Amprenavir in Combination with Lamivudine and Zidovudine Versus Lamivudine and Zidovudine Alone in HIV-1-Infected Antiretroviral-Naive Adults

Home , Amprenavir, Nelfinavir

Antiviral Therapy 5: 215-225 Amprenavir in combination with lamivudine and zidovudine versus lamivudine and zidovudine alone in HIV-1-infected antiretroviral-naive adults

Jeffery C Goodgame1, John C Pottage Jr2*, Helmut Jablonowski3, W David Hardy4, Allan Stein5, Margaret Fischl6, Patrick Morrow7, Judith Feinberg8, Cynthia Hanson Brothers9, Imogen Vafidis10, Pantaleo Nacci10, Jane Yeo10 and Louise Pedneault9 for the Amprenavir PROAB3001 International Study Team

1Central Florida Research Initiative, Altamonte Springs, Fla., USA (currently with Pfizer, Ohio, USA) 2Vertex Pharmaceuticals, Cambridge, Mass., USA 3Universitatsklinik Dusseldorf, Medizinische Klinik & Poliklinik, Dusseldorf, Germany 4Pacific Oaks Research, Beverly Hills, Calif., USA 5Care Resource, Coral Gables, Fla., USA 6University of Miami, AIDS Clinical Trial Unit, Miami, Fla., USA 7ID Associates, Dallas, Tex., USA (Currently with Aesculapius Medical Health Group, Dallas, Tex., USA) 8University of Cincinnati, Holmes Division, Cincinnati, Ohio, USA 9Glaxo Wellcome, Research Triangle Park, N.C., USA 10Glaxo Wellcome, Greenford, UK *Corresponding author: Tel: +1 617 577 6322; Fax: +1 617 577 6501; E-mail: [email protected]

Objectives: To compare the antiviral activity and safety of plasma HIV-1 RNA of <400 copies/ml at 48 weeks. a new protease inhibitor, amprenavir (141W94) in combi- Results: At 48 weeks, a significantly greater proportion of nation with lamivudine and zidovudine, versus lamivudine amprenavir/lamivudine/zidovudine subjects had plasma and zidovudine alone in HIV-1 infected, antiretroviral- HIV-1 RNA levels <400 copies/ml than lamivudine/ zidovu- naive subjects. dine subjects in the overall population: 41 versus 3% Design: Subjects (n=232) with a CD4 T cell count of ≥200 (intent-to-treat missing equals failure analysis) (P<0.001); cells/mm3, plasma HIV-1 RNA levels of ≥10000 copies/ml, 93 versus 42% (as-treated analysis) (P<0.001); and within and ≤4 weeks of prior nucleoside antiretroviral therapy, each of the three randomization strata (P<0.001). Subjects were stratified according to baseline plasma HIV-1 RNA on amprenavir/lamivudine/ zidovudine experienced longer level (10000–30000; 30000–100000; or >100000 time to event (permanent discontinuation of randomized copies/ml). Subjects received double-blind treatment with therapy or viral rebound) than those on lamivudine/ either 1200 mg amprenavir twice daily in combination zidovudine (median of 33 versus 13 weeks; P<0.001). A with lamivudine (150 mg twice daily) and zidovudine (300 significantly greater incidence of drug-related nausea, mg twice daily) (amprenavir/lamivudine/zidovudine) or vomiting, rash and oral/perioral paresthesia was observed matched placebo, lamivudine and zidovudine for 16 weeks. with amprenavir/lamivudine/zidovudine than with lamivu- Thereafter, subjects with confirmed plasma HIV-1 RNA dine/zidovudine. levels of ≥400 copies/ml could add open-label amprenavir Conclusions: Amprenavir, in combination with lamivudine or switch to other antiretrovirals and continue treatment and zidovudine, has potent and durable antiviral activity in for up to a minimum of 48 weeks. The primary endpoint of antiretroviral-naive subjects over 48 weeks. Amprenavir the study was defined as the proportion of subjects with was safe and generally well tolerated.

Introduction

HIV-1 protease inhibitors in combination with other such as zidovudine, lamivudine, didanosine, antiretrovirals have been shown to achieve reduction in zalcitabine or stavudine, have been shown to suppress the levels of plasma HIV-1 RNA to levels below the viral replication in lymphoid tissue [4–8]. Because detection limit of current assays [1–3]. In addition, HIV-1 protease inhibitor-containing regimens have protease inhibitor (PI)-containing regimens including demonstrated potent antiviral activity and a dramatic indinavir, saquinavir, ritonavir or nelfinavir and two reduction in HIV-related deaths and opportunistic nucleoside reverse transcriptase inhibitors (NRTIs), infections [9,10], most HIV/AIDS treatment guidelines

[11–13] currently recommend the consideration of treatment, begin open-label amprenavir therapy, add HIV-1 PIs in the initial treatment of HIV-1 infection. abacavir, add a commercially available HIV protease Amprenavir (141W94; Agenerase), which was struc- inhibitor (except ritonavir), change NRTIs, or switch turally designed based on the 3D structure of the to a commercially available PI (including ritonavir), protease enzyme, is a newly approved and potent and continue study participation for up to a minimum inhibitor of HIV-1 protease (Ki=0.6 nM) [14] and of of 48 weeks (open-label phase). Subjects with HIV-1 HIV-1 replication in vitro (mean IC50=29 nM against RNA levels consistently below 400 copies/ml remained 216 HIV clinical isolates) [15]. Clinical data indicate on blinded, randomized study medication (randomized that amprenavir is well absorbed and can be adminis- phase). tered with or without food, although it should not be taken with a high-fat meal [16]. Twice-daily dosing is Measurements feasible with amprenavir because of its long half-life of Blood was collected at baseline, at weeks 1, 2, 4, 8, 12, between 7 and 10 h [16] compared with that of 1.8–5 16 and every 4 weeks thereafter for analysis of plasma h for currently marketed HIV-1 PIs [17]. Clinical trials HIV-1 RNA levels, absolute CD4 lymphocyte counts, with amprenavir in combination with other antiretro- clinical chemistry and haematology evaluations. Blood virals, such as abacavir, zidovudine, lamivudine, samples for most laboratory evaluations were not indinavir, saquinavir and nelfinavir, are ongoing and collected in the fasting state because subjects were have demonstrated potent antiviral activity, as required to fast overnight only for week 2 and 12 eval- evidenced by decreased plasma HIV-1 RNA levels and uations. Throughout the study period, clinical adverse increased CD4 counts [18–21]. Additionally, ampre- events and haematology or clinical chemistry labora- navir is safe, with most adverse events being transient tory abnormalities were assessed and graded according and mild-to-moderate [19–22]. to ACTG toxicity scales [23] wherever possible (grade This study was designed to evaluate the efficacy and 1 or mild, to grade 4 or very severe). Durability of the safety of amprenavir in combination with lamivudine plasma HIV-1 RNA response over 48 weeks was and zidovudine versus lamivudine and zidovudine in measured as time-to-event. An event occurred when antiretroviral-naive subjects with CD4 counts of ≥200 any one of the following conditions was met: cells/mm3, and to assess the durability of this effect confirmed plasma HIV-1 RNA levels ≥400 copies/ml, over 48 weeks of treatment. However, treatment modi- permanent discontinuation of randomized therapy fications were encouraged from week 16 onwards in (including confirmed plasma HIV-1 RNA level of ≥400 subjects whose HIV-1 RNA levels were not maintained copies/ml), progression to a new confirmed CDC cate- at ≤400 copies/ml. gory C event or death.

Materials and Methods Laboratory methods A reverse transcriptase PCR assay (Amplicor HIV-1 Subjects and study design Monitor, Roche Molecular Systems, Branchburg, N.J., Male or female adults (≥18 years) with a confirmed USA) was used to measure plasma HIV-1 RNA diagnosis of HIV-1 infection gave informed consent to (Primers 1.0 Standard Assay, limit of detection=400 participate in this randomized, placebo-controlled copies/ml; Primers 1.0, Ultrasensitive Assay, limit of study (PROAB3001) which was approved by the detection=50 copies/ml). CD4 lymphocyte counts Institutional Review Board/Ethics Committee at 23 were determined using flow cytometry (FACScalibur, clinical sites, 13 in the USA and 10 in Europe. Subjects Becton Dickinson Three-Color Flow Cytometry, San with CD4 counts of ≥200 cells/mm3, who were anti- Jose, Calif., USA). All efficacy and safety laboratory retroviral-naive (≤4 weeks previous NRTI treatment) measurements were performed at Covance Central and CDC category A or B, were randomly assigned in Laboratories (Geneva, Switzerland; and Indianapolis, a double-blind manner to receive 1200 mg amprenavir Ind., USA). twice daily or matched placebo twice daily without regard to food, in combination with lamivudine (150 Statistical analyses mg twice daily) and zidovudine (300 mg twice daily) A sample size of 115 subjects per treatment group was for 16 weeks. Randomization was stratified according estimated to provide ≥85% power in detecting a 20% to screening plasma HIV-1 RNA (10000–30000, difference between treatment groups at the primary 30000–100000 or >100000 copies/ml). At week 16 endpoint, which was the proportion of subjects with and thereafter, subjects who met the protocol-defined plasma HIV-1 RNA levels of <400 copies/ml at week switch criterion (defined as two consecutive plasma 48. P-values at or below the 0.05 alpha level were HIV-1 RNA values of ≥400 copies/ml measured within considered significant. 3 weeks of one another), could remain on randomized Efficacy variables were analysed on an intent-to-

treat (ITT) basis (which includes data from all random- Table 1. Demographics and baseline characteristics of ized subjects) and on an as-treated basis (which randomized subjects includes data from all subjects who remained on APV/3TC/ZDV 3TC/ZDV randomized treatment only). Two separate ITT Characteristic (n=116) (n=116) analyses were performed. The ITT missing equals Age (years) failure analysis, all premature discontinuations from Median (range) 35 (17–62) 35 (21–62) Gender (%) randomized therapy, new confirmed CDC category C Male 89 89 events, deaths and missing values were considered fail- Female 11 11 ures. To account for subjects who switched therapy but Ethnic origin (%) White 77 73 did not experience virological failure during the open- Black 12 9 label phase, the ITT observed analysis was also Other 11 18 performed, and this analysis included all data CDC disease status (%) Category A (asymptomatic) 78 80 collected, regardless of any treatment switches, with no Category B (symptomatic, not AIDS) 18 20 imputations performed for missing values. The safety Category C (AIDS) 4 0 Plasma HIV-1 RNA (log copies/ml) analysis was based on all subjects exposed to at least 10 Median (range) 4.61 (3.61–6.09) 4.74 (3.06–6.31) one dose of study drug. CD4 cell count (×106 cells/l) ITT missing equals failure and as-treated analyses Median (range) 442 (216–1800) 410 (139–984) were performed for the proportions of subjects with APV, amprenavir (1200 mg bid); 3TC, lamivudine (150 mg bid); ZDV, zidovudine plasma HIV-1 RNA levels of <400 and <50 copies/ml. (300 mg bid); bid = twice daily. The primary analysis was the ITT missing equals failure analysis of the proportion of subjects with HIV- (17%), meeting switch criterion (14%), withdrawal of 1 RNA levels of <400 copies/ml. The time-to-event consent (9%), lost to follow-up (9%), other unspeci- analysis was performed for the ITT population. ITT fied reasons (9%) and protocol violation (<1%). Of the observed analysis was performed for the median 109 subjects who received treatment in the lamivudine/ change from baseline in plasma HIV-1 RNA levels and zidovudine group, 103 subjects (95%) discontinued CD4 cell counts. randomized treatment and the primary reasons for Treatment differences in the proportion of subjects discontinuation included meeting protocol-defined with plasma HIV-1 RNA levels of <400 copies/ml at switch criterion (78%), withdrawal of consent (6%), week 48 were analysed using the Cochran Mantel adverse events (5%), lost to follow-up (4%) and other Haenszel test, controlling for randomization strata, unspecified reasons (4%). Of the subjects who discon- and across randomization strata using Fisher’s exact tinued randomized treatment, 21 subjects (33%) in the test. The distribution of time-to-event data was esti- amprenavir/lamivudine/zidovudine group and 87 mated using the Kaplan–Meier product-limit survival subjects (84%) in the lamivudine/zidovudine group method and treatment differences were assessed using entered the open-label phase of the study. a permutation-based log-rank test. Subjects who never achieved plasma HIV-1 RNA levels of <400 copies/ml Proportion of subjects with plasma HIV-1 RNA levels were considered failures at day 1. Treatment differ- of <400 copies/ml ences in the incidence of adverse events were analysed At week 16, the proportion of subjects with HIV-1 by the Fisher’s exact test. RNA levels of <400 copies/ml was significantly greater in the amprenavir/lamivudine/zidovudine group Results compared with the lamivudine/zidovudine group: 55% (64/116) versus 15% (17/116) by ITT missing equals This study was conducted between February 1997 and failure analysis and 87% (67/77) versus 20% (19/94) November 1998. A total of 232 subjects (116 per treat- by the as-treated analysis. By week 48, the proportion ment group) were randomized into the study. of subjects with HIV-1 RNA of <400 copies/ml was Demographics and baseline characteristics were also significantly greater in the amprenavir/lamivu- comparable in both groups (Table 1). dine/zidovudine group than in the lamivudine/ Of the 232 subjects who were randomized, four zidovudine group: 41% (48/116) versus 3% (4/116) by subjects and seven subjects in the amprenavir/lamivu- ITT missing equals failure analysis (P<0.001) and 93% dine/zidovudine and lamivudine/zidovudine groups, (50/54) versus 42% (5/12) by as-treated analysis respectively, did not initiate treatment. Of the 112 (P<0.001) (Figures 1a and b). A significant difference subjects who received treatment in the amprenavir/ (P<0.001) was observed in each of the three random- lamivudine/zidovudine group, 63 subjects (56%) ization strata at week 48 by the ITT missing equals discontinued randomized treatment and the primary failure analysis: 51% (19/37) versus 8% (3/37) for the reasons for discontinuation included adverse events HIV-1 RNA 10000–30000 copies/ml stratum; 42%

Antiviral Therapy 5:3 217 JC Goodgame et al.

Figure 1. Proportion of subjects with plasma HIV-1 RNA levels of <400 copies/ml (a, b) and 50 copies/ml (c, d) in the amprenavir plus lamivudine plus zidovudine and lamivudine plus zidovudine groups over 48 weeks of treatment

(a) (b) APV/3TC/ZDV 100 3TC/ZDV 100 80

80 60 60

40 40

20 20 centage of subjects with HIV-1 RNA r centage of subjects with HIV-1 RNA <400 copies/ml Pe <400 copies/ml (as-treated analysis) <400 copies/ml (as-treated r 0 0 Pe 04 812162024283236404448 04812162024283236404448 Study week Study week No. of subjects APV/3TC/ZDV 100 91 82 77 73 68 606053505354 3TC/ZDV 102 98 96 94 86 33 261514141312

100

100 APV/3TC/ZDV 80 3TC/ZDV 80 60 60

40 40

20 20 centage of subjects with HIV-1 RNA <50 copies/ml r Pe Subjects with HIV-1 RNA <50 copies/ml 0 0 081624 32 40 48 081624 32 40 48 Study week Study week APV/3TC/ZDV 107 78 68 59 54 52

3TC/ZDV 109 96 30 16 13 11

ITT missing equals failure analysis (a) and as-treated analysis (b). APV, amprenavir; 3TC, lamivudine; ZDV, zidovudine.

(23/55) versus 2% (1/55) for the HIV-1 RNA (Figures 1c and d). 30000–100000 copies/ml stratum; 25% (6/24) versus 0% (0/24) for the HIV-1 RNA >100000 copies/ml Plasma HIV-1 RNA profiles stratum. In both treatment groups, the ITT observed analysis showed median reductions from baseline plasma HIV-

Proportion of subjects with plasma HIV-1 RNA levels 1 RNA of approximately 2 log10 copies/ml by week 4 of <50 copies/ml (Figure 2). These reductions were maintained in the A greater proportion of subjects receiving ampre- amprenavir/lamivudine/zidovudine group through navir/lamivudine/zidovudine had plasma HIV-1 RNA week 48, but diminished after week 4 in the lamivu- of <50 copies/ml at week 48 compared with those dine/zidovudine group. Following therapy switches, receiving lamivudine/zidovudine: 34% (40/116) versus the median reduction from baseline plasma HIV-1 1% (1/116) by ITT missing equals failure analysis and RNA increased in the lamivudine/zidovudine group

79% (41/52) versus 9% (1/11) by as-treated analysis from week 24 (1.8 log10 copies/ml) to week 48 (2.1

Figure 2. Median change from baseline in plasma HIV-1 RNA (log10 copies/ml) through week 48 in the amprenavir plus lamivudine plus zidovudine and lamivudine plus zidovudine groups

Randomized therapy Open-label therapy

-0.5

-1 10

-1.5

-2

-2.5

-3 Median change from baseline in log HIV-1 RNA (copies/ml) Median change from 04812162024283236404448 No. of subjects Study week

APV/3TC/ZDV 116 104 101 101 99 91 93 85 79 77 77 75 74

3TC/ZDV 116 104 100 94 88 81 83717067686971

Error bars are the 25th and 75th percentiles. The values shown for week 16 onwards include subjects with HIV-1 RNA levels of ≥400 copies/ml who had switched to open-label therapy. Plasma HIV-1 RNA limit of detection was 400 copies/ml.

log10 copies/ml). Adverse events The overall incidence of adverse events was signifi- CD4 cell counts cantly greater in the amprenavir/lamivudine/ CD4 cell counts increased in both treatment groups zidovudine group compared with the lamivudine/ throughout the study. At week 48, a median increase of zidovudine group (100 versus 94%; P=0.013). The 128×106 cells/l (range, –237–549) and 125×106 cells/l incidence of drug-related nausea, vomiting, rash and (range, –194–393) was observed in the amprenavir/ oral/perioral paresthesia was significantly higher in the lamivudine/zidovudine and lamivudine/zidovudine amprenavir/lamivudine/zidovudine group compared groups, respectively (Figure 3). with the lamivudine/zidovudine group (P<0.01; Table 2). Most of the adverse events were of mild (ampre- Time-to-event analysis navir/lamivudine/zidovudine 27%; lamivudine/ The survival curves for the two treatment groups (Figure zidovudine 34%) or moderate (amprenavir/lamivu- 4) were significantly different (P<0.001), with longer dine/zidovudine 49%; lamivudine/zidovudine 35%) times-to-event observed in the amprenavir/ intensity. lamivudine/zidovudine group compared with the The most common adverse events that led to study lamivudine/zidovudine group (median time to event of drug discontinuation were nausea (amprenavir/lamivu- 33 versus 20 weeks). By week 48, the proportions of dine/zidovudine 12%; lamivudine/zidovudine <1%), subjects who did not have an event were 47 and 10% in vomiting (amprenavir/lamivudine/zidovudine 5%) and the amprenavir/lamivudine/zidovudine and lamivu- rash (amprenavir/lamivudine/zidovudine 3%; lamivudine/zidovudine groups, respectively. Most events either dine/zidovudine <1%). The median times to onset of occurred at day 1 (amprenavir/lamivudine/zidovudine these adverse events were 4 days for nausea, 33 days 22%; lamivudine/zidovudine 35%) due to subjects never for vomiting and 11 days for rash. The median dura- achieving plasma HIV-1 RNA levels of <400 copies/ml tions were 24 days for nausea, 5 days for vomiting and or never being treated, or at week 20 (amprenavir/ 10 days for rash. In the amprenavir/lamivudine/zidovu- lamivudine/zidovudine 3%; lamivudine/zidovudine dine group, 31 subjects (27%) reported 36 episodes of 31%) due to subjects switching to open-label therapy. rash that did not result in discontinuation of study

Antiviral Therapy 5:3 219 JC Goodgame et al.

Figure 3. Median change from baseline in CD4 cell count (cells/mm3) through week 48 in the amprenavir plus lamivudine plus zidovudine and lamivudine plus zidovudine groups

Randomized therapy Open-label therapy 220 200 180 160 140 120 100 80

3 60

count (ce’’s/mm ) count (ce’’s/mm 40 20 0

Median change from baseline in CD4 cell Median change from –20 –40 –60 Baseline 481216 20 24 28 32 36 40 44 48 Study week No. of subjects (ITT observed analysis) APV/3TC/ZDV 116 104 100 93 8678696667

3TC/ZDV 116 104 103 101 93 88 82 76 71

Figure 4. Kaplan–Meier survival plot of time-to-event

1.0 0.9 0.8 0.7 0.6 0.5 0.4 Probability 0.3 0.2 0.1 0.0 0481216202428323640444852566064687276808488

Time to event (weeks)

Treatment: Placebo Amprenavir

Event defined as plasma HIV-1 RNA level >400 copies/ml, permanent discontinuation of randomized therapy, progression to a new confirmed CDC category C event or death, in the ITT population. Notation: placebo (lamivudine/zidovudine group); amprenavir (amprenavir/lamivudine/zidovudine group).

Table 2. Drug-related adverse events [n (%)] reported by ≥10% of all subjects and grade 3 or 4 adverse events over 48 weeks

All grades* Grade 3 or 4 APV/3TC/ZDV 3TC/ZDV APV/3TC/ZDV 3TC/ZDV Adverse event (n=113) (n=109) (n=113) (n=109) Nausea 84 (74%) 49 (45%) 6 (5%) 2 (2%) Gaseous symptoms 36 (32%) 47 (43%) 0 1 (<1%) Fatigue 37 (33%) 33 (30%) 3 (3%) 2 (2%) Headache 25 (22%) 31 (28%) 0 2 (2%) Vomiting 34 (30%) 15 (14%) 1 (<1%) 0 Diarrhoea 30 (27%) 17 (16%) 2 (2%) 1 (<1%) Paresthesia (oral/perioral) 29 (26%) 7 (6%) 0 0 Rash 28 (25%) 5 (5%) 3 (3%) 0 Loose stools 12 (11%) 16 (15%) 0 0 Abdominal discomfort 11 (10%) 12 (11%) 1 (<1%) 0 Constipation 6 (5%) 14 (13%) 0 0 Taste disorders 11 (10%) 6 (6%) 0 0

APV, amprenavir (1200 mg bid); 3TC, lamivudine (150 mg bid); ZDV, zidovudine (300 mg bid); bid, twice daily. *Data for one placebo subject who inadvertently received APV for 8 weeks are included in both treatment groups. P<0.01 compared with the lamivudine/zidovudine group. drug in the majority of subjects. In 44% of the cases of history of diabetes mellitus after 20 weeks of treatment rash, study medication was temporarily withheld. In and haemolytic anaemia in one subject after 12 weeks only one of these cases did a subject have a recurrence of treatment. One subject randomized to lamivudine/ of rash after re-initiating the study drug. zidovudine who switched on day 285 after initiation of The number of subjects with a grade 3/4 laboratory the study to quadruple therapy with amprenavir/ abnormality is presented in Table 3. The incidence of lamivudine/zidovudine/ abacavir developed a dorsocer- grade 3/4 laboratory abnormalities was low (≤6%) and vical fat accumulation (‘buffalo hump’) on day 424 comparable between treatment groups. Overall, a total after initiation of the study. of 32 subjects (15%) had at least one grade 3/4 laboratory abnormality over the 48-week period (amprenavir/ Discussion lamivudine/zidovudine 14%; lamivudine/zidovudine 15%). The most common clinical chemistry abnor- In antiretroviral-naive HIV-1-infected subjects, a mality was elevated serum transaminases. The most greater proportion of subjects treated with the combi- common haematological abnormality was neutropenia. Cholesterol, triglycerides and glucose levels were Table 3. Number (%) of subjects with grade 3 or 4 not obtained under fasting conditions during the study. treatment-emergent laboratory abnormalities over 48 weeks* No grade 3/4 hypercholesterolaemia was observed in Parameter APV/3TC/ZDV 3TC/ZDV either treatment group. Median baseline cholesterol (n=113) (n=109) values were 167 and 163 mg/dl for the amprenavir/ Clinical chemistry lamivudine/zidovudine and lamivudine/zidovudine ↑ALT 5 (5%) 2 (2%) groups, respectively. Median increases from baseline in ↑AST 4 (4%) 2 (2%) cholesterol of 10.8 mg/dl (week 1) and 23.2 mg/dl ↑ALT and/or AST 6 (6%) 3 (3%) (week 48) were observed in the amprenavir/lamivu- ↑Amylase 1 (<1%) 2 (2%) dine/zidovudine group but no change was observed in ↑Bilirubin 1 (<1%) 1 (<1%) ↑ the lamivudine/zidovudine group. Grade 3/4 hyper- Glucose 1 (<1%) 1 (<1%) ↑ triglyceridaemia was observed in one subject treated Triglycerides 0 1 (<1%) ↑Cholesterol 0 0 with lamivudine/zidovudine during the randomized Haematology phase and in three subjects during the open-label ↓Haemoglobin 1 (<1%) 2 (2%) phase, whereas no grade 3/4 hypertriglyceridaemia was ↓Neutrophil count 6 (5%) 7 (6%) reported in the amprenavir/lamivudine/zidovudine ↓Total WBC 1 (<1%) 1 (<1%) group. Other clinically significant adverse events that were considered to be drug-related in subjects treated Abbreviations: APV, amprenavir (1200 mg BID); 3TC, lamivudine (150 mg BID); with amprenavir/lamivudine/zidovudine included ZDV, zidovudine (300 mg BID); BID, twice daily. *Data for one placebo subject who inadvertently received APV for 8 weeks are grade 4 hyperglycaemia in one subject with a family included in both treatment groups.

Antiviral Therapy 5:3 221 JC Goodgame et al.

nation of amprenavir, zidovudine and lamivudine had copies/ml (53 versus 52%) and <50 copies/ml (49 plasma HIV-1 RNA below detectable levels (<400 and versus 47%) at week 48. <50 copies/ml) than subjects receiving lamivudine and When evaluating HIV-1 RNA response across the zidovudine, after 48 weeks of treatment. Subjects who three randomization strata with different baseline HIV- switched to an open-label amprenavir-containing 1 RNA levels, the proportion of subjects with HIV-1 regimen after 16 weeks demonstrated an immediate RNA <400 copies/ml among those with baseline HIV- and durable response in both plasma HIV-1 RNA 1 RNA levels of >100000 copies/ml was lower levels and CD4 cell count over 48 weeks of therapy. compared with those observed for the other two strata. Subjects in the amprenavir/lamivudine/zidovudine Few subjects (n=24) were in this stratum. Therefore, group had longer times-to-event compared with the these data should be viewed with caution. Larger, lamivudine/zidovudine group. CD4 cell counts controlled studies will be needed to better differentiate increased throughout the study and the increases were and understand the effects of therapy among subjects generally comparable between treatment groups after with varying HIV-1 RNA levels and CD4 cell counts. 48 weeks of treatment. Combination therapy with amprenavir, lamivudine This study had a number of key considerations. and zidovudine was safe and generally well tolerated Firstly, the study was designed with a dual nucleoside over at least 48 weeks. However, since the median regimen for the first 16 weeks. At the time the study duration of exposure to triple therapy was greater than was designed, the dual nucleoside regimen was consid- to double therapy, subjects receiving the triple combi- ered the standard of care for initial antiretroviral nation had a greater probability of experiencing treatment. In view of the emerging data about the adverse events. As with the other currently approved potency and efficacy of triple therapy with two nucle- HIV-1 PIs, gastrointestinal symptoms were the most osides and a PI [24], patients were given the option to common adverse events reported in subjects receiving switch their antiretroviral therapy to include open- amprenavir/lamivudine/zidovudine. The greater label amprenavir and/or switch to other antiretrovirals percentage of subjects discontinuing after 16 weeks of therapy or if they were experiencing amprenavir/lamivudine/zidovudine treatment was virological failure. Most of the patients (84%) chose to primarily attributed to mild-to-moderate (grades 1–2) alter their therapy. Secondly, as demonstrated by Hill et adverse events. al. [25], the efficacy rates as indicated by the propor- The safety profile of amprenavir described in the tion of patients with HIV-1 RNA levels below the limit present study is consistent with that observed in 30 of quantification vary according to the statistical other trials of varying sizes conducted among a total of methods used. Thus, we performed several statistical 1477 diverse group of patients who were treated with analyses in our study to provide a more complete amprenavir in combination with other antiretrovirals understanding of the efficacy of the regimens evalu- [21]. In these trials, most adverse events were mild-to- ated, consistent with the design of our study. moderate, transient and rarely treatment-limiting. The The results of our study are generally consistent most frequently reported adverse events were gastro- with those recently reported in similar protease-naive intestinal in nature. Grade 3/4 laboratory populations receiving PI-containing triple regimens abnormalities, including lipid or glucose abnormalities, [26–28]. Mitsuyasu et al. [26] reported that after 16 and cases of fat redistribution occurred infrequently, weeks of treatment with the soft gelatin capsule formu- although as in other studies with PIs until recently, fat lation of saquinavir and two NRTIs, the proportion of redistribution was not an end-point. subjects with plasma HIV-1 RNA levels below the limit Distinctive and treatment-limiting toxicities have of detection was 67% (<400 copies/ml) and 37% (<50 been associated with the currently marketed protease copies/ml) using an ITT analysis. In the double-blind, inhibitors including nephrolithiasis with indinavir [29], placebo-controlled AVANTI 2 study [27], the indi- diarrhoea with nelfinavir [30], nausea, vomiting and navir/lamivudine/zidovudine regimen was compared to abdominal pain with ritonavir [17], and nausea and the lamivudine/zidovudine regimen, with the former diarrhoea with the soft gelatin capsule formulation of regimen showing a superior response in the proportion saquinavir [17]. In the present study, amprenavir of subjects with plasma HIV-1 RNA levels <500 therapy was mostly associated with mild-to-moderate copies/ml (60 versus 18%) and <20 copies/ml (46 adverse events. Nausea was the most common treat- versus 4%) at week 52 by the ITT (missing equals ment-limiting adverse event. Mild-to-moderate rash failure) analysis. In the open-label START I trial [28], occurred most commonly during the second week of a comparison of the stavudine/didanosine/indinavir therapy, and was almost always maculopapular, with regimen to the indinavir/lamivudine/zidovudine or without pruritis. It was not accompanied by fever regimen indicated comparable results for the propor- and subjects were generally able to continue treatment tion of subjects with plasma HIV-1 RNA levels <500 through the event. Oral/perioral paresthesia was

usually mild and not treatment-limiting. The incidence General Hospital of Athens, Athens, Greece; Johan N. of grade 3/4 adverse events reportedly associated with Brunn, MD, Infeksjonsmedisinsk avdeling, Ulleval the use of PI-containing regimens, such as Sykehus, Norway; Rui Proenca, MD, Servico de hypertriglyceridaemia, hypercholesterolaemia, hyper- Doencas Infecciosas, Hospital Curry Cabral, Lisbon, glycaemia, diabetes mellitus, haemolytic anaemia and Portugal; Melico Silvestre, MD, PhD, Departmento de lipodystrophy, was low. Doencas Infecciosas, Hospitais da Univeridade de Studies have shown that when ritonavir is coadmin- Coimbra, Coimbra, Portugal; Jose Mallolas Masferrer, istered with other PIs the plasma concentrations of the MD, Servicio de Enfermedades Infecciosas, Hospital coadministered PI increase [31–34]. Thus, studies have Clinic i Provincial, Barcelona, Spain; Bonaventura been conducted to study the interaction between riton- Clotet Sala, MD, Hospital Germans Trias i Pujol, avir and amprenavir. Results indicate that ritonavir Barcelona, Spain; Edmund GL Wilkins, MD, causes clinically significant increases in amprenavir Department of Infectious Diseases & Tropical plasma concentrations [35]. The combination of twice Medicine, North Crumpsall General Hospital, daily 600 mg amprenavir with 100 mg ritonavir, or Crumpsall, Manchester, UK; Chris Taylor, MD, once daily 1200 mg amprenavir with 200 mg ritonavir, Directorate of Genitourinary Medicine, The Caldecott was predicted to increase amprenavir trough concen- Center, London, UK; Anton Pozniak, MD, FRCP, trations to levels significantly above the mean IC50 for Kings College Hospital, London, UK. amprenavir- or multiple PI-resistant viruses [36]. These Glaxo Wellcome team members: Judith Millard, dosage regimens as well as the development of the PhD for the clinical conduct of the study; Barbara amprenavir prodrug [37–39] offer the potential for McGarry, MS, Kamlesh Patel, BSc (Hons), and improved potency concurrently with reductions in Michelle Cartland, BS, MAT, for data management. dosing frequency and pill count. Supported by a grant from Glaxo Wellcome The results of this study confirm that amprenavir in combination with NRTIs is potent, successfully References achieving plasma HIV-1 RNA levels <400 copies/ml in 1. Gulick RM, Mellors JW, Havlir D, Eron JJ, Gonzalez C, a significant number of antiretroviral therapy-naive McMahon D, Rickman DD, Valentine FT, Jonas L, subjects. With its potent and durable antiviral effect, Meibohm A, Emini EA, Chodakewitz JA, Deutsch P, Holder D, Schleif WA & Condra JH. Treatment with indi- safety and the convenience of twice-daily dosing navir, zidovudine and lamivudine in adults with human without dietary requirements, amprenavir, when immunodeficiency virus infection and prior antiretroviral therapy. New England Journal of Medicine 1997; administered in combination with other antiretroviral 337:734–739. agents, should prove to be an important therapeutic 2. Saag M, Gersten M, Chang Y, Greenberg SL, Yu G & option for HIV-1-infected patients. Clendeninn NJ. Long term virological and immunological effect of the HIV protease inhibitor Viracept (nelfinavir mesylate) in combination with zidovudine (AZT) and Acknowledgments lamivudine (3TC). Programs and Abstracts of the Infectious Diseases Society of America 35th Annual Meeting, San Francisco, Calif., USA, 13–16 September We thank all of the participating subjects and the 1997. Abstract 221. following individuals for their contributions to the 3. Markowitz M, Cao Y, Hurley A, O’Donovan R, Peterkin J, Andersen B, Smiley L, Keller A, Johnson P & Ho DD. study: Amprenavir PROAB3001 Study Team Triple therapy with AZT, lamivudine, and ritonavir in 12 Investigators and Staff; Harold A Kessler, MD, subjects newly infected with HIV8 Chicago Center for Clinical Research, Chicago, Ill, 7–12 July 1996. Abstract LB.B.6031. USA; Calvin Cohen, MD, Community Research 4. Ruiz L, van Lunzen J, Arno A, Stellbrink H-J, Schneider C, Rull M, Castella E, Ojanguren I, Richman DD, Clotet B, Initiative of New England, Brookline, Mass., USA; Tenner-Racz K & Racz P. Protease inhibitor-containing Steven Davis, MD, ID Associates, Dallas, Tex., USA; regimens compared with nucleoside analogues alone in the suppression of persistent HIV-1 replication in lymphoid Ramon A Torres, MD, AIDS Center, St. Vincent’s tissue. AIDS 1999; 13:F1–F8. Hospital, NY, USA; Danny Lancaster, MD, Methodist 5. Lafeuillade A, Chollet L, Hittinger G, Profizi N, Costes O Hospital, Memphis, Tenn., USA; Susan Swindells, MD, & Poggi C. Residual human immunodeficiency virus type 1 RNA in lymphoid tissue of patients with sustained plasma UNMC, Omaha, N.E., USA; Bruce Rashbaum, MD, RNA of <200 copies/ml. Journal of Infectious Diseases Washington DC, USA; Robert Anderson, BS, Pacific 1998; 177:235–238. Oaks Research, Beverly Hills, Calif., USA; Jarlath 6. Wong JK, Hezareh M, Gunthard HF, Havlir DV, Ignacio CC, Spina CA & Richman DD. Recovery of replication- Black, RN, and Josette Robinson, RAII, University of competent HIV despite prolonged suppression of plasma Cincinnati, Cincinnati, Ohio, USA; Graham Lovell, viremia. Science 1997; 278:1291–1295. Central Florida Research Initiative, Altamonte Springs, 7. Finzi D, Hermankova M & Pierson T. Identification of a reservoir for HIV-1 in patients on highly active antiretro- Fla., USA; Schlomo Staszewski, MD, Klinikum der viral therapy. Science 1997; 278:1295–1300. Johann Wolfgang Goethe-Universitat, Frankfurt, 8. Chun T-W, Stuyver L, Mizell SB, Ehler LA, Mican JM, Germany; Panagiotis Gargalianos-Kakolyris, MD, Baseler M, Lloyd AL, Nowak MA & Fauci AS. Presence of

Antiviral Therapy 5:3 223 JC Goodgame et al.

an inducible HIV-1 latent reservoir during highly active PG & Volberding PA. Antiretroviral therapy for HIV infec- antiretroviral therapy. Proceedings of the National tion in 1996. Journal of the American Medical Association Academy of Sciences, USA 1997; 94:13193–13197. 1996; 276:146–154. 9. Palella FJ, Delaney KM, Moorman AC, Loveless MO, 25. Hill AM. Analysis of HIV-1 clinical trials: statistical magic? Fuhrer J, Satten GA, Aschman DJ & Holmberg SD. Lancet 1999; 353:2061–2064. Declining morbidity and mortality among patients with 26. Mitsuyasu RT, Skolnik PR, Cohen SR, Conway B, Gill MJ, advanced human immunodeficiency virus infection. New Jensen PC, Pulvirenti JJ, Slater LN, Schooley RT, England Journal of Medicine 1998; 338:853–860. Thompson MA, Torres RA & Tsoukas CM. Activity of the 10. Ledergerber B, Egger M, Opravil M, Telenti A, Hirschel B, soft gelatin formulation of saquinavir in combination Battegay M, Vernazza P, Sudre P, Flepp M, Furrer H, therapy in antiretroviral-naïve patients. AIDS 1998; Francioli P & Weber R. Clinical progression and virolog- 12:F103–F109. ical failure on highly active antiretroviral therapy in HIV-1 27. The AVANTI study group. AVANTI 2. Randomized, patients: a prospective cohort study. Lancet 1999; double-blind trial to evaluate the efficacy and safety of 353:863–868. zidovudine plus lamivudine versus zidovudine plus lamivu- 11. CDC: Guidelines for the use of antiretroviral agents in dine plus indinavir in HIV-infected antiretroviral naive HIV-infected adults and adolescents. Morbidity Mortality patients. AIDS 2000; 14:367–374. Weekly Report 1998; 47:43–82. 28. Squires KE. An open-label, randomized, comparative study 12. Gazzard B & Moyle G. 1998 revision to the British HIV of stavudine (d4T)+lamivudine (3TC)+indinavir (IDV) Association guidelines for antiretroviral treatment of HIV versus zidovudine (ZDV)+3TC+IDV in treatment of HIV- seropositive individuals. Lancet 1998; 352:314–316. infected patients: START I. 39th Interscience Conference on 13. Carpenter CCJ, Cooper DA, Fischl MA, Gatell JM, Antimicrobial Agents & Chemotherapy, San Francisco, Gazzard BG, Hammer SM, Hirsch MS, Jacobsen DM, Calif., USA, September 1999. Abstract 506. Katzenstein DA, Montaner JS, Richman DD, Saag MS, 29. Kopp JB, Miller KD, Mican JAM, Feuerstein IM, Vaughan Schechter M, Schooley RT, Thompson MA, Vella S, Yeni E, Baker C, Pannell LK & Falloon J. Crystalluria and PG & Volberding PA. Antiretroviral therapy in adults: urinary tract abnormalities associated with indinavir. updated recommendations of the International AIDS Annals of Internal Medicine 1997; 127:119–125. Society-USA Panel. Journal of the American Medical 30. Moyle GL, Youle M, Higgs C, Monaghan J, Prince W, Association 2000; 283:381–390. Chapman S, Clendeninn N & Nelson MR. Safety, pharma- 14. St Clair M, Millard J, Rooney J, Tisdale M, Parry N, cokinetics, and antiretroviral activity of the potent, specific Sadler BM, Blum MR & Painter G. In vitro activity of human immunodeficiency virus protease inhibitor nelfi- 141W94 (VX-478) in combination with other antiretro- navir: Results of a phase I/II trial and extended follow-up viral agents. Antiviral Research 1996; 29:53–56. in patients infected with human immunodeficiency virus. 15. Data on file. Glaxo Wellcome Inc. Journal of Clinical Pharmacology 1998; 38:736–743. 16. Sadler BH, Hanson CD, Chittick GE, Symonds WT & 31. Hsu A, Granneman GR, Cao G, Carothers L, el-Shourbagy Roskell NS. The safety and pharmacokinetics of ampre- T, Baroldi P, Erdman K, Brown F, Sun E & Leonard JM. navir (141W94), a human immunodeficiency virus (HIV) Pharmacokinetic interactions between two human type 1 protease inhibitor, following oral administration of immunodeficiency virus protease inhibitors, ritonavir and single doses to HIV-infected adults. Antimicrobial Agents saquinavir. Clinical Pharmacology & Therapeutics 1998; & Chemotherapy 1999; 43:1686–1692. 63:453–464. 32. Hsu A, Granneman GR, Cao G, Carothers L, Japour A, El- 17. Flexner C. HIV protease inhibitors. New England Journal Shourbagy T, Dennis S, Berg J, Erdman K, Leonard JM & of Medicine 1998; 338:1281–1291. Sun E. Pharmacokinetic interaction between ritonavir and 18. Kelleher D, Mellors J, Lederman M, Haas D, Cooney E, indinavir in healthy volunteers. Antimicrobial Agents & Horton J, Stanford J & Haubrich R. Activity of abacavir Chemotherapy 1998; 42:2784–2791. (1592, ABC) combined with protease inhibitors (PI) in 33. van Heeswijk RP, Veldkamp AI, Hoetelmans RM, Mulder therapy naive patients. 12th World AIDS Conference, JW, Schreij G, Hsu A, Lange JM, Beijnen JH & Meenhorst Geneva, Switzerland, June–July 1998. Abstract 12210. PL. The steady-state plasma pharmacokinetics of indinavir 19. Bart P-A, Rizzardi GP, Gallant S, Meylan P, Spreen W, alone and in combination with a low dose of ritonavir in McDade H & Pantaleo UK. Combination abacavir twice daily dosing regimens in HIV-1-infected individuals. (1592)/amprenavir (141W94) therapy in HIV-1 infected AIDS 1999; 13:F95–F99. antiretroviral-naive subjects with CD4+ counts >400 µ th 34. Cameron DW, Japour AJ, Xu Y, Hsu A, Mellors J, cells/ l and viral load >5000 copies/ml. 12 World AIDS Farthing C, Cohen C, Poretz D, Markowitz M, Follansbee Conference, Geneva, Switzerland, June 1998. Abstract 286. S, Angel JB, McMahon D, Ho D, Devanarayan V, Rode R, 20. Eron J, Haubrich R, Richman D, Lang W, Tisdale M, Salgo M, Kempf DJ, Granneman R, Leonard JM & Sun E. Myers R, Pagano G & Ro M. Preliminary assessment of Ritonavir and saquinavir combination therapy for the 141W94 in combination with other protease inhibitors. 5th treatment of HIV infection. AIDS 1999; 13:213–224. Conference on Retroviruses and Opportunistic Infections, 35. Sadler BK, Piliero PJ, Preston SL, Lou Y & Stein DS. Chicago, Ill., USA, February 1–5 1998. Abstract 6. Pharmacokinetic (PK) interaction between amprenavir 21. Pedneault L, Fetter A, Hanson C, Wilson J, Nacci P & (APV) and ritonavir (RTV) in HIV-seronegative subjects Millard J. Amprenavir (141W94, APV): Review of overall after multiple, oral dosing. 7th Conference on Retroviruses safety profile. 6th Conference on Retroviruses and and Opportunistic Infections, San Francisco, Calif., USA, Opportunistic Infections, Chicago, Ill., USA. February January 2000. Abstract 77. 1999. Abstract 386. 36. Stein DS, Sadler B & Sale M. Pharmacodynamic effects of 22. Haubrich R, Thompson M, Schooley R, Lang W, Stein A, amprenavir (APV)+ritonavir (RTV) on different viral popu- Sereni D, van der Ende ME, Antunes F, Richman D, lations. 1st International Workshop on Clinical Pagano G, Kahl L, Fetter A, Brown D & Clumeck N. A Pharmacology of HIV Therapy, The Netherlands, March phase II safety and efficacy study of amprenavir in combi- 2000. Abstract 5.2. nation with zidovudine and lamivudine in HIV-infected 37. Falcoz C, Jenkins JM, Bye C, Kenney KB, Studenberg S, patients with limited antiretroviral experience. AIDS 1999; Fuder H & Prince WT. Single-dose safety and pharmacoki- 13:2411–2420. netics of GW433908, a prodrug of amprenavir (141W94, 23. Division of AIDS. Division of AIDS table for grading APV), administered as an oral liquid formulation to severity of adult adverse experiences. Rockville, Md., USA: healthy male subjects. 39th Interscience Conference on National Institute of Allergy and Infectious Diseases, 1996. Antimicrobial Agents and Chemotherapy, San Francisco, 24. Carpenter CJC, Fischl MA, Hammer SM, Hirsch MS, Calif., USA, 1999. Abstract 918. Jacobsen DM, Katzenstein DA, Montaner JSG, Richman 38. Furfine ES, Baker C, Boehlert C, Dahl R, Good S, Huang DD, Saag MS, Schooley RT, Thompson MA, Vella S, Yeni L, Kenney K, Klein R, Kutz S, Patrick M, Reynolds D,

Salisbury J, Searle A, Studenberg S, Todd D, Tung R, 39. Baker C, Chaturvedi PR, Hale MR, Bridson G, Heiser A, Varlashkin P, Wheelan P, Whitesell G, Woolley J & Furfine ES, Spaltenstein A & Tung RD. Discovery of VX- Spaltenstein A. Preclinical development of GW433908/VX- 175/GW433908, a novel, water-soluble prodrug of 175 the water-soluble phosphate prodrug of amprenavir. amprenavir. 39th Interscience Conference on Antimicrobial 39th Interscience Conference on Antimicrobial Agents and Agents and Chemotherapy, San Francisco, Calif., USA, Chemotherapy, San Francisco, Calif., USA, 1999. Abstract 1999. Abstr. 916. 917.

Received 24 January 2000; accepted 9 August 2000

Antiviral Therapy 5:3 225