Twice-daily amprenavir 1200 mg versus amprenavir 600 mg/ritonavir 100 mg, in combination with at least 2 other antiretroviral drugs, in HIV-1-infected patients

Male and non-pregnant, non-lactating female outpatients were eligible for study enrollment if they were at least 18 years of age; had HIV-1 infection documented by HIV-1 antibody enzyme-linked immunosorbent assay (ELISA) and confirmed by Western blot test, positive HIV-1 culture, positive HIV-1 serum antigen, or plasma viremia; and CD4+ cell counts ≥ 50/mm³. Women of childbearing potential had to have a negative serum β-human chorionic gonadotropin (HCG) pregnancy test at screening, and had to be willing to use an adequate method of contraception during the study. Patients could be either antiretroviral-naïve or -experienced. If patients were antiretroviral-experienced, they had to be naïve to APV, have a plasma HIV-1 RNA >1000 copies/mL, show susceptibility to APV and ≥ 2 other antiretroviral drugs, and remain on their most recent treatment regimen until completion of the screening visit. For all drugs, susceptibility was defined as an HIV isolate with <4.0-fold change in IC₅₀ (determined by VIRCOGEN™ [Tibotec-Virco NV, Mechelen, Belgium], a virtual phenotype assay) in comparison to the control virus. Patients were excluded if they had active Centers for Disease Control (CDC) Class C status; could not comply with the study schedule; were in another investigational drug study; were undergoing opiate detoxification; had a malabsorption syndrome or pre-existing condition that interfered with normal gastrointestinal transit; had clinically significant laboratory abnormalities, required radiation therapy or cytotoxic chemotherapy, or had received immunomodulating agents, within 4 weeks pre-study; or had received an HIV-1 immunotherapeutic vaccine within 3 months pre-study. Patients were allowed to take pravastatin, fluvastatin, cerivastatin, or atorvastatin for hyperlipidemia at the cautionary discretion of the investigators, but were not allowed to take lovastatin, simvastatin, or any other drug known to significantly affect metabolism by the cytochrome P450 CYP3A4 enzyme system.

In this multicenter, open-label clinical trial, patients were first stratified according to prior antiretroviral therapy exposure (i.e., naïve or experienced), then randomized 3:1 to treatment for 24 weeks with either APV 600 mg BID plus RTV 100 mg BID or to APV 1200 mg BID in combination with ≥ 2 non-PI antiretroviral drugs. APV was supplied as 150 mg soft-gelatin capsules of Agenerase^® (GlaxoSmithKline, Research Triangle Park, North Carolina); thus, eight capsules were given for each 1200 mg dose of APV and four capsules for each 600 mg dose. RTV was supplied as 100 mg soft-gelatin capsules of Norvir^® (Abbott Laboratories, North Chicago, Illinois). Late in the study, patients who had not yet completed 24 weeks of treatment were given the option to continue treatment in a subsequent 24-week extension phase to assess the durability of virologic response.

During the 6-week period prior to the start of the study, candidates were screened for study eligibility, demography, CDC classification, HIV risk factors, mode of HIV transmission, antiretroviral therapy history, and medical history, and given a physical examination. Prior to study initiation, the ESS40011 study protocol was approved by the Institutional Review Boards at each participating study site. All participants provided written informed consent before any study-related procedures were commenced. The initial 24-week portion of the study was conducted between June 13, 2000 and October 18, 2001 at 44 treatment centers in the United States.

The primary efficacy parameter was the proportion of patients with plasma HIV-1 RNA levels <200 copies/mL at week 24. Plasma HIV-1 RNA was assessed by the Roche Amplicor MONITOR Ultrasensitive assay (version 1.5; lower limit of quantitation [LLOQ] 50 copies/mL and quantitation range of 50 to 75,000 copies/mL, Roche Diagnostics, Branchburg, New Jersey). Patients were classified as having reached a virologic endpoint if, at week 24, plasma HIV-1 RNA levels were ≥ 200 copies/mL; or if the plasma HIV-1 RNA had not decreased by at least 0.5-log₁₀ from baseline at week 8, with confirmation at week 12.

Secondary efficacy measures included assessment of the proportion of patients achieving HIV-1 RNA <50 copies/mL (using the above mentioned assay); changes in CD4+ cell counts (measured by flow cytometry) compared with baseline values; and progression of HIV disease from baseline status to the occurrence of the first new event involving a change in CDC class for each patient. Patients who remained at their entry CDC class were considered to be clinical non-progressors.

Change from baseline in plasma HIV-1 RNA levels was assessed at weeks 4, 8, 12, and 24 in all patients, as well as at weeks 32, 40 and 48 in patients who participated in the treatment extension phase. CD4+ cell counts were assessed at baseline, and at weeks 12 and 24. Patients had plasma collected for genotype/virtual phenotype at screening (for experienced patients only) and at the time of virologic failure (results to be reported elsewhere).

The study was powered to evaluate whether the virologic efficacy (proportion of patients with plasma HIV-1 HIV <200 copies/mL at week 24) in the APV 600/RTV arm was at least as good (non-inferior) as in the APV1200 arm. Non-inferiority of the APV600/RTV regimen to the APV1200 regimen was established if the 95% lower confidence limit (LCL) for the difference in proportion of patients achieving HIV-1 RNA <200 copies/mL at week 24 with APV 600/RTV minus APV1200 was ≥-0.12. The non-inferiority margin of 0.12 was chosen based on previous regulatory studies and was pre-specified in the protocol. Target enrollment was 198 in the APV600/RTV arm and 66 in the APV1200 arm. The primary analysis was made in the intent-to-treat (ITT) population, which consisted of all eligible patients who were randomized into the study regardless of what treatment was actually received and the eventual outcome of study participation. Two types of analyses were performed: an ITT: observed analysis, in which only available assessments were used (no imputation for missing values), regardless of whether the patient was still receiving their original therapy; and an ITT: missing = failure (ITT: M = F) analysis, in which all missing values constituted failure. Comparisons between treatment arms regarding proportion of patients achieving HIV-1 RNA <200 copies/mL and <50 copies/mL were made using the Cochran-Mantel-Haenszel test controlling for prior antiretroviral drug experience. Changes from baseline in plasma HIV-1 RNA (log₁₀ copies/mL) and CD4+ cell count, and average area under the curve minus baseline (AAUCMB) in plasma HIV-1 RNA, were tabulated by treatment arm and visit, then analyzed using analysis of variance (ANOVA). The safety population consisted of all patients who consumed at least 1 dose of study drug. Safety parameters included incidence of treatment-limiting toxicities (clinical and laboratory adverse events); and change from baseline in selected laboratory variables at weeks 4, 8, 12 and 24. Fisher's exact test was used to compare adverse event rate and ANOVA was used to compare change from baseline for laboratory variables between the two treatment arms. A P value of <0.05 was considered statistically significant. Statistical analyses were performed using SAS version 6.12 (SAS Institute, Inc., Cary, NC).

A total of 211 patients enrolled; they were predominantly male (87%), ethnically diverse (48% Caucasian, 37% African American, 14% Hispanic), and had a mean age of 42 years, median baseline HIV-1 RNA level of 4.33 log₁₀ copies/mL, and median baseline CD4+ cell count of 257 cells/mm³. Most of the study population was antiretroviral-experienced (81%). One hundred and fifty-eight (158) patients were randomized to the APV600/RTV regimen and 53 to the APV1200 regimen. The two treatment arms did not differ with respect to any baseline characteristic (Table 1). The most common background antiretroviral regimens taken by patients in the APV600/RTV and APV1200 treatment arms were stavudine/didanosine (25% vs 30%), Combivir (19% for both arms), abacavir/stavudine (11% for both arms), lamivudine/stavudine (9% vs 4%), and abacavir/didanosine (5% vs 6%). A total of 150 patients (71%) completed all 24 weeks of the study, including a similar proportion of patients in the APV600/RTV arm (72%) and APV1200 arm (68%). Reasons for premature withdrawal from treatment were also similar between the APV600/RTV and APV1200 arms, except for fewer patients in the APV600/RTV arm withdrawing due to protocol-defined virologic failure (1% vs 6%) (Table 1). Antihyperlipidemic agents were used by 1 patient in each treatment arm, neither of whom had taken this type of medication pre-study. No deaths or clinical progressions of HIV disease occurred. Twenty patients enrolled into the extension phase, including 15 in the APV600/RTV arm and 5 in the APV1200 arm.

At week 24, the proportion of patients with HIV-1 RNA <200 copies/mL in the APV600/RTV arm and APV1200 arm was 46% (73/158) and 38% (20/53), respectively, in the ITT: M = F analysis (Figure 1) and 62% (73/118) and 53% (20/38), respectively, in the ITT: observed analysis (Figure 2). The APV600/RTV regimen proved to be similar to or better than the APV1200 regimen because the 95% LCL for the difference in proportions of patients achieving HIV-1 RNA <200 copies/mL with APV600/RTV minus APV1200 was ≥ -0.12 (-0.04 [ITT: M = F analysis] and -0.06 [ITT: observed analysis]). The proportion of antiretroviral-naïve patients who achieved HIV-1 RNA <200 copies/mL in the APV600/RTV arm was twice that for the APV1200 arm in the ITT: M = F analysis (63% [19/30] vs 30% [3/10], P = 0.141), and 26% higher in the ITT: observed analysis (76% [19/25] vs 50% [3/6], P = 0.320). In contrast, the proportions in antiretroviral-experienced patients were similar in the treatment arms in both analyses (ITT: M = F analysis: 42% [54/128] vs 40% [17/43], P = 0.859; ITT: observed analysis: 58% [54/93] vs 53% [17/32], P = 0.682).

Virologic findings using the more stringent virologic suppression endpoint of HIV-1 RNA <50 copies/mL paralleled those with the <200 copies/mL endpoint, although differences attained statistical significance. Thus, at week 24, significantly more patients in the APV600/RTV arm than the APV1200 arm achieved HIV-1 RNA <50 copies/mL (ITT: M = F analysis: 36% [57/158] vs 21% [11/53], P = 0.039; ITT: observed analysis: 48% [57/118] vs 29% [11/38], P = 0.042) (Figures 1 and 2, respectively). The proportion of antiretroviral-naïve patients who achieved HIV-1 RNA <50 copies/mL was 47% (14/30) in the APV600/RTV arm and 20% (2/10) in the APV1200 arm (P = 0.141), whereas the proportion of antiretroviral-experienced patients who achieved HIV-1 RNA <50 copies/mL was 34% (43/128) in the APV600/RTV arm and 21% (9/43) in the APV1200 arm (P = 0.119) (ITT:M = F analysis). In the ITT: observed analysis, mean reduction in HIV-1 RNA from baseline was significantly greater at week 24 in the APV600/RTV arm than the APV1200 arm (-2.21 vs -1.59 log₁₀ copies/mL, P = 0.028) (Figure 3), as was the mean AAUCMB in HIV-1 RNA in both the total population (-1.56 vs -1.25 log₁₀ copies/mL, P = 0.045) and antiretroviral-naïve subgroup (-2.40 vs -1.60 log₁₀ copies/mL, P = 0.020). No significant differences were observed between the APV600/RTV and APV1200 treatment arms with respect to mean AAUCMB in HIV-1 RNA for the antiretroviral-experienced subgroup (-1.40 vs -1.20 log₁₀ copies/mL P = 0.209), or proportion of antiretroviral-naïve patients achieving HIV-1 RNA <50 copies/mL (56% vs 33%, P = 0.394) (ITT: observed analysis).

Of the 20 patients (15 in the APV600/RTV arm and 5 in the APV1200 arm) enrolled into the extension phase, 15 (11 APV600/RTV, 4 APV1200) had HIV-1 RNA <200 copies/mL at week 24. Eleven (8 APV600/RTV, 3 APV1200) of these 15 patients continued to have HIV-1 RNA levels <200 copies/mL between week 24 and week 48, 3 rebounded, and 1 discontinued at week 32. Of the 5 patients who had HIV-1 RNA >200 copies/mL at week 24, 1 discontinued when an HIV-1 RNA of 186 copies/mL was measured at week 32, 1 missed week 32 and 40 visits (but had an HIV-1 RNA of 878 copies/mL at week 48), and 3 continued to have HIV-1 RNA >200 copies/mL from week 24 to week 48. In these 3 patients, HIV-1 RNA levels ranged from 316 to 3645 copies/mL, 1247 to 5175 copies/mL, and 143 to 12,149 copies/mL, respectively, between weeks 24 and 48.

At baseline, the median CD4+ cell count in the APV600/RTV and APV1200 arms was 271 and 255 cells/mm³, respectively. Over the 24 weeks of the study, the median CD4+ cell count remained higher than baseline, with median elevations above baseline peaking at week 12 in both treatment arms (+51 and +52 cells/mm³, respectively) (Figure 3). At week 24, the median change from baseline in CD4+ cell count in the APV600/RTV and APV1200 arms was +35 and +46 cells/mm³, respectively, and the final median CD4+ cell count was 321 and 346 cells/mm³, respectively. In patients who participated in the 24-week extension phase, 12 of 15 patients in the APV600/RTV arm and 4 of 5 in the APV1200 arm had a median change from baseline in CD4+ cell count at week 48 of +156 cell/mm³, and +143 cell/mm³, respectively. The final median CD4+ cell count in these patients was 404 and 407 cells/mm³, respectively. The remaining patients did not have CD4+ data reported.

Table 2 shows the drug-related adverse events that were reported in ≥ 5% of patients. Nausea, diarrhea, vomiting, and fatigue were the most common adverse events in both treatment arms. The incidence of drug-related oral/perioral paresthesia was lower in the APV600/RTV treatment arm than the APV1200 arm (2% vs 8%). No differences between the APV600/RTV and APV1200 treatment arms were observed regarding the frequency of drug-related grade 1–4 adverse events (44% vs 45%), frequency of discontinuing treatment due to adverse events (7% vs 8%), incidence of hyperglycemia (1 vs 0 patient), nonspecific lipodystrophy (1 vs 0), buffalo hump (1 vs 0), or hypercholesterolemia (1 vs 0). More cases of hypertriglyceridemia were reported as adverse events in the APV600/RTV arm (11 [7%] vs 0). However, review of laboratory changes revealed that the incidence of grade 3–4 hypertriglyceridemia was the same (4%) in each treatment arm.