Abstract

Objectives

We evaluate the efficacy and tolerability of ritonavir-boosted dual protease inhibitor as a nucleoside reverse transcriptase inhibitor-sparing regimen in a prospective open-label randomized pilot trial in antiretroviral-naive patients.

Methods

Thirty patients received fosamprenavir/atazanavir/ritonavir (Group 1) and 31 patients received saquinavir/atazanavir/ritonavir (Group 2). The primary endpoint for efficacy was the rate of early virological success, defined as plasma viral load <50 copies/mL at week 16. The study is registered with ClinicalTrials.gov (NCT00122603).

Results

At baseline, median (range) viral load was 4.8 log10 copies/mL (4.0–5.7) and the median CD4 cell count was 271/mm3 (197–740). Viral load was <50 copies/mL in 12/30 patients [40%, 95% confidence interval (CI) 23%–58%] and 13/31 patients (42%, 95% CI 25%–59%) at week 16 in Groups 1 and 2, respectively. Patients with failing regimens (viral load ≥400 copies/mL at week 16 or ≥50 copies/mL at week 24) were switched to a standard antiretroviral regimen. At week 48, by an intention-to-treat analysis, 23/30 patients (77%) and 26/31 patients (84%) had plasma HIV-1 RNA <50 copies/mL in Groups 1 and 2, respectively. Four patients discontinued treatment for adverse events, all before week 4. No major changes in the protease gene were detected at treatment failure relative to baseline. Baseline viral load <50 000 copies/mL was the only predictor of virological success at week 16.

Conclusions

Ritonavir-boosted dual protease inhibitor regimens targeting only one step of viral replication were insufficient to rapidly suppress plasma HIV RNA to <50 copies/mL in antiretroviral-naive patients with high viral load at baseline.

Introduction

A combination of three antiretroviral drugs is recommended for initial treatment of HIV infection and is usually based on two nucleoside reverse transcriptase inhibitors (NRTIs) plus either one ritonavir-boosted protease inhibitor (PI/r) or one non-nucleoside reverse transcriptase inhibitor (NNRTI).1 NRTIs are a mainstay of highly active antiretroviral therapy (HAART) but can cause troublesome mitochondrial toxicity and lipoatrophy.2 Different strategies were discussed to build a NRTI-sparing regimen, including the use of NNRTI + PI/r,3 PI/r monotherapy4 and dual PI/r regimens. Dual PI/r regimens, such as lopinavir/ritonavir + saquinavir, have been used as salvage or maintenance therapy for patients with reverse transcriptase (RT) mutations or NRTI intolerance.5,6 Nevertheless, at the time our study was designed, the intrinsic viral potency and tolerability of dual PI/r regimens in induction therapy in naive patients was not known. The objective of our study was to assess the possible value of dual PI/r regimens as first-line regimens. We chose to study atazanavir/ritonavir plus fosamprenavir and atazanavir/ritonavir plus saquinavir regimens based on their favourable pharmacokinetics and low risk of drug–drug interactions.7–9

Methods

Patients and study design

An early endpoint at week 16 was selected in our trial as likely to be the most sensitive and discriminative of a potent regimen and limited the consequences of treatment modifications and missing data. Based on published data, specifically those of the ACTG 5095 trial, which provided precise estimates of response rates to three different triple combination regimens in a large population of untreated HIV-positive patients,10 we considered that a minimum of 50% efficacy at week 16 was required while rates >70% could be expected from the most potent regimens.

This open-label trial with two parallel groups was performed in 19 centres in France. The treatments were randomly allocated in order to ensure that the groups were comparable at baseline. Randomization was performed by a centralized procedure. Permuted blocks were used and no stratification was performed.

Patients were eligible if they were aged 18 years or more, had never received antiretroviral therapy, were infected with a virus with no resistance mutations in the RT and no major resistance mutations in the protease genes, and had a CD4 cell count >200/mm3 and a plasma HIV RNA level between 10 000 and 750 000 copies/mL. Patients with an active hepatitis B infection were excluded. Patients with hepatitis C with no major liver dysfunction were allowed to participate in our study.

The study protocol was approved by an institutional ethics review board (Saint-Germain en Laye, France), and all the patients provided their written informed consent. The study is registered with ClinicalTrials.gov (NCT00122603). An ongoing safety monitoring was performed and controlled by a Data and Safety Monitoring Board.

The patients were randomly assigned at a ratio of 1:1 to receive either fosamprenavir (700 mg twice daily) + atazanavir (300 mg once daily)/ritonavir (100 mg twice daily) (Group 1) or saquinavir (1500 mg once daily) + atazanavir (300 mg once daily)/ritonavir (100 mg once daily) (Group 2).

The patients were assessed at baseline, and at weeks 4, 8, 12, 16, 24 and 48, with a physical examination, CD4 cell count, plasma HIV-1 RNA assay, haematological analyses, plasma chemistry and fasting lipid levels.

Plasma HIV-1 RNA was determined at each study visit in a local laboratory to monitor response to therapy. All laboratories participated in a quality control programme. Virological failure was defined as a plasma viral load ≥400 copies/mL at week 16 or ≥50 copies/mL from weeks 24 to 48 on a single measurement. Patients with failing regimens were switched to a standard antiretroviral regimen chosen with their clinician based on genotypic results.

A virological substudy evaluating the kinetics of early virus decay in plasma was performed on 19 patients who agreed to participate (11 in Group 1 and 8 in Group 2), based on viral load values obtained on days 0, 3, 6 and 9.

HIV-1 RNA measures for the study of both the efficacy outcome (weeks 0 and 16) and virus decay were performed centrally in batches on stored plasma samples.

Plasma samples were collected for genotypic resistance testing at baseline and in case of virological failure. Sequencing was performed in each centre belonging to ANRS resistance group Quality Control and all the sequences were then analysed in a central virology laboratory. The RT and protease gene sequences were determined by population sequencing, using the consensus method of the Agence nationale de recherches sur le sida et les hépatites virales (http://www.hivfrenchresistance.org).

Plasma concentrations of amprenavir, saquinavir and atazanavir were determined at weeks 4, 8, 16, 24 and 48 by using HPLC with fluorimetric or UV detection. The lowest limits of quantification were 5 ng/mL (amprenavir), 20 mg/mL (saquinavir), 30 ng/mL (atazanavir) and 30 ng/mL (ritonavir).11,12

Plasma concentrations determined 24 ± 4 h (atazanavir and saquinavir) or 12 ± 2 h after the last drug intake (amprenavir) were interpreted according to the respective efficacy cut-offs for wild-type HIV-1, i.e. 150 ng/mL (atazanavir), 100 ng/mL (saquinavir) and 228 ng/mL (amprenavir).13,14

Adherence to treatment was assessed by self-administered questions on missed doses.

Quality of life was assessed by a self-administered questionnaire (WHOQOL-HIV 31-BREF) that explores six domains.15 For each of these domains, a score ranging from 4 to 20 was calculated with higher values denoting better quality of life.

Primary efficacy endpoint

The primary efficacy endpoint was the rate of early virological responses, defined as plasma HIV RNA level <50 copies/mL at week 16. To replace conventional HAART, we considered that each dual PI/r combination should provide at least a 50% early success rate, using the lower boundary of the one-sided 95% confidence interval (CI) of the observed success rate at week 16.

Samples used to assess the primary endpoint were centralized for HIV RNA assay.

Statistical analysis

Primary efficacy analysis was planned as a test for a single proportion (binomial distribution) performed in each of the two groups. Based on a virological success rate of at least 50% at week 16 (see previous text), 30 subjects per group were required, with a type I error of 5% and 80% power, to test the null hypothesis of <50% efficacy against the alternative hypothesis of at least 72% efficacy. In each of the two groups, the one-sided 95% CI of the observed proportion of patients reaching the efficacy endpoint was calculated and the lower boundary was compared with 50%. Primary efficacy analysis was performed on an intention-to-treat (ITT) basis, and patients who modified the allocated treatment including dosage or were missing were analysed as failures.

After week 16, virological evaluations were analysed by ITT analyses whether the subjects had discontinued study treatments or not, missing data being considered as failures. On-treatment analyses (data were censored at the stop of the study treatment or after virological failure) were also performed.

Baseline characteristics including gender, CD4, HIV-1 RNA, HIV-1 subtype (B versus non-B) and averaged plasma concentrations at weeks 4, 8 and 16 of each study drug (atazanavir, saquinavir and amprenavir) were investigated as potential predictive factors for virological success at week 16. Univariate analysis (Wilcoxon rank sum test or χ2 test) was performed on the basis of on-treatment analysis. Quantitative data were described with median and interquartile range (IQR).

Viral load reduction from baseline to week 16 was estimated using the Kaplan–Meier method to take account of the values censored by the lower limit of detection of the assay.16

The safety analyses were performed on all subjects through the follow-up on an ITT approach ignoring the changes in the allocated treatment. Subsidiary on-treatment analyses (data were censored 30 days after the stop of the study treatment) were also performed.

Kinetics of viral decay using log10-transformed HIV-1 RNA measures at days 3, 6 and 9 was estimated by a mixed linear model to determine the mean slope after day 3 and half-life, e.g. time to reduce by 50% the number of HIV RNA copies per mL of plasma.17

Results

Baseline characteristics and patient disposition

Recruitment took place between January and September 2006. Eighty-nine patients were screened for inclusion in the study, and 61 were randomized, 30 to Group 1 and 31 to Group 2 (Figure 1). Fifty-five patients (90%) completed the study. Baseline characteristics were similar in the two groups (Table 1). The median baseline HIV-1 RNA value was 4.9 log10 copies/mL in Group 1 and 4.6 log10 copies/mL in Group 2 (4.8 log10 copies/mL overall). Nineteen patients (31%) had HIV-1 RNA ≥100 000 copies/mL. The median CD4 cell count was 271/mm3 (265 in Group 1 and 274 in Group 2). Thirty-seven patients (61%) were infected by HIV-1 subtype B [18 (60%) in Group 1 and 19 (61%) in Group 2].

Figure 1

Patient disposition. ATV, atazanavir; FPV, fosamprenavir; SQV, saquinavir; r, ritonavir.

Figure 1

Patient disposition. ATV, atazanavir; FPV, fosamprenavir; SQV, saquinavir; r, ritonavir.

Table 1

Baseline characteristics

 ATV + FPV/r, n = 30 ATV + SQV/r, n = 31 Total, n = 61 
Women 4 (13%) 7 (23%) 11 (18%) 
Agea (years) 43.0 (34.4–46.0) 36.4 (30.5–42.4) 39.2 (32.8–44.0) 
Weighta (kg) 68.5 (61.0–77.0) 67.2 (60.0–74.7) 68.0 (61.0–75.0) 
CDC class    
 A 23 (77%) 24 (77%) 47 (77%) 
 B 7 (23%) 7 (23%) 14 (23%) 
HIV-1 risk factors    
 homosexual 17 (57%) 12 (39%) 29 (48%) 
 heterosexual 11 (37%) 15 (48%) 26 (43%) 
 injection drug use 1 (3%) 3 (10%) 4 (7%) 
 not known 1 (3%) 1 (3%) 2 (3%) 
HIV-1 RNAa (log10 copies/mL) 4.9 (4.6–5.1) 4.6 (4.3–5.0) 4.8 (4.4–5.1) 
 min–max 4.1–5.7 4.0–5.7 4.0–5.7 
 <50 000 copies/mL 10 (33%) 17 (55%) 27 (44%) 
 ≥100 000 copies/mL 11 (37%) 8 (26%) 19 (31%) 
Subtype B 18 (60%) 19 (61%) 37 (61%) 
CD4a (/mm3265 (223–320) 274 (255–312) 271 (243–312) 
 ATV + FPV/r, n = 30 ATV + SQV/r, n = 31 Total, n = 61 
Women 4 (13%) 7 (23%) 11 (18%) 
Agea (years) 43.0 (34.4–46.0) 36.4 (30.5–42.4) 39.2 (32.8–44.0) 
Weighta (kg) 68.5 (61.0–77.0) 67.2 (60.0–74.7) 68.0 (61.0–75.0) 
CDC class    
 A 23 (77%) 24 (77%) 47 (77%) 
 B 7 (23%) 7 (23%) 14 (23%) 
HIV-1 risk factors    
 homosexual 17 (57%) 12 (39%) 29 (48%) 
 heterosexual 11 (37%) 15 (48%) 26 (43%) 
 injection drug use 1 (3%) 3 (10%) 4 (7%) 
 not known 1 (3%) 1 (3%) 2 (3%) 
HIV-1 RNAa (log10 copies/mL) 4.9 (4.6–5.1) 4.6 (4.3–5.0) 4.8 (4.4–5.1) 
 min–max 4.1–5.7 4.0–5.7 4.0–5.7 
 <50 000 copies/mL 10 (33%) 17 (55%) 27 (44%) 
 ≥100 000 copies/mL 11 (37%) 8 (26%) 19 (31%) 
Subtype B 18 (60%) 19 (61%) 37 (61%) 
CD4a (/mm3265 (223–320) 274 (255–312) 271 (243–312) 

ATV, atazanavir; FPV, fosamprenavir; SQV, saquinavir; r, ritonavir.

Data are numbers of patients (%) unless otherwise stated.

aMedian (IQR).

Primary efficacy endpoint at week 16

At week 16, 12/30 patients in Group 1 (40%, 95% CI 23%–58%) and 13/31 patients in Group 2 (42%, 95% CI 25%–59%) had an HIV RNA level <50 copies/mL (Table 2). The estimated falls in viral load in Groups 1 and 2 were 3.1 and 2.9 log10 copies/mL, respectively. Neither of the two dual PI/r regimens thus reached the specified threshold of early efficacy.

Table 2

Primary efficacy endpoint at week 16 (HIV-1 RNA <50 copies/mL)

 ATV + FPV/r, n = 30 ATV + SQV/r, n = 31 Total, n = 61 
Primary endpoint (HIV-1 RNA <50 copies/mL at week 16), 95% CI 12 (40%), 23%–58% 13 (42%), 25%–59% 25 (41%), 29%–53% 
Reasons for failure    
 drug modification before week 16 4 (13%) 4 (13%) 8 (13%) 
  adverse event, n 
  investigator decision, n 
  non-compliance, n — 
 HIV-1 RNA ≥50 copies/mL and <400 copies/mL 11 (37%) 10 (32%) 21 (34%) 
 HIV-1 RNA ≥400 copies/mL 3 (10%) 4 (13%) 7 (11%) 
 ATV + FPV/r, n = 30 ATV + SQV/r, n = 31 Total, n = 61 
Primary endpoint (HIV-1 RNA <50 copies/mL at week 16), 95% CI 12 (40%), 23%–58% 13 (42%), 25%–59% 25 (41%), 29%–53% 
Reasons for failure    
 drug modification before week 16 4 (13%) 4 (13%) 8 (13%) 
  adverse event, n 
  investigator decision, n 
  non-compliance, n — 
 HIV-1 RNA ≥50 copies/mL and <400 copies/mL 11 (37%) 10 (32%) 21 (34%) 
 HIV-1 RNA ≥400 copies/mL 3 (10%) 4 (13%) 7 (11%) 

ATV, atazanavir; FPV, fosamprenavir; SQV, saquinavir; r, ritonavir.

Data are numbers of patients (%) unless otherwise stated.

Missing data and treatment modifications equal failure.

Therapy was modified before week 16 in eight patients, owing to adverse events (two patients in each group), to poor compliance (one patient in Group 1) or physician decision (one patient in Group 1 and two patients in Group 2).

Eight patients with viral load ≥400 copies/mL at week 16 were switched based on genotypic results and physician decision to a standard triple antiretroviral treatment, as required by the study protocol.

Predictors of virological efficacy

In the on-treatment analysis, the following parameters were not associated with the virological response at week 16: gender (P = 0.71), the CD4 cell count at baseline (P = 0.87), the HIV subtype (B versus non-B, P = 0.61), the treatment group (P = 0.88) and the averaged plasma concentrations of atazanavir (n = 41, P = 0.47), amprenavir (n = 23, P = 1.0) and saquinavir (n = 19, P = 0.9). The only factor predictive of virological success was baseline viral load (median 4.3 log10 copies/mL in responders, 4.8 log10 copies/mL in non-responders; P = 0.0003). In a post hoc analysis, using a plasma viral load cut-off of 50 000 copies/mL close to the median baseline value, virological response was obtained at week 16 in 21/33 patients (64%, CI 95% 47%–80%) with baseline viral load <50 000 copies/mL and in only 3/19 patients (16%) with baseline viral load ≥50 000 copies/mL (P = 0.0009). This figure becomes 23/44 (52%) in patients with a baseline HIV-1 RNA <100 000 copies/mL compared with 1/8 (13%) in patients with baseline HIV-1 RNA ≥100 000 copies/mL.

Early viral kinetics

For the 19 patients participating in the kinetics of virus decay substudy, the median change in viral load from day 0 to day 3 was −0.1 log10 copies/mL and from day 0 to day 9 was −1.4 log10 copies/mL. At day 9, the estimated mean slope from day 3 to day 9 was −0.2 log10 copies/mL per day (95% CI −0.23 to −0.18), with a half-life of 1.5 days and a standard deviation of 0.14 (inter-subjects coefficient of variation of 9%). No link was found between the predicted individual mean slope and the virological response at week 16. Median (IQR) of the predicted individual mean slope was −0.21 (−0.22 to −0.20) in responders and −0.20 (−0.21 to −0.19) log10 copies/mL in non-responders.

Follow-up from weeks 16 to 48

Virological responses by ITT and on-treatment analyses at weeks 24 and 48 are reported in Table 3. At week 48, 23/30 patients (77%) and 26/31 patients (84%) achieved plasma HIV-1 RNA <50 copies/mL in Groups 1 and 2, respectively.

Table 3

HIV-1 RNA at weeks 24 and 48 by ITT and on-treatment analyses

 ATV + FPV/r ATV + SQV/r Total 
Week 24    
 ITTa 30 31 61 
  HIV-1 RNA <50 copies/mL 22 (73%) 19 (61%) 41 (67%) 
  HIV-1 RNA <400 copies/mL 25 (83%) 27 (87%) 52 (85%) 
 on-treatmentb 22 21 43 
  HIV-1 RNA <50 copies/mL 18 (82%) 16 (76%) 34 (79%) 
  HIV-1 RNA <400 copies/mL 21 (95%) 20 (95%) 41 (95%) 
Week 48    
 ITTa 30 31 61 
  HIV-1 RNA <50 copies/mL 23 (77%) 26 (84%) 49 (80%) 
  HIV-1 RNA <400 copies/mL 25 (83%) 27 (87%) 52 (85%) 
 on-treatmentb 16 15 31c 
  HIV-1 RNA <50 copies/mL 15 (94%) 15 (100%) 30 (97%) 
  HIV-1 RNA <400 copies/mL 16 (100%) 15 (100%) 31 (100%) 
 ATV + FPV/r ATV + SQV/r Total 
Week 24    
 ITTa 30 31 61 
  HIV-1 RNA <50 copies/mL 22 (73%) 19 (61%) 41 (67%) 
  HIV-1 RNA <400 copies/mL 25 (83%) 27 (87%) 52 (85%) 
 on-treatmentb 22 21 43 
  HIV-1 RNA <50 copies/mL 18 (82%) 16 (76%) 34 (79%) 
  HIV-1 RNA <400 copies/mL 21 (95%) 20 (95%) 41 (95%) 
Week 48    
 ITTa 30 31 61 
  HIV-1 RNA <50 copies/mL 23 (77%) 26 (84%) 49 (80%) 
  HIV-1 RNA <400 copies/mL 25 (83%) 27 (87%) 52 (85%) 
 on-treatmentb 16 15 31c 
  HIV-1 RNA <50 copies/mL 15 (94%) 15 (100%) 30 (97%) 
  HIV-1 RNA <400 copies/mL 16 (100%) 15 (100%) 31 (100%) 

ATV, atazanavir; FPV, fosamprenavir; SQV, saquinavir; r, ritonavir.

aMissing data equal failure.

bData were censored at the study treatment discontinuation or at time of virological failure.

cOne additional patient in Group 1 missed the week 48 visit but was still under trial treatment.

For patients with baseline viral load <50 000 copies/mL, 21/27 (78%) and 23/27 (85%) had HIV-1 RNA <50 copies/mL at weeks 24 and 48, respectively (ITT analysis). Using on-treatment analysis, 19/21 (90%) and 19/19 (100%) had HIV-1 RNA <50 copies/mL, at weeks 24 and 48, respectively.

Among the 61 patients, 29 discontinued their initial dual PI/r therapy; 13 changed for two NRTIs and a PI/r and 9 for two NRTIs and efavirenz, 4 had increased dose of saquinavir, 1 was lost to follow-up, 1 had withdrawn his consent and 1 had stopped treatment.

At week 48, 32 patients remained on their initial dual PI/r therapy. Patients who reached HIV-1 RNA <50 copies/mL at week 16 (26 patients) kept viral load <50 copies/mL until week 48 except for two patients (one had 69 copies/mL and another had discontinued the study treatments).

Genotyping results

At baseline, 24 (39%) patients were infected by HIV-1 subtype non-B strains [CRF02_AG (n = 10), CRF01_AE (n = 3), A1 (n = 4), CRF06_cpx (n = 2), G (n = 1), J (n = 1), CRF11_cpx (n = 1), CRF18_cpx (n = 1) and unknown (n = 1)]. No major mutations associated with resistance to PIs (IAS-USA expert list of mutations, http://www.iasusa.org) were detected at baseline. The median (IQR) number of minor PI mutations was 3 (2–4), and was similar in the two groups.

The virus was genotyped at the time of virological failure in 21 of the 22 patients concerned (one sample was not amplifiable). Compared with baseline, no major PI resistance mutations were detected. Nine patients had viruses with minor PI resistance mutations (IAS-USA 2008) and other substitutions with an unknown influence on PI efficacy (L10I + V11I, M36I + K43R, A71V + I72V, V62I, L63P, I10V + K57R, V11I + D35N, L10I + K14R + E16G + L23I+I72V and I93L in one case each). Six patients had substitutions not listed in IAS-USA 2008 (W6G, V75I, K21E + S63L, D35E, V15I + K41R + H63L and L10P + L19I in one case each). The six remaining patients had no changes in the protease gene between baseline and virological failure.

Immunological response

The median (IQR) increase in the CD4 cell count at week 16 relative to baseline was 122/mm3 (27–190) in the ITT analysis and 120/mm3 (52–187) in the on-treatment analysis. At week 48, the median (IQR) increase from baseline was 203 cells/mm3 (112–254) in the ITT analysis and 224 cells/mm3 (160–255) in the on-treatment analysis.

Adverse events

The most frequent adverse events (grade ≥2) were gastrointestinal disorders (23% of patients, 30% in Group 1 and 16% in Group 2), including diarrhoea (13%), abdominal pain (5%), nausea (5%) and vomiting (3%). Eight patients (13%) had a skin rash, mainly in Group 1 (7/8).

Eighteen patients (30%) experienced at least a grade 3 or 4 adverse event, most of which were hyperbilirubinaemia [10 patients (16%); 3 (10%) in Group 1 and 7 (23%) in Group 2]. Two patients experienced hepatic cytolysis (Group 2). One patient (Group 1) developed acute cholecystitis after 4 weeks of treatment and underwent cholecystectomy. This patient continued on the same treatment and had plasma HIV-1 RNA <50 copies/mL at week 48.

Four patients (two in each group) discontinued treatment for adverse effects: two for grade 3 skin rash, one for dizziness and tachycardia, and one for grade 3 hepatic cytolysis. All adverse events leading to study drug discontinuation occurred before week 4.

At week 16, bilirubin was higher in Group 2 than in Group 1 (42 versus 26 mM, P = 0.007), and triglyceride and LDL cholesterol in Group 1 than in Group 2 (1.4 versus 1.1 mM, P = 0.036 and 3.7 versus 2.8 mM, P = 0.026).

At week 48, the lipid changes between groups remained statistically different. For bilirubin, the difference in change between groups was not significant (0.9 versus 1.3, P = 0.23).

Pharmacokinetics

Plasma samples taken within 24 ± 4 h (atazanavir and saquinavir) or 12 ± 2 h (amprenavir) after the last drug intake were available for at least one visit in 46 patients. No deleterious pharmacokinetic interactions were observed between atazanavir and amprenavir/ritonavir or between atazanavir and saquinavir/ritonavir. Potentially effective trough plasma concentrations were obtained with boosted atazanavir, amprenavir and saquinavir (Figure 2). The median (IQR) of averaged atazanavir plasma concentrations was 616 (367–899) ng/mL in Group 1 and 742 (528–1042) in Group 2. The medians (IQRs) of averaged plasma saquinavir and amprenavir concentrations, respectively, were 164 (112–365) and 1568 (1210–2079) ng/mL. The intra-individual coefficients of variation of these concentrations were 16% for atazanavir, 22% for saquinavir and 5% for amprenavir.

Figure 2

Cumulative data of atazanavir (ATV) (24 ± 4 h), amprenavir (APV) (12 ± 2 h) and saquinavir (SQV) (24 ± 4 h) plasma concentrations at weeks 4, 8, 16, 24 and 48.

Figure 2

Cumulative data of atazanavir (ATV) (24 ± 4 h), amprenavir (APV) (12 ± 2 h) and saquinavir (SQV) (24 ± 4 h) plasma concentrations at weeks 4, 8, 16, 24 and 48.

No relationship was observed between PI trough plasma concentrations at week 4 and the virological response at week 16.

Adherence

In the on-treatment analysis at weeks 4, 8 and 16, 51/55 (93%), 47/52 (90%) and 34/41 (83%) patients reported no missed doses during the previous 4 days and 52/57 (91%), 52/53 (98%) and 49/49 (100%) patients reported no missed doses during the previous 4 weeks, respectively.

Quality of life

At baseline, median scores in the WHOQOL-HIV 31-BREF questionnaire were high in each domain: a median of 16 for physical, 15 for psychological, 17 for level of independence, 14 for social relationships, 14 for environment and 12 for spirituality.

At weeks 16 and 48, by on-treatment analysis, the median changes in scores from baseline were low, ranging from 0 to 1 for each domain. Results were similar between treatment groups, and there were no differences in changes according to the virological response at week 16.

Discussion

This clinical randomized pilot study showed that dual PI/r alone with atazanavir/ritonavir plus saquinavir or fosamprenavir are insufficient to provide a high rate of early virological success in antiretroviral-naive HIV-1-infected patients. Neither of these two regimens reached a success rate of at least 50% that would have qualified them as potent antiretroviral regimens for initial therapy according to our pre-specified criterion. These results were confirmed by two other studies evaluating dual PI/r in naive patients. The first reported a slow viral decline with a different dosing of saquinavir plus lopinavir/ritonavir;6 the second was terminated prematurely because the treatment lacked the potency of atazanavir plus lopinavir/ritonavir compared with a lopinavir/ritonavir-containing triple-drug regimen.18

An early efficacy endpoint was selected in our trial as likely to be the most sensitive and discriminative of a potent regimen. We considered that a minimum of 50% efficacy at week 16 was required while rates >70% could be expected from the most potent regimens.10 Recent data from ACTG 5142 on two NRTIs plus efavirenz or lopinavir/ritonavir, as well as other studies with NRTIs plus atazanavir/ritonavir, fosamprenavir/ritonavir or darunavir/ritonavir, confirmed a virological efficacy of at least 55%–60% at week 16.3,19–21 Indeed, in our study, all patients with an early response at week 16 (<50 copies/mL), except one, maintained plasma HIV RNA <50 copies/mL through week 48.

Obviously, limitations of our study include the lack of a triple-drug group and the small number of patients included.

Significant predictors of virological efficacy of dual PI/r regimens were previously described during salvage therapy and included the CD4 cell count and viral load at baseline, past antiretroviral exposure and previous PI resistance.6,22

The poor virological results obtained in this study with naive patients were not due to problems of tolerability or compliance. Only four patients discontinued therapy because of clinical or biological adverse events, and adherence to both regimens was good. Trough plasma concentrations of boosted atazanavir, amprenavir and saquinavir were adequate, and PI trough plasma concentrations at week 4 did not influence the virological response at week 16. Antagonism has been observed in vitro between saquinavir and other PIs on wild-type HIV and on saquinavir-resistant variants.23 We cannot exclude a molecular antagonism with atazanavir or fosamprenavir as this has not been described so far. The low penetration of these PIs in sanctuaries where production of viruses can occur might also explain some of the failures observed.

HIV-1 subtype (B or non-B) could not explain the poor virological efficacy at week 16. PIs are designed to fit the active site of the HIV-1 protease and are thus sensitive to structural changes in the viral protein. It was recently reported that the virological response to tipranavir/ritonavir- or lopinavir/ritonavir-containing regimens might be compromised in patients with HIV-1 non-clade B infection.24,25 As stated above, however, subtype non-B infection was not associated with poorer virological outcome at week 16 in our patients.

We found no link between the predicted individual mean slopes of early viral load and the subsequent virological response at week 16. Decline in viral load on day 9 (−1.4 log10 copies/mL) was similar to that obtained with two NRTIs + efavirenz on day 7 (−1.49 log10 copies/mL) in a substudy of the ACTG 5142 trial.26 In our setting of first-line therapy, we demonstrated that a low baseline viral load was the only significant predictor for the early virological response. Indeed, in a post hoc analysis, patients with a baseline plasma HIV-1 RNA <50 000 copies/mL achieved a 64% rate of virological responses at week 16 much closer to the expected value.

This difference in antiretroviral power notably affects the response rate in patients with higher levels of HIV RNA.

The most probable reason for the limited efficacy observed in our trial is that dual PI/r regimens target only one step of viral replication, in the same way as triple NRTI and single PI regimens, which has consistently been insufficient to rapidly suppress plasma HIV RNA to <50 copies/mL in antiretroviral-naive patients with high viral load at baseline.4,10,27,28

Furthermore, although direct comparisons with single PI regimens are lacking in our study, our data do not suggest that dual boosted PI combinations are more potent than single PI regimens.

Finally, the recent availability of new drugs, such as integrase inhibitors, second-generation NNRTIs, new PIs or CCR5 inhibitors, which bring more options to build nucleoside analogue-free combinations, further undermines the potential value of dual PI/r regimens.

Funding

This work was supported by grants from the Agence nationale de recherches sur le sida et les hépatites virales (ANRS).

Transparency declarations

None to declare.

Author contributions: R. L. was the principal investigator of the trial. D. D. and F. B.-V. were responsible for the virological results and interpretation. G. Peytavin was responsible for the pharmacological results and interpretation. R. L., D. D., J.-P. A., G. Peytavin and P. Y. were responsible for the conception and design of the study. R. L., G. Pialoux, M. B., C. K. and P. Y. were responsible for the recruitment of patients and supervision of the trials in their clinical centre. J.-P. A., C. Capitant and C. Chazallon provided statistical expertise and performed analysis. R. L., D. D., C. Capitant, C. Chazallon, J.-P. A., G. Peytavin, G. Pialoux, M. B., C. K., P. Y. and F. B.-V. provided interpretation of the data, critical revision of the article for important intellectual content and final approval of the article.

Acknowledgements

We thank the patients who participated in the study. The drugs were supplied by Bristol-Myers Squibb, GlaxoSmithKline and Roche.

Members of the ANRS 127 study group, ANRS, Paris, France.

Scientific Committee: P. Y., R. L., F. B.-V., D. D., G. Peytavin, M. B., C. Piketty, J.-P. A., C. Capitant, C. Chazallon, M. J. Commoy, Y. Bennai, B. Hadacek and D. Merah.

Participating Clinical Departments: Groupe Hospitalier Bichat, Paris (R. L. and G. Fraqueiro), Groupe Hospitalier Pitié-Salpétrière, Paris (C. K. and M. Pauchard), Hôpital Tenon, Paris (G. Pialoux and C. Fontaine), Centre Hospitalier de la Région Annecienne, Annecy (C. Michon and M. Bensalem), Centre Hospitalier Départemental, La Roche sur Yon (P. Perre and I. Suaud), Hôpital Saint-Louis, Paris (J. M. Molina and A. Rachline), Hôpital de Bicêtre, Le Kremlin-Bicêtre (C. Goujard and M. Môle), Hôpital Pontchaillou—CHRU, Rennes (C. Arvieux and M. Ratajczak), Hôpital Avicenne, Bobigny (M. B. and F. Touam), Hôpital Saint-Antoine, Paris (P. M. Girard and J. L. Lagneau), CHU Hôtel Dieu, Nantes (F. Raffi and H. Hue), Hôpital Foch, Suresnes (D. Zucman), Hôpital Raymond Poincaré-Vidal, Garches (P. de Truchis and H. Berthe), Hôpital Henri Mondor, Créteil (Y. Levy and C. Jung), Hôpital Européen Georges Pompidou, Paris (C. Piketty), CHU Hôpital Gui de Chauliac, Montpellier (J. Reynes and M. Vidal), Hôpital Saint-Louis, Paris (D. Sereni) and Hôpital Louis Mourier, Colombes (M. Bloch).

Data and Safety Monitoring Board: M. L. Chaix, P. Flandre and Y. Yazdanpanah.

Coordinating Centre: INSERM SC10, Villejuif—J.-P. A., C. Capitant, N. Leturque, E. Netzer, V. Foubert and C. Chazallon (statistics), and S. Izard and A. Arulananthan (data management).

References

1
Hammer
SM
Eron
JJ
Jr
Reiss
P
, et al.  . 
Antiretroviral treatment of adult HIV infection: 2008 recommendations of the International AIDS Society-USA panel
JAMA
 , 
2008
, vol. 
300
 (pg. 
555
-
70
)
2
Caron
M
Auclairt
M
Vissian
A
, et al.  . 
Contribution of mitochondrial dysfunction and oxidative stress to cellular premature senescence induced by antiretroviral thymidine analogues
Antivir Ther
 , 
2008
, vol. 
13
 (pg. 
27
-
38
)
3
Riddler
SA
Haubrich
R
DiRienzo
AG
, et al.  . 
Class-sparing regimens for initial treatment of HIV-1 infection
N Engl J Med
 , 
2008
, vol. 
358
 (pg. 
2095
-
106
)
4
Delfraissy
JF
Flandre
P
Delaugerre
C
, et al.  . 
Lopinavir/ritonavir monotherapy or plus zidovudine and lamivudine in antiretroviral-naive HIV-infected patients
AIDS
 , 
2008
, vol. 
22
 (pg. 
385
-
93
)
5
Cameron
DW
Becker
S
King
MS
, et al.  . 
Exploratory study comparing the metabolic toxicities of a lopinavir/ritonavir plus saquinavir dual protease inhibitor regimen versus a lopinavir/ritonavir plus zidovudine/lamivudine nucleoside regimen
J Antimicrob Chemother
 , 
2007
, vol. 
59
 (pg. 
957
-
63
)
6
Staszewski
S
Babacan
E
Stephan
C
, et al.  . 
The LOPSAQ study: 48 week analysis of a boosted double protease inhibitor regimen containing lopinavir/ritonavir plus saquinavir without additional antiretroviral therapy
J Antimicrob Chemother
 , 
2006
, vol. 
58
 (pg. 
1024
-
30
)
7
Winston
A
Mallon
P
Satchell
C
, et al.  . 
The pharmacokinetic profile and safety of saquinavir–ritonavir administered once daily with atazanavir or twice daily with a nucleoside backbone using the saquinavir-500 mg formulation in HIV-1 infected subjects
Abstracts of the Seventh International Workshop on Clinical Pharmacology of HIV Therapy, Lisbon, Portugal, 2006
  
Abstract 64
8
Ford
J
Boffito
M
Maitland
D
, et al.  . 
Influence of atazanavir 200 mg on the intracellular and plasma pharmacokinetics of saquinavir and ritonavir 1600/100 mg administered once daily in HIV-infected patients
J Antimicrob Chemother
 , 
2006
, vol. 
58
 (pg. 
1009
-
16
)
9
von Hentig
N
Muller
A
Haberl
A
The ATSAQ-1 cohort study: pharmacokinetic interactions of atazanavir (AZV) and saquinavir (SQV) in a ritonavir (RTV) boosted protease inhibitor therapy regimen
Abstracts of the Fifteenth International AIDS Conference, Bangkok, Thailand, 2004
  
Abstract WeOrB1235
10
Gulick
RM
Ribaudo
HJ
Shikuma
CM
, et al.  . 
Triple-nucleoside regimens versus efavirenz-containing regimens for the initial treatment of HIV-1 infection
N Engl J Med
 , 
2004
, vol. 
350
 (pg. 
1850
-
61
)
11
Sadler
BM
Hanson
CD
Chittick
GE
, et al.  . 
Safety and pharmacokinetics of amprenavir (141W94), a human immunodeficiency virus (HIV) type 1 protease inhibitor, following oral administration of single doses to HIV-infected adults
Antimicrob Agents Chemother
 , 
1999
, vol. 
43
 (pg. 
1686
-
92
)
12
Marsh
KC
Eiden
E
McDonald
E
Determination of ritonavir, a new HIV protease inhibitor, in biological samples using reversed-phase high-performance liquid chromatography
J Chromatogr B Biomed Sci Appl
 , 
1997
, vol. 
704
 (pg. 
307
-
13
)
13
Sadler
BM
Gillotin
C
Lou
Y
, et al.  . 
Pharmacokinetic study of human immunodeficiency virus protease inhibitors used in combination with amprenavir
Antimicrob Agents Chemother
 , 
2001
, vol. 
45
 (pg. 
3663
-
8
)
14
Mouroux
M
Yvon-Groussin
A
Peytavin
G
, et al.  . 
Early virological failure in naive human immunodeficiency virus patients receiving saquinavir (soft gel capsule)–stavudine–zalcitabine (MIKADO trial) is not associated with mutations conferring viral resistance
J Clin Microbiol
 , 
2000
, vol. 
38
 (pg. 
2726
-
30
)
15
WHOQOL-HIV Group
WHOQOL-HIV for quality of life assessment among people living with HIV and AIDS: results from the field test
AIDS Care
 , 
2004
, vol. 
16
 (pg. 
882
-
9
)
16
Flandre
P
Durier
C
Descamps
D
, et al.  . 
On the use of magnitude of reduction in HIV-1 RNA in clinical trials: statistical analysis and potential biases
J Acquir Immune Defic Syndr
 , 
2002
, vol. 
30
 (pg. 
59
-
64
)
17
Cnaan
A
Laird
NM
Slasor
P
Using the general linear mixed model to analyse unbalanced repeated measures and longitudinal data
Stat Med
 , 
1997
, vol. 
16
 (pg. 
2349
-
80
)
18
Ulbricht
KU
Stoll
M
Behrens
G
, et al.  . 
Double-protease inhibitor, RTI-sparing therapy regimen in naive HIV-1-infected patients: 24-week virologic response analysis of the LORAN trial
Abstracts of the Fifteenth Conference on Retroviruses and Opportunistic Infections, Boston, MA, USA, 2008
 
Alexandria, VA, USA
Foundation for Retrovirology and Human Health
 
Abstract 780
19
Molina
JM
Andrade-Villanueva
J
Echevarria
J
, et al.  . 
Once-daily atazanavir/ritonavir versus twice-daily lopinavir/ritonavir, each in combination with tenofovir and emtricitabine, for management of antiretroviral-naive HIV-1-infected patients: 48 week efficacy and safety results of the CASTLE study
Lancet
 , 
2008
, vol. 
372
 (pg. 
646
-
55
)
20
Eron
J
Jr
Yeni
P
Gathe
J
Jr
, et al.  . 
The KLEAN study of fosamprenavir–ritonavir versus lopinavir–ritonavir, each in combination with abacavir–lamivudine, for initial treatment of HIV infection over 48 weeks: a randomised non-inferiority trial
Lancet
 , 
2006
, vol. 
368
 (pg. 
476
-
82
)
21
Ortiz
R
Dejesus
E
Khanlou
H
, et al.  . 
Efficacy and safety of once-daily darunavir/ritonavir versus lopinavir/ritonavir in treatment-naive HIV-1-infected patients at week 48
AIDS
 , 
2008
, vol. 
22
 (pg. 
1389
-
97
)
22
von Hentig
N
Babacan
E
Staszewski
S
, et al.  . 
Predictive factors for response to a boosted dual HIV-protease inhibitor therapy with saquinavir and lopinavir in extensively pre-treated patients
Antivir Ther
 , 
2007
, vol. 
12
 (pg. 
1237
-
46
)
23
Dam
E
Lebel-Binay
S
Rochas
S
, et al.  . 
Synergistic inhibition of protease-inhibitor-resistant HIV type 1 by saquinavir in combination with atazanavir or lopinavir
Antivir Ther
 , 
2007
, vol. 
12
 (pg. 
371
-
80
)
24
Flandre
P
Delaugerre
C
Ghosn
J
, et al.  . 
Prognostic factors of virological success in antiretroviral-naive patients receiving LPV/r monotherapy in the MONARK trial
Abstracts of the Eleventh European AIDS Conference, Madrid, 2007
  
Abstract PS1.2
25
Marcelin
AG
Masquelier
B
Descamps
D
, et al.  . 
Tipranavir–ritonavir genotypic resistance score in protease inhibitor-experienced patients
Antimicrob Agents Chemother
 , 
2008
, vol. 
52
 (pg. 
3237
-
43
)
26
Haubrich
RH
Riddler
S
Ribaudo
H
, et al.  . 
Initial viral decay to assess the relative antiretroviral potency of PI-, NNRTI-, and NRTI-sparing regimens for first line therapy of HIV-1 infection: ACTG 5160s (sub-study of A5142)
Abstracts of the Fourteenth Conference on Retroviruses and Opportunistic Infections, Los Angeles, CA, USA
 
Alexandria, VA, USA
Foundation for Retrovirology and Human Health
 
Abstract 137
27
Khanlou
H
Yeh
V
Guyer
B
, et al.  . 
Early virologic failure in a pilot study evaluating the efficacy of therapy containing once-daily abacavir, lamivudine, and tenofovir DF in treatment-naive HIV-infected patients
AIDS Patient Care STDS
 , 
2005
, vol. 
19
 (pg. 
135
-
40
)
28
Landman
R
Descamps
D
Peytavin
G
, et al.  . 
Early virologic failure and rescue therapy of tenofovir, abacavir, and lamivudine for initial treatment of HIV-1 infection: TONUS study
HIV Clin Trials
 , 
2005
, vol. 
6
 (pg. 
291
-
301
)

Author notes

Members of the ANRS 127 study group are listed in the Acknowledgements section.