suppression was conducted, in which patients who discontinued therapy within the first 2 months were excluded.
The durability of viral suppression was analyzed in two ways. First, in the subset of patients who achieved viral suppression within 8 months of starting HAART, time to viral rebound (> 1000 HIV-1 RNA copies/ml) was compared between the three regimen types with a Kaplan–Meier survival function and the log-rank test. A Cox proportional hazards model was used to estimate relative hazards of viral rebound in the three groups after adjustment for potential confounding factors. The proportional hazards assumption was checked with graphical means and Schoenfeld residuals [31,32]. Second, in patients who were followed in the clinic for > 8 months after starting HAART, we dichotomized virologic response as durable viral suppression (more than three consecutive HIV-1 RNA measurements < 400 copies/ml for a period of > 6 months) or failure to achieve durable viral suppression. The proportions of patients who achieved durable viral suppression were compared between the three groups.
The change in CD4 cell count was defined as the difference between the highest CD4 cell count recorded within 12 months of starting HAART and the baseline CD4 cell count. Variables associated with change in CD4 cell count were analyzed with the Wilcoxon rank-sum test. All analyses were conducted using Stata 6.0 (College Station, Texas, USA).
Patients and their baseline characteristics
A total of 545 patients started a first HAART regimen between February 1996 and May 2000 and met inclusion criteria for the study. The HAART regimen included a single PI in 416 (indinavir, 164; RTV, 35; nelfinavir, 214; amprenavir, 3), RTV/SQV in 68, and EFV in 61. Patients in the three groups did not differ significantly by sex, race, or sexual exposure categories (Table 1). Patients in the EFV group were older than patients starting a single PI (P = 0.01) or patients starting RTV/SQV (P = 0.004, Table 1). Thirty-four percent of patients in the RTV/SQV group had a history of injecting drug use, compared to 49% in the single PI group (P = 0.03) and 52% in the EFV group (P = 0.05). The median baseline CD4 cell count was significantly lower and the baseline HIV-1 RNA was significantly higher in the RTV/SQV group than in the single PI group (Table 1).
Sixty-nine percent of patients starting EFV were NRTI-naive, compared to 49% in the single PI group (P = 0.004) and 47% in the RTV/SQV group (P = 0.014). As anticipated from the inclusion criteria, 72% of patients in the RTV/SQV group used a regimen with four or more antiretroviral agents compared to 13% in the single PI group and 11% in the EFV group (P < 0.001 for both comparisons with RTV/SQV). Consistent with the availability of different antiretroviral preparations and temporal trends in the clinic, the median start time was June 1997 in the single PI group, January 1998 in the RTV/SQV group, and June 1999 in the EFV group (P < 0.01 for all two-group comparisons). Twenty patients (3.7%) received combination antiretroviral therapy as part of a clinical trial (16 receiving indinavir, and one each receiving nelfinavir, amprenavir, RTV/SQV, and EFV).
Initial viral suppression
By intent-to-treat analysis (missing = failure), viral suppression within 8 months of initiating HAART was achieved by 72% of patients in the EFV group, compared to 49% of patients in the single PI group (P = 0.001) and 51% of patients in the RTV/SQV group (P = 0.019; Table 2). Viral load measurements prior to beginning HAART were unavailable in 8% (35/416) receiving a single PI, 6% (4/68) receiving RTV/SQV, and 0% (0/61) receiving EFV. The proportions of patients without a follow-up HIV-1 RNA measurement within 8 months of initiating HAART were similar in the three groups (single PI, 13%; RTV/SQV, 15%; EFV, 10%). Rates of viral suppression were similar for the individual PI in the single PI group: 45% (73/164) for indinavir, 46% (16/35) for RTV, 54% (115/214) for nelfinavir, and 33% (1/3) for amprenavir (P = 0.3 for overall comparison). The difference in viral suppression between EFV-based therapy and PI-based therapy was similar in patients with low and high baseline HIV-1 RNA levels (Fig. 1). Other factors significantly associated with viral suppression included race, history of injecting drug use, prior NRTI exposure, baseline CD4 cell count, and baseline HIV-1 RNA level (Table 2). Sex, participation in a clinical trial, and men-having-sex-with-men exposure were not associated with viral suppression (data not shown).
After adjustment for multiple factors in a logistic regression model, EFV-based therapy remained significantly associated with viral suppression (OR, 2.7; 95%CI, 1.1–6.7, compared with single PI; Table 2). There were no statistically significant interactions among the variables evaluated. Because EFV was used only during the latter part of the study, we assessed the possibility that a temporal trend in viral suppression rates explained the differences observed in the three regimen types. Although calendar year was significantly associated with viral suppression in the cohort overall (Table 2), calendar time was not significantly associated with viral suppression among the subset of patients who used a single PI or RTV/SQV (P = 0.191, Fisher's exact test). In addition, calendar year was not significantly associated with viral suppression in the multivariate model (Table 2).
Sixteen percent of patients discontinued their antiretroviral regimen within 2 months of starting [indinavir 10% (17/164), RTV 29% (10/35), nelfinavir 14% (31/214), amprenavir 0% (0/3), RTV/SQV 29% (20/68), EFV 13% (8/61)]. Patients using a regimen that included RTV were significantly more likely to discontinue HAART within 2 months of starting than patients using other regimens (29% versus 13%; P < 0.001). There were no significant differences in the proportions discontinuing HAART among regimens that did not include RTV (P = 0.6). In the as-treated analysis, from which patients who discontinued HAART within 2 months were excluded, viral suppression was achieved by 77% in the EFV group, 55% in the single PI group (P = 0.003 compared with EFV), and 67% in the RTV/SQV group (P = 0.16 compared to single PI, and P = 0.27 compared to EFV).
Durability of viral suppression
Among patients who initially achieved viral suppression, we compared time to viral rebound in the three groups with a Kaplan–Meier survival function (Fig. 2). The time-to-rebound curves were similar for patients treated with a single PI, RTV/SQV, and EFV (P > 0.5 for all two-group comparisons, log-rank test). The estimated probability of viral rebound at 1 year was 43% in the single PI group, 39% in the RTV/SQV group, and 41% in the EFV group. In a Cox proportional hazards model including race, injecting drug use, baseline CD4 cell count, and baseline HIV-1 RNA level; use of EFV [hazards ratio (HR), 0.9; 95%CI, 0.5–1.5] or RTV/SQV (HR, 0.8; 95%CI, 0.5–1.3) were not associated with viral rebound, compared to use of a single PI. Durable viral suppression, as defined by more than three consecutive HIV-1 RNA measurements < 400 copies/ml for > 6 months, was achieved by 53% in the EFV group, compared with 26% in the single PI group and 29% in the RTV/SQV group (P = 0.001 and P = 0.034 compared to EFV, respectively). After adjustment for race, age, history of injecting drug use, prior NRTI exposure, baseline CD4 cell count, and baseline HIV-1 RNA level, EFV remained strongly associated with durable viral suppression (OR, 2.9; 95%CI, 1.4–5.8).
CD4 cell count increases were similar in the single PI group (median 139 × 106 cells/l), the RTV/SQV group (median 120 × 106 cells/l), and the EFV group (median, 151 × 106 cells/l) (P > 0.3 for all two-group comparisons). Among patients achieving viral suppression, CD4 cell count increases were similar in the single PI group (185 × 106 cells/l), the RTV/SQV group (193 × 106 cells/l) and the EFV group (158 × 106 cells/l). Among patients who did not achieve viral suppression, there was a trend toward a lower median CD4 cell count increase in the RTV/SQV group (66 × 106 cells/l), compared to 83 × 106 cells/l in the single PI group (P = 0.08) and 114 × 106 cells/l in the EFV group (P = 0.19).
The median CD4 cell count increase was 215 × 106 cells/l in White patients and 121 × 106 cells/l in African–Americans (P < 0.001). Injecting drug users had a smaller median CD4 cell count increase than non-users (113 × 106 versus 157 × 106 cells/l, P = 0.003). Men who have sex with men had a median CD4 cell count increase of 179 × 106 cells/l compared to 120 × 106 cells/l in patients without this exposure (P < 0.001). The median CD4 cell count increase was higher in patients who were NRTI-naive (156 × 106 cells/l), compared to patients with prior NRTI exposure (125 × 106 cells/l, P = 0.032). Changes in CD4 cell counts did not differ significantly according to sex, age, baseline CD4 cell count, or baseline HIV-1 RNA level (data not shown).
As therapeutic options for HIV-infected patients rapidly expand, treatment strategies must be continually reassessed [15,33–35]. Important issues that remain unresolved include: determining the optimal time to initiate HAART in order to preserve immune function and yet minimize long-term toxicities; identifying the most effective initial regimen; and determining the best criteria to decide when therapy should be changed. Cohort studies provide important and timely information germane to current clinical dilemmas, and are an important complement to clinical trials. In minimally antiretroviral-experienced patients attending a large urban clinic, we compared the virologic and immunologic effectiveness of initial therapy for HIV-1 infection with NRTI combined with a single PI, RTV/SQV, or EFV. By intent-to-treat analysis (missing = failure), the proportion achieving viral suppression was approximately 20% higher in patients treated with an EFV-containing regimen than those treated with a PI-containing regimen. This difference remained statistically significant after adjustment for multiple baseline factors. Among the subset of patients who achieved initial viral suppression, time to viral rebound was similar in the three groups. Overall, durable viral suppression was achieved by 52% of patients in the EFV group, 26% in the single PI group, and 29% in the RTV/SQV group. Absolute increases in CD4 cell counts were higher in the EFV group than in the single PI and RTV/SQV groups, but not significantly so.
Our results are similar to those reported from the 006 Study , in which 70% of patients randomized to EFV plus zidovudine and lamivudine attained HIV-1 RNA < 400 copies/ml, compared with 48% of patients assigned to indinavir plus zidovudine and lamivudine. In this study, patients assigned to the indinavir plus two NRTI arm were given four 200 mg capsules of indinavir three times daily, twice the number of capsules used in routine clinical practice. The proportion of patients who discontinued treatment for any reason was higher in the group assigned to indinavir plus two NRTI (43%), than the group assigned to EFV plus two NRTI (27%, P = 0.005). An accompanying editorial by Clumeck  questioned whether the superiority of EFV over indinavir could be inferred from this open-label trial where design issues may have fostered better compliance and acceptance of the regimen by patients assigned EFV.
In our study, discontinuation of therapy was strongly associated with use of RTV-containing regimens. Higher rates of adverse effects in RTV recipients compared to recipients of other PI have been widely reported [4,8,9]. The proportions discontinuing therapy in our study were similar in patients taking EFV and those taking PI other than RTV. In the as-treated analysis, where patients who discontinued therapy in the first 2 months were excluded, EFV-containing therapy remained associated with significantly higher rates of viral suppression than regimens based on a single PI. However, in this analysis, the proportion achieving viral suppression in the RTV/SQV group was similar to the EFV group. This suggests that RTV/SQV, while associated with a high rate of early discontinuation, has similar antiretroviral potency to EFV. In our Clinic 400 mg of RTV was generally combined with 400 mg of SQV twice daily in the RTV/SQV combination. Regimens based on lower doses of RTV in conjunction with a second PI , or the newly released combination preparation of lopinavir and low-dose RTV , may prove highly effective if tolerated better than higher doses of RTV.
The ultimate goal of antiretroviral therapy is preservation of immune function and prevention of clinical disease progression and death. Treatment-mediated reductions in HIV-1 RNA levels have been strongly linked to clinical benefits [37–40]. Robust increases in the CD4 cell count and declines in the risk of HIV-related morbidity and mortality have been documented with HAART, even in the setting of incomplete viral suppression [4,41,42]. We chose HIV-1 RNA < 400 copies/ml as our primary outcome because this level of viral suppression has been associated with significantly reduced rates of HIV-1 disease progression in several large cohorts [1,3,43]. In the experience of our clinic, the probability of HIV disease progression or death was 18% in patients who did not achieve viral suppression on combination antiretroviral therapy versus 9% in patients who did (P < 0.001) . Interestingly, at 2 years of follow-up, there was no difference in disease progression in patients who achieved suppression with subsequent viral load rebound and in those with durable viral suppression.
Despite this, however, viral replication in the setting of HAART sets the stage for the progressive accumulation of resistance-conferring mutations [12,44], and leads to challenging management issues and expensive resistance testing [11,45,46]. The durability of viral suppression has been strongly correlated with the nadir level of HIV-1 RNA suppression [30,47]. Because HIV-1 RNA assays with different limits of detection were used throughout the time period of our study, we were not able to address this issue directly. However, there were no differences in time to rebound in the three groups, once initial viral suppression had been achieved. Overall, durable viral suppression was significantly associated with EFV-based therapy.
Our study has several limitations. First, the results reported reflect the experience of a single clinic and may not be applicable to other settings. The generally low rates of durable viral suppression in our cohort have been described previously , and this experience is probably representative of HIV-infected patients in other poor, urban treatment settings. Second, patients were not randomized to the treatments received, raising the possibility of confounding by indication. For example, it is possible the EFV was avoided in patients in whom adherence was perceived to be a potential problem. However, it is notable that EFV recipients were more likely to be African–American and have a history of injecting drug use, factors associated with poorer response to therapy in our cohort. Additionally, the difference we observed in viral suppression between EFV- and PI-containing regimens was large, and persisted after adjustment for potentially confounding factors. Third, because of temporal differences in the availability of antiretroviral agents, EFV was used exclusively during the latter part of the study period. Although, adjustment for calendar time in our multivariate model did not effect the association between regimen type and viral suppression, we cannot completely rule out that temporal effects, such as increased experience with HAART among clinicians or a greater emphasis on adherence, explain part of the association between EFV and viral suppression.
In conclusion, we found that EFV-based antiretroviral therapy was associated with a greater likelihood of achieving viral suppression than PI-based therapy in HAART-naive patients treated in an urban HIV clinic. In the subset of patients who suppressed HIV-1 RNA levels < 400 copies/ml, durability of suppression was similar in the three groups. Studies with long-term follow-up are needed to determine if initial therapy with NNRTI-containing regimens is associated with a lower incidence of adverse metabolic effects than PI-based therapy. Lastly, it will be important to identify differences in antiretroviral resistance profiles associated with early virologic failure of a first NNRTI-based or PI-based regimen.
Sponsorship: Supported by the National Institute on Drug Abuse (RO1-DA11602) and the Agency for Health Care Policy and Research (RO1-H507809).
1. Moore RD, Chaisson RE. Natural history of HIV infection in the era of combination antiretroviral therapy. AIDS 1999, 13: 1933–1942.
2. Palella FJ Jr, Delaney KM, Moorman AC. et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. HIV Outpatient Study Investigators. N Engl J Med. 1998, 338: 853–860.
3. Ledergerber B, Egger M, Opravil M. et al. Clinical progression and virological failure on highly active antiretroviral therapy in HIV-1 patients: a prospective cohort study. Swiss HIV Cohort Study. Lancet 1999, 353: 863–868.
4. Mocroft A, Devereux H, Kinloch-de-Loes S. et al. Immunological, virological and clinical response to highly active antiretroviral therapy treatment regimens in a complete clinic population. Royal Free Centre for HIV Medicine. AIDS 2000, 14: 1545–1552.
5. Paterson DL, Swindells S, Mohr J. et al. Adherence to protease inhibitor therapy and outcomes in patients with HIV infection. Ann Intern Med 2000, 133: 21–30.
6. Haubrich RH, Little SJ, Currier JS. et al. The value of patient-reported adherence to antiretroviral therapy in predicting virologic and immunologic response. California Collaborative Treatment Group. AIDS 1999, 13: 1099–1107.
7. Altice FL, Friedland GH. The era of adherence to HIV therapy. Ann Intern Med 1998, 129: 503–505.
8. 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 1999, 131: 81–87.
9. Sulkowski MS, Thomas DL, Chaisson RE, Moore RD. Hepatotoxicity associated with antiretroviral therapy in adults infected with human immunodeficiency virus and the role of hepatitis C or B virus infection. JAMA 2000, 283: 74–80.
10. Behrens G, Dejam A, Schmidt H. et al. Impaired glucose tolerance, beta cell function and lipid metabolism in HIV patients under treatment with protease inhibitors. AIDS 1999, 13: F63–F70.
11. Hirsch MS, Brun-Vezinet F, D'Aquila RT. et al. Antiretroviral drug resistance testing in adult HIV-1 infection: recommendations of an International AIDS Society-USA Panel. JAMA 2000, 283: 2417–2426.
12. Erickson JW, Gulnik SV, Markowitz M. Protease inhibitors: resistance, cross-resistance, fitness and the choice of initial and salvage therapies. AIDS 1999, 13 (suppl A): S189–S204.
13. Murphy RL. New antiretroviral drugs in development. AIDS 2000, 14 (suppl 3): S227–S234.
14. Carpenter CC, Cooper DA, Fischl MA. et al. Antiretroviral therapy in adults: updated recommendations of the International AIDS Society-USA Panel. JAMA 2000, 283: 381–390.
15. Guidelines for the Use of Antiretroviral Agents in HIV-Infected Adults and Adolescents.
Department of Health and Human Services and Henry J. Kaiser Family Foundation. http://www
. hivatis.org. Accessed January 2001.
16. Gulick RM, Mellors JW, Havlir D. et al. 3-year suppression of HIV viremia with indinavir, zidovudine, and lamivudine. Ann Intern Med 2000, 133: 35–39.
17. van Heeswijk RP, Veldkamp AI, Hoetelmans RM. et al. The steady-state plasma pharmacokinetics of indinavir alone and in combination with a low dose of ritonavir in twice daily dosing regimens in HIV-1-infected individuals. AIDS 1999, 13: F95–F99.
18. van Heeswijk RP, Veldkamp AI, Mulder JW. et al. Once-daily dosing of saquinavir and low-dose ritonavir in HIV-1-infected individuals: a pharmacokinetic pilot study. AIDS 2000, 14: F103–F110.
19. Hsu A, Granneman GR, Cao G. et al. Pharmacokinetic interactions between two human immunodeficiency virus protease inhibitors, ritonavir and saquinavir. Clin Pharmacol Ther 1998, 63: 453–464.
20. Kempf DJ, Marsh KC, Kumar G. et al. Pharmacokinetic enhancement of inhibitors of the human immunodeficiency virus protease by coadministration with ritonavir. Antimicrob Agents Chemother 1997, 41: 654–660.
21. Staszewski S, Morales-Ramirez J, Tashima KT. et al. Efavirenz plus zidovudine and lamivudine, efavirenz plus indinavir, and indinavir plus zidovudine and lamivudine in the treatment of HIV-1 infection in adults. Study 006 Team. N Engl J Med 1999, 341: 1865–1873.
22. Adkins JC, Noble S. Efavirenz. Drugs 1998, 56: 1055–1064.
23. Molina JM, Ferchal F, Rancinan C. et al. Once-daily combination therapy with emtricitabine, didanosine, and efavirenz in human immunodeficiency virus-infected patients. J Infect Dis 2000, 182: 599–602.
24. Bacheler LT, Anton ED, Kudish P. et al. Human immunodeficiency virus type 1 mutations selected in patients failing efavirenz combination therapy. Antimicrob Agents Chemother 2000, 44: 2475–2484.
25. Schmidt B, Walter H, Moschik B. et al. Simple algorithm derived from a geno-/phenotypic database to predict HIV-1 protease inhibitor resistance. AIDS 2000, 14: 1731–1738.
26. Moore RD. Understanding the clinical and economic outcomes of HIV therapy: The Johns Hopkins HIV clinical practice cohort. J Acquir Immune Defic Syndr 1998, 17 (suppl.1): S38–S41.
27. Deeks SG, Smith M, Holodniy M, Kahn JO. HIV-1 protease inhibitors. A review for clinicians. JAMA 1997, 277: 145–153.
28. Casado JL, Perez-Elias MJ, Antela A. et al. Predictors of long-term response to protease inhibitor therapy in a cohort of HIV-infected patients. AIDS 1998, 12: F131–F135.
29. Fatkenheuer G, Theisen A, Rockstroh J. et al. Virological treatment failure of protease inhibitor therapy in an unselected cohort of HIV-infected patients. AIDS 1997, 11: F113–F116.
30. Powderly WG, Saag MS, Chapman S, Yu G, Quart B, Clendeninn NJ. Predictors of optimal virological response to potent antiretroviral therapy. AIDS 1999, 13: 1873–1880.
31. Hess KR. Graphical methods for assessing violations of the proportional hazards assumption in Cox regression. Stat Med 1995, 14: 1707–1723.
32. Schoenfeld D. Partial residuals for the proportional hazards regression model. Biometrika 1982, 69: 239–241.
33. Henry K. The case for more cautious, patient-focused antiretroviral therapy. Ann Intern Med 2000, 132: 306–311.
34. Cohen OJ. Antiretroviral therapy: time to think strategically. Ann Intern Med 2000, 132: 320–322.
35. Clumeck N. Choosing the best initial therapy for HIV-1 infection. N Engl J Med 1999, 341: 1925–1926.
36. Murphy RL, Brun S, Hicks C. et al. ABT-378/ritonavir plus stavudine and lamivudine for the treatment of antiretroviral-naive adults with HIV-1 infection: 48-week results. AIDS 2001, 15: 1–9.
37. O'Brien WA, Hartigan PM, Martin D. et al. Changes in plasma HIV-1 RNA and CD4+ lymphocyte counts and the risk of progression to AIDS. Veterans Affairs Cooperative Study Group on AIDS. N Engl J Med 1996, 334: 426–431.
38. Marschner IC, Collier AC, Coombs RW. et al. Use of changes in plasma levels of human immunodeficiency virus type 1 RNA to assess the clinical benefit of antiretroviral therapy. J Infect Dis 1998, 177: 40–47.
39. Katzenstein DA, Hammer SM, Hughes MD. et al. The relation of virologic and immunologic markers to clinical outcomes after nucleoside therapy in HIV-infected adults with 200 to 500 CD4 cells per cubic millimeter. AIDS Clinical Trials Group Study 175 Virology Study Team. N Engl J Med 1996, 335: 1091–1098.
40. Murray JS, Elashoff MR, Iacono-Connors LC, Cvetkovich TA, Struble KA. The use of plasma HIV RNA as a study endpoint in efficacy trials of antiretroviral drugs. AIDS 1999, 13: 797–804.
41. Kaufmann D, Pantaleo G, Sudre P, Telenti A. CD4-cell count in HIV-1-infected individuals remaining viraemic with highly active antiretroviral therapy (HAART). Swiss HIV Cohort Study. Lancet 1998, 351: 723–724.
42. Deeks SG, Barbour JD, Martin JN, Swanson MS, Grant RM. Sustained CD4+ T cell response after virologic failure of protease inhibitor-based regimens in patients with human immunodeficiency virus infection. J Infect Dis 2000, 181: 946–953.
43. Thiebaut R, Morlat P, Jacqmin-Gadda H. et al. Clinical progression of HIV-1 infection according to the viral response during the first year of antiretroviral treatment. Groupe d'Epidemiologie du SIDA en Aquitaine (GECSA). AIDS 2000, 14: 971–978.
44. Belec L, Piketty C, Si-Mohamed A. et al. High levels of drug-resistant human immunodeficiency virus variants in patients exhibiting increasing CD4+ T cell counts despite virologic failure of protease inhibitor-containing antiretroviral combination therapy. J Infect Dis 2000, 181: 1808–1812.
45. Durant J, Clevenbergh P, Halfon P. et al. Drug-resistance genotyping in HIV-1 therapy: the VIRADAPT randomised controlled trial. Lancet 1999, 353: 2195–2199.
46. Baxter JD, Mayers DL, Wentworth DN. 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 2000, 14: F83–F93.
47. Pilcher CD, Miller WC, Beatty ZA, Eron JJ. Detectable HIV-1 RNA at levels below quantifiable limits by amplicor HIV-1 monitor is associated with virologic relapse on antiretroviral therapy. AIDS 1999, 13: 1337–1342.
© 2001 Lippincott Williams & Wilkins, Inc.