Secondary Logo

Journal Logo

Clinical: Original Papers

The use of plasma HIV RNA as a study endpoint in efficacy trials of antiretroviral drugs

Murray, Jeffrey S.; Elashoff, Michael R.; Iacono-Connors, Lauren C.; Cvetkovich, Therese A.; Struble, Kimberly A.

Author Information
  • Free

Abstract

Introduction

An increase in the number of approved antiretroviral drugs, identification of potent combination regimens, and widespread availability of assays for monitoring HIV RNA levels in plasma have resulted in new HIV treatment guidelines which emphasize maintaining maximal suppression of HIV replication [1,2]. Of the several assays for monitoring HIV RNA that are commercially available, only Amplicor HIV-1 Monitor (Roche Molecular Systems Inc., Somerville, New Jersey, USA) has been approved for diagnostic use by the Food and Drug Administration (FDA). These advances warrant a reassessment of clinical trial design for the approval of antiretroviral agents.

Under accelerated approval regulations (CFR§314, Subpart H), FDA has approved antiretroviral drugs based on favorable changes in CD4 T-lymphocyte counts and/or plasma HIV RNA, but has required clinical endpoint trials for traditional approval. A disadvantage of clinical endpoint studies for AIDS is that they rely on the willingness of study participants to remain on randomized study regimens until a clinical endpoint (e.g., opportunistic infections, malignancies, wasting, death) occurs. In the setting of routine viral load monitoring, clinical protocols which rely on continuation of a study regimen to document clinical disease progression, despite an apparent lack or loss of virologic suppression, are no longer feasible.

Compared with an endpoint of disease progression, plasma HIV RNA has practical advantages including: routine use in clinical practice, ease of measurement, patient acceptability, earlier evaluation of drug activity, the opportunity to rapidly identify loss of response, and economy of sample size.

Given these practical advantages, drug development and traditional approval may be facilitated by the use of plasma HIV RNA similar to the way in which blood pressure and cholesterol measurements have been used for the traditional approval of antihypertensive and hypolipidemic drugs, respectively. The use of endpoints such as cholesterol or plasma HIV RNA is appropriate when treatment effects on these measurements have been shown to predict clinical benefit. Results of several clinical studies [3-7] have suggested that treatment-induced decreases in plasma HIV RNA lower the risk of clinical progression; however, relationships between magnitude, duration, and nadir of virologic change and risk of disease progression have not been characterized fully. A better description of the behavior of plasma HIV RNA changes in response to potent combination therapy may help with the design of studies using plasma HIV RNA as a primary endpoint.

Materials and methods

Variation in HIV RNA measurements

Total variability in plasma HIV RNA was evaluated using serial plasma samples from participants in AIDS Clinical Trials Group (ACTG) studies 175, 229, 076 [7-9] and the Women and Infants Transmission Study [10]. HIV RNA measurements were assayed using versions of the Amplicor Monitor Assay (Roche Molecular Systems, Inc.) at laboratories participating in the ACTG viral quality assurance proficiency testing program.

Changes in plasma HIV RNA and HIV disease progression

Table 1 lists the studies used to explore the relationship between treatment-induced reductions in viral load and the risk of HIV disease progression. Characteristics of study participants including prior antiretroviral experience, pretreatment CD4 T-lymphocyte counts, and pretreatment plasma HIV RNA levels are shown. For most analyses, studies and/or treatment arms were pooled. Clinical progression was defined as the occurrence of new AIDS-defining events or death. HIV RNA in plasma was measured using versions of the Amplicor HIV-1 Monitor assay (Roche Molecular Systems, Inc.) except where indicated otherwise.

T1-8
Table 1:
Characteristics of studies used for analyzing changes in HIV RNA in response to treatment.

For ACTG studies, the relative risks of clinical progression (first new AIDS-defining event or death) subsequent to week 24 for several HIV RNA response categories were estimated using a proportional hazards model, adjusting for both baseline HIV RNA level and CD4 T-lymphocyte count. The analyses was conducted in 1000 individuals who had not experienced clinical progression prior to week 24 and who had at least one HIV RNA measurement between weeks 16 and 24.

For Glaxo-Wellcome studies, the incidence of clinical progression for multiple AIDS-defining events per 100 patient-years was estimated for five response category groupings defined according to the single lowest HIV RNA achieved: <5000, 5001-20000, 20001-50000, 50001-200000, and >200000 copies/ml. Analyses were stratified by median baseline HIV RNA level.

For Pharmacia and Upjohn studies, HIV RNA was measured using a PCR assay developed by Pharmacia and Upjohn. Results using this assay were divided by a factor of 4.68 to convert values to an Amplicor HIV-1 Monitor assay equivalent. For the analysis that evaluated duration of response and risk of disease progression, only those study participants who were in the study for at least 24 weeks and did not experience clinical progression during the first 24 weeks were included. Duration of virologic response categories (response defined as a 0.5 log10 copies/ml reduction) were divided into quintiles. Proportional hazards regression models with adjustments for baseline CD4 T-lymphocyte counts and HIV RNA levels were used to evaluate the association between the duration of virologic response and the risk of clinical progression in reference to a non-responder group.

Characteristics of HIV RNA changes in response to antiretroviral therapy

Analyses assessing time to an HIV RNA level below 500 copies/ml used data pooled from indinavir- containing treatment arms from Merck studies 028, 033 and 035 [11,20,21]. Only patients who responded (n=204), defined as two consecutive HIV RNA levels below 500 copies/ml, were included in the analyses; time to response was defined as the first time point below 500 copies/ml. The proportion of patients below the assay limit of quantification (LoQ) was estimated every 4 weeks after start of treatment to week 24. Analyses comparing the time to HIV RNA level below 500 copies and 50 copies/ml were conducted on participants receiving indinavir-containing treatment regimens in study 035 [20]. Only patients who achieved HIV RNA levels <50 copies/ml (n=34) were included. The proportions of patients below 500 and 50 copies/ml were calculated every 4 weeks post-treatment onset to 28 weeks. HIV RNA in serum was measured using versions of the Amplicor Monitor assay in which 500 copies/ml was the estimated LoQ and 50 copies/ml was the estimated lower limit of detection (LoD) [21].

Multivariable analyses using Cox proportional hazards models were used to examine the effects of baseline CD4 T-lymphocyte counts, HIV RNA, and prior treatment (<6 months versus ≥6 months of nucleoside analogs) on the time to loss of virologic response for patients receiving zidovudine and lamivudine in Glaxo-Wellcome studies. In these studies plasma HIV RNA was measured every 4-8 weeks. Loss of virologic response was defined as a plasma HIV RNA level ≥400 copies/ml, the estimated LoQ.

For analysis of Boehringer-Ingelheim Pharmacuticals Inc. (BIPI) study 1046 [19], HIV RNA was assayed using both the Amplicor HIV-1 Monitor assay, LoQ approximately 400 copies/ml and the Roche ultradirect assay, estimated LoD approximately 20 copies/ml. Loss of virologic response was defined as an increase in HIV RNA to within 1 log10 copies/ml of baseline.

Fluctuations of HIV RNA levels around the LoQ were analyzed among a subset of participants (n=177) in Agouron study 511 [11] who had sustained suppression of HIV RNA below the LoQ out to approximately 40 weeks. HIV RNA measurements were assessed using a Chiron branched-chain DNA assay that has an estimated LoQ of approximately 500 copies/ml.

Results

Variation in plasma HIV RNA measurements

Based on serial plasma HIV RNA measurements (Amplicor HIV-1 Monitor, Roche Molecular Systems Inc.) sampled 2 weeks to 6 months apart from 663 participants in three ACTG studies and the Women and Infants Transmission Study [7-10], the total variability of sequential plasma HIV RNA measurements was estimated to be 0.5-0.7 log10 copies/ml, approximately three- to fivefold for absolute plasma HIV RNA levels [11]. This estimate includes intra-assay and biologic variability but not treatment effects as patients were not initiating antiretroviral therapy between measurements. Measurement variability may exceed this estimate for other types of HIV RNA assays or when the approved Amplicor HIV-1 Monitor assay is performed in a laboratory not meeting performance standards.

Relationships between changes in plasma HIV RNA and reduction in HIV disease progression

Five separate groups of analyses (Table 1), including a total of more than 5000 study participants, showed an association between treatment-induced decreases in plasma HIV RNA and a reduction in the risk of AIDS disease progression [11]. Mean decreases in plasma HIV RNA of 0.5 log10 copies/ml (approximating estimates of measurement variability) were associated with decreases in the risk of disease progression. A dose-response relationship for plasma HIV RNA and clinical disease progression was observed (i.e., larger mean decreases in plasma HIV RNA were associated with larger decreases in the risk of clinical progression).

Selected examples from these analyses illustrate the relationship between the risk of disease progression and virologic changes as assessed by: (i) the magnitude of initial plasma HIV RNA decrease; (ii) the nadir of plasma HIV RNA achieved [22]; and (iii) the duration of plasma HIV RNA decrease.

Magnitude of plasma HIV RNA decrease and the risk of progression

Risks of disease progression calculated for specified categories of plasma HIV RNA reduction for patients participating in several ACTG trials were calculated and have been published [23]. The magnitude of plasma HIV RNA reduction was categorized according to the change from baseline at a time point measured 16-24 weeks after initiation of treatment. Results showed that larger reductions in plasma HIV RNA were associated with smaller risks of disease progression. Patients in the category comprising the smallest magnitude of plasma HIV RNA decrease (0-0.5 log10) had a lower risk of disease progression compared with those with no plasma HIV RNA reduction from baseline. Patients with plasma HIV RNA reductions of 0.5-1.0 and 1.0 log10 copies/ml had incrementally smaller risks for disease progression in a dose-response manner.

Plasma HIV RNA nadir and the risk of clinical progression

When decreases in plasma HIV RNA were evaluated according to nadir achieved (lowest value within 52 weeks from randomization), a direct relationship between plasma HIV RNA reduction and risk of disease progression was also demonstrated (Fig. 1). Based on pooled data from several Glaxo-Wellcome trials (Table 1), patients achieving lower plasma HIV RNA nadirs had fewer new clinical events (incidence per 100 patient-years). The relationship between disease progression and plasma HIV RNA reduction was similar for patients with pretreatment plasma HIV RNA levels above and below the median value (Fig. 2), suggesting that the association between plasma HIV RNA nadir and the risk of progression is independent of baseline plasma HIV RNA.

F1-8
Fig. 1. :
Incidence of clinical progression to new AIDS event or death (per 100 patient-years) for multiple clinical events per patient. Patients were stratified by median baseline plasma HIV RNA: >63000 copies/ml (solid line) and baseline plasma HIV RNA £63000 copies/ml (broken line). 95% Confidence intervals are shown as vertical bars.
F2-8
Fig. 2. :
Time to maximal virologic response. (a) Proportion of patients who remain above the assay limit of detection (LoD), 50 copies/ml (broken line) or limit of quantification, 500 copies/ml (solid line) for indinavir-containing arms of study 035. 95% Confidence intervals are shown as vertical bars. Only patients whose plasma HIV RNA level fell below the LoD, of <50 copies/ml are included (n= 34). (b) Proportion who remain above the assay quantification limit of 500 copies/ml. Only patients whose plasma HIV RNA level fell below the limit are included (71% of patients reached 500 copies/ml). 95% Confidence intervals are shown as vertical bars. Subjects are stratified by baseline plasma HIV RNA: >50000 (n=45; dotted line), >20000 and £50000 (n=50; dashed line), >10 000 and £20000 (n=41; solid line), and £10000 (n= 64; broken line).

Duration of plasma HIV RNA response and risk of clinical progression

Based on data pooled from two Pharmacia and Upjohn studies of delavirdine (Table 1), a longer duration of virologic response within the first 24 weeks was associated with a lower risk of disease progression. For this analysis virologic response was defined as maintaining plasma HIV RNA levels at least 0.5 log10 copies/ml below pretreatment levels. Hazard ratios for the risk of disease progression are presented in Table 2 for several duration of virologic response categories (quintiles). In this analysis the association between the risk for clinical progression and maintaining a virologic response past 24 weeks was not evaluated.

T2-8
Table 2:
Risk of AIDS disease progression according to duration of virologic response for participants in two delavirdine studies.

Characteristics of plasma HIV RNA changes in response to antiretroviral therapy

Providing evidence that treatment-induced decreases in plasma HIV RNA confer eventual clinical benefit supports the usefulness of plasma HIV RNA as an endpoint in clinical trials; however, the design of protocols requires additional knowledge regarding the behavior of HIV RNA in response to treatment. Therefore, time to maximal virologic response, time to loss of maximal response, and the duration of maximal response were characterized and evaluated with respect to baseline characteristics and nadir achieved. Fluctuations of plasma HIV RNA levels in patients with sustained virologic suppression were also explored.

Time to maximal virologic response below assay limits

In an analysis of patients receiving indinavir-containing regimens in Merck study 035 [20], all patients who achieved a serum HIV RNA nadir below the LoD of 50 copies/ml (n=34) did so by 28 weeks (Although the Amplicor HIV-1 Monitor Assay, standard version, has an approved LoQ of approximately 400-500 copies/ml, the qualitative LoD has been estimated to be 50 copies/ml [21]). It took 16 weeks for all 34 patients to achieve serum HIV RNA levels below the LoQ of 500 copies/ml. The proportions of patients remaining above the assay LoD and LoQ at 4-week intervals after treatment onset are shown in Fig 2a.

Based on data pooled from several Merck studies (Table 1) in which participants received indinavir- containing regimens, time to virologic response below LoQ was dependent on baseline serum HIV RNA level. Patients with higher baseline serum HIV RNA values took somewhat longer to achieve serum HIV RNA levels below 500 copies/ml as shown in Figure 2b. However, almost all patients who responded (defined as a serum HIV RNA level <500 copies/ml) had levels below the LoQ by 16 weeks regardless of baseline HIV RNA level.

Pretreatment plasma HIV RNA was also found to be an important variable with respect to the proportion of individuals who achieved levels below the LoQ. In several Glaxo-Wellcome studies (Table 1), in which most patients received therapy with two nucleoside analogs (mainly zidovudine and lamivudine), the proportion of patients achieving plasma HIV RNA levels <400 copies/ml decreased as pretreatment plasma HIV RNA levels increased. The proportion of patients with a treatment-induced virologic nadir below the LoQ (400 log10 copies/ml) were 87%, 91%, 68%, 44%, and 9% for patients with pretreatment HIV RNA levels of <5000, 5000-20000, 20001-50000, 50001-200000 and >200000 log10 copies/ml, respectively.

Time to loss of virologic response

The risks of losing a virologic response according to pretreatment characteristics was calculated for patients participating in Glaxo-Wellcome studies. For these analyses, loss of virologic response was defined retrospectively as an increase in plasma HIV RNA level above 400 copies/ml after an initial decrease below this level. Time to loss of virologic response was analyzed in relation to pretreatment plasma HIV RNA level, CD4 T-lymphocyte count, and prior nucleoside reverse transcriptase inhibitor (NRTI) therapy. The risk of losing a virologic response was significantly associated with pretreatment plasma HIV RNA and length of previous treatment with NRTI. Patients with lower baseline HIV RNA levels and <6 months of prior NRTI therapy had lower risks of losing a virologic response than patients with higher baseline HIV RNA levels and >6 months of prior NRTI therapy [hazard ratio, 0.45; 95% confidence interval (CI) 0.35-0.58 and hazard ratio, 0.58; 95% CI, 0.40-0.83 respectively]. Pretreatment CD4 T-lymphocyte count did not independently affect the risk of losing a virologic response in the statistical model (hazard ratio, 0.95; 95% CI, 0.91-1.04).

A similar relationship between pretreatment plasma HIV RNA and loss of virologic response was observed for patients receiving a triple combination regimen of zidovudine, lamivudine, and nelfinavir (750 mg three times per day) in Agouron study 511 [11]. After approximately 40 weeks of follow-up, the percentages of patients experiencing confirmed virologic rebound (>1200 copies/ml using a branched-chain DNA assay) were 15%, 24%, and 32% for patients with pretreatment HIV RNA levels of £50000, >50000-100000 and >100000 copies/ml, respectively.

Based on analyses of the regimens listed in Table 1, time to loss of virologic response, defined as time from treatment onset to HIV RNA level above the LoQ, varied according to treatment regimen. Double nucleoside regimens appeared to have more limited virologic response times of approximately 12-20 weeks in treatment-naïve patients. Triple combination regimens with a non-NRTI (NNRTI) or protease inhibitor provided median response times below the assay LoQ for 1 year or greater for patients who were treatment-naïve (NNRTI combinations) or had no prior experience with protease inhibitors and lamivudine (protease combinations).

Virologic analyses of BIPI study 1046 [19] and data pooled from several Merck studies [11,20,21] showed that patients who achieved lower plasma HIV RNA nadirs maintained longer virologic suppression. Fig. 3 shows this relationship for participants in BIPI study 1046. Patients who achieved <20 copies/ml of HIV RNA maintained longer viral suppression than those patients with nadirs >20 copies/ml. There was no difference in response time for patients who achieved nadirs of >400 copies/ml compared to those whose nadir was between 20 and 400 copies/ml (between the LoQ for the Amplicor Monitor assay and estimated LoD for the more sensitive assay).

F3-8
Fig. 3. :
Time to loss of virologic response. (a) Time to plasma HIV RNA rebound to within 1 log10 copies/ml of baseline in Boehringer-Ingelheim Pharmacuticals Inc. study 1046. Subjects stratified by plasma HIV RNA nadir: <20 copies/ml (solid line), 20-400 copies/ml (dotted line), >400 copies/ml (broken line).

Similarly, for participants receiving indinavir- containing drug regimens in several Merck studies, those achieving serum HIV RNA levels <50 copies/ml (estimated LoD) maintained more durable viral suppression than those patients who achieved a nadir >500 copies/ml (LoQ). Those patients with a virologic nadir between 50 and 500 copies/ml (i.e., between the estimated LoD and LoQ) maintained an intermediate duration of virologic suppression [21].

Plasma HIV RNA fluctuations around the LoQ

Among 177 patients who achieved sustained suppression of plasma HIV RNA below the LoQ (<500 copies/ml) for approximately 40 weeks of follow-up in Agouron study 511 [11], 34 had at least one measurement above LoQ in between time points of sustained suppression below LoQ. Of the 34 patients with measurements temporarily rising above LoQ, 28 had one measurement above the LoQ, five patients had two measurements above the LoQ and one patient had three measurements above the LoQ. These results indicate that there may be fluctuation around the assay LoQ in patients who otherwise appear to have a sustained virologic response.

Discussion

The data summarized in this review may be helpful in guiding clinical trial design for assessing efficacy of antiretroviral drugs. These data support the use of plasma HIV RNA as an endpoint that is likely to confer clinical benefit and also provide information regarding how plasma HIV RNA may be analyzed and used for managing trial participants. However, much of the data were derived from studies evaluating dual therapy regimens that are no longer considered standard of care. Therefore, additional investigations to evaluate characteristics of plasma HIV RNA reductions for a greater duration in response to current standard antiretroviral regimens are needed.

In five separate analyses of more than 5000 patients, there was a clear association between initial decreases in plasma HIV RNA (within the first 24 weeks) and a reduction in the risk of disease progression and death. This relationship was observed across a range of patient characteristics including: pretreatment CD4 T- lymphocyte counts and RNA levels, prior drug experience, and treatment regimen. This relationship was observed when initial RNA reductions were evaluated using nadir, average change over time, or change from baseline at a specified time point. Evidence for this clinical association was strengthened by a dose-response relationship between plasma HIV RNA reduction and disease progression. In addition, a longer duration of initial HIV RNA reduction appeared to correlate with a decreased risk of clinical progression. Based on presentation of these data, an FDA advisory committee for the Division of Antiviral Drug Products concurred that favorable changes in plasma HIV RNA over a 16-24 week treatment period are a suitable basis for accelerated approval, whereas clinical trials that demonstrate durability (at least 48 weeks) of these favorable changes are a suitable basis for traditional approval of antiretroviral drugs.

Two analyses of triple combination regimens showed that patients who achieved plasma HIV RNA levels below the LoQ had longer viral suppression than those patients with nadirs above the lower LoQ using the Roche Amplicor Monitor assay. Patients (study 1046) who achieved even lower levels of HIV RNA, as measured using a more sensitive version of this assay had more sustained viral suppression. Based on the data presented, use of sensitive plasma HIV RNA assays with limits below that of the Roche Amplicor Monitor (400 copies/ml) assay may offer greater discrimination of treatment effects in clinical trials; however, quantitative and qualitative performance characteristics of new assays should be known when these measurements are used for the primary study endpoint. In addition, observations showing that larger plasma HIV RNA reductions confer both greater reductions in disease progression and more durable virologic responses support the concept of maximal and sustained suppression as reasonable goals of antiretroviral therapy. Although the incidence of AIDS-defining events was relatively low for patients with HIV RNA reductions to <5000 copies/ml (Fig. 1), greater reductions may be necessary to minimize the emergence of resistance and maximize long-term suppression of viremia. Given the goals of therapy, comparing the proportion of participants with plasma HIV RNA below the LoQ or evaluating time to loss of virologic response above a LoQ may be useful for stringently evaluating potent combination drug regimens. Such an endpoint also has practical advantages for clinical trial design. For example, those participants demonstrating loss of suppression would have the option of seeking other treatment without jeopardizing interpretation of the study endpoint.

Analyses characterizing the behavior of plasma HIV RNA in response to potent antiretroviral regimens may help to establish criteria for decision making in clinical studies. Data summarized in this review illustrate the importance of allowing adequate time to assess a virologic response. When plasma HIV RNA was measured using ‚ultra-sensitive‚ assays, some patients did not achieve plasma HIV RNA levels below quantification for 20-28 weeks, even when initial RNA decreases were abrupt. This data is consistent with data from previous studies of HIV kinetics [24] which indicate a biphasic reduction in RNA representing clearance of virus from productively and latently infected cellular reservoirs.

In addition, the data illustrate the importance of confirming virologic rebound with repeat measurements prior to implementing treatment changes. Isolated fluctuations around the LoQ may occur in patients who otherwise appear to have sustained virologic response. These fluctuations may be secondary to assay variability, non-adherence, or ongoing replication.

The appropriate duration of efficacy studies using virologic endpoints will depend on patient characteristics and the regimens studied. Sample size calculations and proposals for study duration should take these factors into account. Exploratory analyses presented in this review showed that the time to loss of virologic response differed according to baseline HIV RNA, previous antiretroviral treatment, and current regimen; therefore study investigators should consider stratifying participants for these factors.

Double NRTI therapy appeared to produce a median duration of viral suppression of <6 months. In patients with no prior antiretroviral treatment, a regimen containing two NRTI and either a protease inhibitor or an NNRTI appeared to produce more sustained viral suppression than regimens containing only two NRTI. For a regimen containing a protease inhibitor and at least one new NRTI, a majority of individuals on treatment maintained levels below the LoQ at 40-52 weeks; thus, the ability to detect potential differences between active triple combination regimens may require continuing studies for at least 40-52 weeks.

Although plasma HIV RNA appears to offer practical advantages as an endpoint and is likely to confer clinical benefit, the evaluation of CD4 T-lymphocyte changes and clinical response, including the incidence of adverse events, and AIDS-defining events and death will add further support for the effectiveness of an antiretroviral drug. Studies with statistical power to detect differences in clinical endpoints measuring disease progression and survival may also be necessary to address certain clinical questions or strategies.

Acknowledgements

The authors thank Abbott Laboratories, Agouron Pharmaceuticals, Inc., Boehringer Ingelheim Pharmaceuticals, Inc., Glaxo-Wellcome Inc., Harvard School of Public Health/ACTG, Hoffmann La Roche, Merck and Co., Inc., New England Research Institutes, Pharmacia and Upjohn.

References

1. Carpenter CC, Fishl MA, Hammer SM, et al.: Antiretroviral therapy for HIV infection in 1997. Updated recommendations of the International AIDS Society-USA panel. JAMA 1997, 277:1962-1969.
2. Fauci AS, Bartlett JG, Goosby EP, et al.: Guidelines for the use of antiretroviral agents in HIV infected adults and adolescents. MMWR 1998, 47(RR-5):43-82.
3. 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. N Engl J Med 1996, 355:1142-1143.
4. CAESAR Coordinating Committee: Randomized trial of addition of lamivudine or lamivudine plus loviride to zidovudine-containing regimens for patients with HIV-1 infection: CAESAR trial. Lancet 1997, 349:1413-1421.
5. Hammer SM, Squires KE, Hughes MD, et al.: A controlled trial of two nucleoside analogues plus indinavir in persons with human immunodeficiency virus infection and CD4 cell counts of 200 per cubic millimeter or less. N Engl J Med 1997, 337:725-733.
6. Cameron B, Heath-Chiozzi M, Danner S, et al.: Randomised placebo-controlled trial of ritonavir in advanced HIV-1 disease. Lancet 1998, 351:543-549.
7. Hammer SM, Katzenstein DA, Hughes MD et al.: A trial comparing nucleoside monotherapy with combination therapy in HIV-infected adults with CD4 cell counts from 200 to 500 per cubic millimeter. N Engl J Med 1996, 335:1081-1090.
8. Collier AC, Coombs RW, Schoenfeld DA, et al.: Treatment of human immunodeficiency virus infection with saquinavir, zidovudine, and zalcitabine. N Engl J Med 1996, 334:1011-1017.
9. Connor EM, Sperling RS, Gelber R, et al.: Reduction of maternal-infant transmission of human immunodeficiency virus type 1 with zidovudine treatment. N Engl J Med 1994, 331:1173-1180.
10. Sheon AR, Fox HE, Rich KC, et al.: The women and infants transmission study (WITS) of maternal-infant HIV transmission: study design, methods, and baseline data. J Women‚s Health 1996, 5:69-78.
11. FDA Division of Antiviral Drug Products Advisory Committee Meeting. July 1997, [Transcripts].
12. Dolin R, Amato D Fischl M, et al.: Zidovudine compared to didanosine in patients with advanced HIV-1 infection and little or no previous experience with zidovudine. Arch Intern Med 1995, 155: 961-974.
    13. D‚Aquila RT, Hughes MD, Johnson VA, et al.: Nevirapine, zidovudine, and didanosine compared with zidovudine and didanosine in patients with HIV-1 infections. A randomized, double-blind, controlled trial. Ann Intern Med 1996, 124:1019-1030.
    14. Fischl MA, Richman DD, Saag M, et al.: Safety and antiviral activity of combination therapy with zidovudine, zalcitabine and two doses of interferon alpha 2a in patients with HIV. AIDS Clinical Trials Group Study 197. J Acquir Immune Def Syndr Hum Retrovirol 1997, 16:247-253.
      15. Eron JJ, Benoit SL, Jemsek J, et al.: Treatment with lamivudine, zidovudine, or both in HIV-positive patients with 200 to 500 CD4+ cells/mm3. N Engl J Med 1995, 333:1662-1669.
      16. Katlama C, Ingrand D, Loveday C, et al.: Safety and efficacy of lamivudine-zidovudine combination therapy in antiretroviral naïve patients. JAMA 1996, 276:118-125.
      17. Bartlett JA, Benoit SL, Johnson VA, et al.: Lamivudine plus zidovudine compared with zalcitabine plus zidovudine in patients with HIV infection. Ann Intern Med 1996, 125:161-172.
      18. Staszewski S, Loveday C, Picazo JJ, et al.: Safety and efficacy of lamivudine-zidovudine combination therapy in zidovudine experienced patients. JAMA 1996, 276:111-117.
      19. Montaner JS, Reiss P, Cooper D, et al.: A randomized double-blind trial comparing combinations of nevirapine, didanosine, and zidovudine for HIV-infected patients: the INCAS trial, Netherlands, Canada and Australia Study. JAMA 1998, 279:930-937.
      20. Gulick RM, Mellors JW, Havlir D, et al.: Treatment with indinavir, zidovudine, and lamivudine in adults with human immunodeficiency virus infection and prior antiretroviral therapy. N Engl J Med 1997, 337:734-739.
      21. Drusano GL, Bilello JA, Stein DA, et al.: Factors influencing the emergence of resistance to indinavir: role of virologic, immunologic, and pharmacologic variables. J Infect Dis 1998, 178:360-367.
      22. Kempf DJ, Rode RA, Xu Y, et al.: The duration of viral suppression during protease inhibitor therapy for HIV-infection is predicted by plasma HIV-1 RNA at the nadir. AIDS 1998, 12:F9-F14.
      23. Marschner IC, Collieer AC, Coombs RW, et al.: The use of changes in plasma levels of human immunodeficiency virus type 1 RNA levels to assess the clinical benefit of antiretroviral therapy. J Infect Dis 1998, 177:40-47.
      24. Perelson AS, Essunger P, Cao Y, et al.: Decay characteristics of HIV-1-infected compartments during combination therapy. Nature 1997, 387:188-191.
      Keywords:

      Antiretroviral therapy; combination therapy; clinical trials; viral load; limit of assay quantification; clinical progression

      © 1999 Lippincott Williams & Wilkins, Inc.