| |
Substitution of nevirapine or efavirenz for protease inhibitor versus lipid-lowering therapy for the management of dyslipidaemia
|
| |
| |
AIDS, July 2005
Calza, Leonardo; Manfredi, Roberto; Colangeli, Vincenzo; Tampellini, Livia; Sebastiani, Teresa; Pocaterra, Daria; Chiodo, Francesco
From the Department of Clinical and Experimental Medicine, Division of Infectious Diseases, Alma Mater Studiorum University of Bologna, S. Orsola Hospital, Bologna, Italy.
"..... In our study, pharmacological therapy with pravastatin or bezafibrate proved significantly more effective in the management of diet-resistant mixed hyperlipidaemia in HAART-treated patients than the switch from PIs to non-nucleoside analogues.
Particularly, hypolipidaemic agents obtained a significantly greater reduction in both triglyceride, total and LDL and cholesterol levels than simplification therapy with nevirapine or efavirenz. With regard to simplification therapy, switch to nevirapine led to a significantly higher decrease both in plasma triglyceride and cholesterol concentrations than substitution of efavirenz for PI. However, only about 50% of patients in pravastatin and bezafibrate groups reached normal plasma lipid levels after a 12-month follow-up, and a slight-to-moderate hyperlipidaemia persisted in the remaining subjects. Hypolipidaemic compounds and simplification therapy with NNRTIs showed a favourable tolerability and adherence profile, and there were no appreciable differences in the incidence of clinical and laboratory adverse effects in all the four study groups.
In conclusion, simplification of a PI-based treatment with nevirapine or efavirenz in patients with a sustained virological response and no prior suboptimal antiretroviral therapy had a significant probability of maintaining viral suppression, and lowering plasma lipid levels, but pharmacological treatment with pravastatin or bezafibrate added to current HAART had a superior hypolipidaemic efficacy and was usually well tolerated....."
ABSTRACT
Objectives: To evaluate simplified protease inhibitor (PI)-sparing antiretroviral treatment versus lipid-lowering therapy for the management of highly active antiretroviral therapy (HAART)-induced hyperlipidaemia.
Design: Randomized, open-label clinical trial assessing the efficacy on hyperlipidaemia of a switching therapy from PI to non-nucleoside reverse transcriptase inhibitor (NNRTI) nevirapine or efavirenz versus a hypolipidaemic treatment (with pravastatin or bezafibrate) added to current, unchanged antiretroviral combination.
Methods: All HIV-infected patients on their first HAART regimen, with stable immuno-virological features, naive to all NNRTIs, and with mixed hyperlipidaemia, were randomized to replace PI with nevirapine (arm A) or efavirenz (arm B), or to receive pravastatin (arm C) or bezafibrate (arm D) with unchanged HAART regimen, and were followed-up for 12 months.
Results: One hundred and thirty patients were evaluated: 29 patients were randomized to arm A, 34 to arm B, 36 to arm C, and 31 to arm D. At the end of the 12-month follow-up, a reduction of 25.2, 9.4, 41.2 and 46.6% in mean triglyceridaemia versus respective baseline values was reported in groups A, B, C and D, respectively, with statistically significant difference between arms A-B and C-D (P < 0.01). Similar results were reported for total and low-density lipoprotein cholesterol levels. Viro-immunological efficacy and tolerability profile were comparable in all considered arms.
Conclusion: Pravastatin and bezafibrate proved significantly more effective in the management of HAART-related hyperlipidaemia than the switching therapy from PI to nevirapine or efavirenz.
INTRODUCTION
Since new protease inhibitor (PI)-based antiretroviral regimens have led to a notable extension of life expectancy in HIV-positive patients, prolonged hyperlipidaemia could significantly act on the long-term prognosis and outcome of HIV-infected persons, so that there is increasing concern particularly about the higher risk of cardiovascular complications [1-6]. Increased plasma lipid levels are associated with higher low-density lipoprotein (LDL) cholesterol concentrations and atherogenic ratios, and recent studies have shown an increased risk of myocardial infarction and cerebrovascular events in patients treated with highly active antiretroviral therapy (HAART) in comparison with the general population [7-9].
Even though hypolipidaemic diet and physical exercise may certainly improve dyslipidaemia, substitution of PIs with non-nucleoside reverse transcriptase inhibitors (NNRTIs) or the nucleoside reverse transcriptase inhibitor (NRTI) abacavir, or lipid-lowering drugs (such as statins or fibrates) can be considered when plasma lipid levels excessively increase or persist for more than 3 months [10,11]. (Ed Note: another option is to switch to Reyataz).
On the other hand, lipid-lowering pharmacological treatment is advisable when PI-sparing simplified therapy is not applicable, or serum lipid concentrations excessively increase or persist. Moreover, the choice of hypolipidaemic drugs is often influenced by adjunctive problems because of expected drug-drug interactions with antiretroviral compounds, toxicity, and decreased patient adherence to multiple pharmacological regimens [3,12,13].
The aim of our study was to compare efficacy and safety of substituting nevirapine or efavirenz for a PI with those of adding a lipid-lowering drug (pravastatin or bezafibrate) to the current HAART regimen in HIV-infected patients with mixed hyperlipidaemia and stable virological suppression. This is the first randomized, open-label, clinical trial, to the best of our knowledge, comparing PI-sparing simplified therapy with hypolipidaemic pharmacological treatment for the management of HAART-induced dyslipidaemia.
Results
Hypertriglyceridaemia was found in 422 out of the 696 (60.6%) HIV-infected subjects referred to our tertiary care outpatient centre and on a PI-based antiretroviral regimen for at least 12 months, while hypercholesterolaemia was detected in 294 individuals (42.2%).
In detail, 173 out of the 696 patients (24.8%) had isolated hypertriglyceridaemia, 45 (6.5%) had isolated hypercholesterolaemia, whereas 249 (35.8%) showed mixed hyperlipidaemia. Lipodystrophy was observed in 222 individuals (31.9%), associated with hypertriglyceridaemia and/or hypercholesterolaemia in 110 overall cases (15.8%). In particular, lipoatrophy was reported in 91 subjects out of 696 (13%), fat accumulation in 73 (10.5%), while a mixed form (peripheral fat loss associated with central fat accumulation) was seen in 58 individuals (8.3%). The prevalence rate of hypertriglyceridaemia and/or hypercholesterolaemia in patients with lipoatrophy, fat accumulation and a mixed form, was 28 out of 91 (30.8%), 39 out of 73 (53.4%), and 43 out of 58 (74.1%), respectively.
Between January 2001 and December 2002, of the 696 patients observed, 142 (20.4%) were found eligible and enrolled for the study. The remaining 107 patients with mixed hyperlipidaemia were not enrolled in the study because they met one or more exclusion criteria, including previous PI- or NNRTI-based treatments (n = 91 patients), previous suboptimal single or dual NRTI therapy (n = 67), chronic hepatitis B or C (n = 62), diabetes mellitus (n = 6), corticosteroid therapy (n = 5), beta-blocker therapy (n = 4), dysthyroidism (n = 3), major psychiatric illness (n = 3), thyroid preparations (n = 2), severe hyperlipidaemia (n = 2), hyperlipidaemia with duration lower than six months (n = 41), or responsive to hypolipidaemic diet and physical activity (n = 39). Twelve out of the 142 enrolled patients were excluded from final evaluation due to early drop-out: these 12 subjects did not take any dose of the new treatment. Epidemiological, clinical and laboratory features of these 130 evaluable patients are summarized in Table 1.
Ongoing NRTI therapy included stavudine in 60 cases (46.1%), zidovudine in 60 (46.1%), lamivudine in 68 (52.3%), didanosine in 62 (47.7%), abacavir in seven (5.4%), and tenofovir in three (2.3%). Ongoing PI therapy included lopinavir/ritonavir in 36 cases (27.7%), nelfinavir in 28 (21.5%), indinavir in 22 (16.9%), indinavir/ritonavir in 11 (8.5%), saquinavir in 18 (13.8%), saquinavir/ritonavir in 10 (7.8%), and amprenavir/ritonavir in five (3.8%). Treatment duration ranged from 12 to 42 months, with a mean length ± one standard deviation (SD) of 29.2 ± 11.4 months. Concomitant lipodystrophy was diagnosed in 75 cases (57.7%): in detail, concurrent lipoatrophy, fat accumulation and mixed form were shown by 34 (26.1%), 25 (19.2%), and 16 out of 130 patients (12.3%), respectively.
Mean plasma triglyceride level ± SD was 293.2 ± 106.1 mg/dl, and tested significantly higher in patients treated with lopinavir/ritonavir (358.6 ± 152.1 mg/dl) compared with other boosted and unboosted PIs (265.2 ± 91.8 mg/dl) (P < 0.005). Mean plasma total cholesterol concentration ± SD was 266.8 ± 72.4 mg/dl, and did not show any significant difference according to the administered PI-based regimen, as well as the lipodystrophy prevalence (data not shown).
Concomitant hyperinsulinaemia (C-peptide levels > 4 ng/ml) was found in 13 patients (10%), whereas no cases of fasting hyperglycaemia (glucose levels > 110 mg/dl) were observed among the 130 evaluable subjects.
Thirty-four patients switched from PI to efavirenz (arm B), and 29 to nevirapine (arm A), while continuing to take concomitant NRTI regimen. Thirty-six subjects started a lipid-lowering therapy with pravastatin (arm C), and 31 with bezafibrate (arm D), without changes in their HAART combination. A hypolipidaemic diet and regular physical exercise were persistently and consistently recommended in all considered arms.
The four compared patient groups did not show any significant difference according to age and gender distribution, type of exposure to HIV infection, HIV disease stage, mean CD4 cell count, type and mean duration of antiretroviral treatment, and mean baseline plasma triglyceride and cholesterol levels (total concentration and LDL-HDL components) (Table 2). All of these patients were treated for at least 12 months, and were followed-up for efficacy, and clinical and laboratory safety of switching therapy to NNRTIs and hypolipidaemic drugs.
After the first 6 months of follow-up, the reduction of mean plasma triglyceride levels versus baseline respective values was 26.4% in arm A, 9.8% in arm B, 42.3% in arm C, and 47.2% in arm D. These results were maintained after 12 months without relevant modifications: at the end of a 1-year follow-up, the decrease in mean triglyceride levels versus respective baseline values was 25.2% in arm A, 9.4% in arm B, 41.2% in arm C, and 46.6% in arm D. The reduction of mean triglyceride concentrations was significantly higher in patients treated with lipid-lowering drugs [mean decrease of 44.1% or 129 mg/dl; 95% confidence intervals (CI), -238 to -20 mg/dl), than in those switched from PI to nevirapine or efavirenz (mean decrease of 18.4% or 53 mg/dl; 95% CI, -97 to -9 mg/dl) (P < 0.01). Bezafibrate treatment showed a more obvious effect than pravastatin on the hypertriglyceridaemia, but no statistically significant differences of therapeutic responses were observed between these drugs. With regard to simplification therapy, switch to nevirapine obtained a significantly greater decrease in mean triglyceride levels (65 mg/dl; 95% CI, -115 to -15 mg/dl) than switch to efavirenz (-19 mg/dl; 95% CI, -34 to -4 mg/dl) (P < 0.05).
At the same time, the mean reduction of plasma total cholesterol levels versus baseline values after the first 6 months of treatment was 28.5% in arm A, 11.4% in arm B, 44.2% in arm C, and 36.8% in arm D. At the end of the 1-year follow-up, the decrease in mean total cholesterol concentrations versus baseline levels was 27.1% in arm A, 10.2% in arm B, 45.8% in arm C, and 37.6% in arm D. Similar changes in plasma lipid LDL cholesterol concentrations were reported, while a slight increase in plasma HDL cholesterol levels was observed only in patients treated with lipid-lowering agents. Variations in plasma lipid profile obtained in the four evaluated groups at the close of the 12-month follow-up are summarized in Table 3, while temporal trends of mean plasma triglyceride and total cholesterol levels are represented in Figures 1 and 2, respectively.
Decreases in mean plasma total and LDL cholesterol levels were significantly higher in subjects treated with pravastatin or bezafibrate (mean decreases of 41.8% or 109 mg/dl; 95% CI, -173 to -45 mg/dl, for total cholesterol levels; 37.6% or 59 mg/dl; 95% CI, -87 to -31 mg/dl, for LDL cholesterol levels) than in nevirapine- and efavirenz-treated patients (mean decreases of 19.2% or 50 mg/dl; 95% CI, -88 to -12 mg/dl, for total cholesterol levels; 16.5% or 26 mg/dl; 95% CI, -48 to -8 mg/dl, for LDL cholesterol levels) (P < 0.01). Pravastatin therapy showed a more evident effect than bezafibrate on hypercholesterolaemia, but no statistically significant differences were observed. Moreover, mean reductions of total and LDL cholesterol values were significantly greater in nevirapine group (27.1% or 69 mg/dl; 95% CI, -119 to -19 mg/dl, for total cholesterol levels; 25.2% or 38 mg/dl; 95% CI, -65 to -11 mg/dl, for LDL cholesterol levels) than in efavirenz one (10.2% or 25 mg/dl; 95% CI, -46 to -4 mg/dl, for total cholesterol levels; 8.7% or 14 mg/dl; 95% CI, -26 to -2 mg/dl, for LDL cholesterol levels) (P < 0.05).
The percentage of patients who reached normal triglyceride levels after a 12-month follow-up was significantly higher in pravastatin and bezafibrate-treated group (34 out of 67, or 50.7%) than in the nevirapine and efavirenz-treated one (12 out of 63, or 19%) [odds ratio (OR), 4.48;, 95% CI, 1.5 to 6.2; P < 0.01]. Similarly, the percentage of patients who obtained normal total cholesterol concentrations was significantly greater in arms C and D (33 of 67, 49.2%) than in arms A and B (11 of 63, 17.5%) (OR, 4.62; 95% CI, 1.7 to 6.6; P < 0.01).
No appreciable changes of either prevalence or severity of clinical signs of fat redistribution syndrome (lipoatrophy, fat accumulation and mixed form) were reported throughout the study. No significant differences of hypolipidaemic therapeutic responses were detected according to underlying PI-based regimens, baseline lipid levels, and CD4 cell counts.
Both simplification therapy with nevirapine or efavirenz and lipid-lowering treatment with pravastatin or bezafibrate showed a favourable tolerability profile, and in all the evaluated arms there were no appreciable differences in the onset of side-effects, as reported spontaneously by the patients and observed by physical examination and laboratory tests. Mild gastroenteric signs and symptoms (nausea, dyspepsia, abdominal pain, flatulence, and diarrhoea) were found with similar incidence in the four evaluated arms (10.3% in arm A, 11.8% in arm B, 8.3% in arm C, 6.4% in arm C), and they did not require any therapy modification or suspension. Neuropsychic symptoms (anxiety, insomnia, and nightmares) were observed in six patients treated with efavirenz, but they were mild-to-moderate and spontaneously disappeared within the first 6 weeks of treatment. A maculo-papular and pruriginous skin rash was observed in two patients receiving nevirapine and in one treated with efavirenz, leading to the interruption of treatment and early drop-out within the first 8 weeks of therapy. Neither myalgia nor myositis were observed in all the considered arms, and no significant laboratory adverse events occurred, with particular regard to serum CPK levels. Mild increases in alanine amino-transferase concentrations (always below 120 U/l) were reported in two subjects treated with nevirapine and in one patient treated with efavirenz, but they did not require any therapy modifications and remained mild during the entire follow-up.
At the close of the 1-year observation period, the virological efficacy was similar in all the four arms, without statistically significant differences. The number of patients with plasma HIV viral load persistently lower than 50 copies/ml was 28 in arm A (96.5%), 33 in arm B (97%), 36 in arm C (100%), and 30 in arm D (96.8%). At the same time, the immunological responses were comparable in all the evaluated groups: after a 12-month follow-up, the mean increases in CD4 lymphocyte counts versus respective baseline values were 92.2 106 cells/l in arm A, 108.4 106 cells/l in arm B, 89.2 106 cells/l in arm C, and 102.9 106 cells/l in arm D.
AUTHOR DISCUSSION
The potential clinico-pathological consequences of HIV-associated dyslipidaemia are not still completely known, but several anecdotal observations and small case series report an increased risk of premature coronary artery disease and myocardial infarction in young HIV-infected individuals treated with HAART, besides peripheral atherosclerosis, acute pancreatitis, and cutaneous xantomas [13-28]. Insulin resistance syndrome, hypercholesterolaemia, hypertriglyceridaemia, low levels of HDL cholesterol, and truncal adiposity are known to increase cardiovascular risk in HIV-negative population, and may similarly predispose HIV-infected subjects to accelerated cardiovascular events [26-31].
The concerns about the long-term risk of coronary artery disease from these lipid metabolism disorders have been recently confirmed by the DAD study [9]. In this prospective, observational trial involving 23 468 HIV-positive patients from 11 previously established cohorts, Friis-Möller et al. showed that combination antiretroviral therapy was associated with a 26% relative increase in the rate of myocardial infarction per year of exposure during the first 4 to 6 years of treatment, although the absolute event rate was low (incidence of 3.5 events per 1000 person-years) and must be balanced against the marked benefits from antiretroviral therapy.
These findings led to a series of clinical trials and observational studies where different kinds of management for PI-related hyperlipidaemia were evaluated for reducing the risk of premature coronary events.
Several studies have demonstrated that an antiretroviral regimen in which a PI is replaced with nevirapine, efavirenz or abacavir in patients with long-lasting viral suppression usually maintains optimal antiviral activity. Moreover, as compared with PIs, these agents reduce serum lipid abnormalities, offer more convenient dosing regimens, involve fewer pills, and result in fewer potentially serious drug-drug interactions, reduced side effects, and improved adherence to antiretroviral therapy. However, the rate of virological failure might eventually increase among patients who have previously received prolonged non-suppressive antiretroviral treatment, such as single or dual NRTI therapy, as a result of the re-emergence of archived viral resistance [32,33].
In a recent meta-analysis, Bucher et al. [34] have compared efficacy and safety of continued PI-based therapy versus a switch from PI to abacavir, efavirenz, or nevirapine. In this study, there are indications of an increased risk of virological failure for patients with prior suboptimal NRTI treatment, and this risk seems to be highest for patients switched to abacavir. With regard to metabolic parameters, this meta-analysis suggested a trend towards lower cholesterol and triglyceride levels in patients receiving PI-sparing simplification regimens, and the most evident reduction was detected in subjects switched from PI to abacavir.
Significant improvement in serum lipid concentrations was also reported after switching from stavudine to abacavir or tenofovir [35-37], or after replacing current PI with atazanavir [38,39].
Lipid-lowering therapy becomes suitable when dietary changes, physical exercise and switching treatment are ineffective or not applicable. Drug therapy for dyslipidaemia in HIV-infected persons receiving HAART is problematic, because of potential drug interactions, toxicity, intolerance, and decreased patient adherence to multiple pharmacologic regimens.
Consequently, it is reasonable to recommend the use of pravastatin or low dosage of atorvastatin as first-line treatment for hypercholesterolaemia in PI-treated patients, and the use of fluvastatin (characterized by a slightly lower efficacy), as second-line regimen. On the other hand, simvastatin and lovastatin should be avoided, because they present a great risk of pharmacological interactions with PIs [13,40-45].
The available data on management of HAART-related dyslipidaemia with pravastatin, atorvastatin or fluvastatin has been limited to few clinical studies (mostly non-randomized or non-placebo-controlled), involving a small number of patients and with a limited follow-up. Significant toxicities have not been reported in these studies, but the cholesterol-lowering effects of statins in subjects receiving PIs were usually modest as compared with HIV-negative general population, and many patients have not reached cholesterol goals with statin monotherapy. This lack of efficacy of statins in HIV-positive patients is not explained by drug interactions between PIs and statins, as the levels of most statins increase in association with the use of PIs, but may be due to the hyperlipidaemic effect of continued PI-based therapy [44,45].
Fibrates represent the cornerstone of drug therapy for hypertriglyceridaemia and mixed hyperlipidaemia. Treatment with gemfibrozil, bezafibrate or fenofibrate generally results in a significant reduction in triglyceride and cholesterol levels in HIV-infected patients receiving a PI-containing therapy, with a more evident improvement of hypertriglyceridaemia. Recent data demonstrate additive effect for the statin-fibrate combination therapy: the mean reduction in both cholesterol and triglyceride levels was significantly higher than with statin or fibrate monotherapy. However, the statin-fibrate association therapy should only be used with great caution because of the increased risk of skeletal muscle toxicity [13,42-45].
In our study, pharmacological therapy with pravastatin or bezafibrate proved significantly more effective in the management of diet-resistant mixed hyperlipidaemia in HAART-treated patients than the switch from PIs to non-nucleoside analogues.
Particularly, hypolipidaemic agents obtained a significantly greater reduction in both triglyceride, total and LDL and cholesterol levels than simplification therapy with nevirapine or efavirenz. With regard to simplification therapy, switch to nevirapine led to a significantly higher decrease both in plasma triglyceride and cholesterol concentrations than substitution of efavirenz for PI. However, only about 50% of patients in pravastatin and bezafibrate groups reached normal plasma lipid levels after a 12-month follow-up, and a slight-to-moderate hyperlipidaemia persisted in the remaining subjects. Hypolipidaemic compounds and simplification therapy with NNRTIs showed a favourable tolerability and adherence profile, and there were no appreciable differences in the incidence of clinical and laboratory adverse effects in all the four study groups.
In conclusion, simplification of a PI-based treatment with nevirapine or efavirenz in patients with a sustained virological response and no prior suboptimal antiretroviral therapy had a significant probability of maintaining viral suppression, and lowering plasma lipid levels, but pharmacological treatment with pravastatin or bezafibrate added to current HAART had a superior hypolipidaemic efficacy and was usually well tolerated.
Methods
Plasma lipid levels and body fat distribution of HAART-treated patients referring to our tertiary care outpatient centre between January 2001 and December 2002 were evaluated.
Hypertriglyceridaemia was defined by plasma fasting triglyceride levels > 200 mg/dl, hypercholesterolaemia by plasma fasting total cholesterol levels > 250 mg/dl, and mixed hyperlipidaemia by the association of hypertriglyceridaemia with hypercholesterolaemia. An evident body fat redistribution syndrome or lipodystrophy was established by physical examination and observation of peripheral fat wasting (involving face and/or buttocks and limbs), and/or abdominal, mammary or dorsocervical ('buffalo hump') fat accumulation.
Eligible patients were HIV-infected adults who were receiving their first HAART regimen (consisting of two NRTIs plus one or more PIs) of at least 12-month duration, who had stable immuno-virological parameters (plasma HIV viral load < 50 copies/ml and CD4 lymphocyte count > 350 106 cells/l for at least 6 months), and with diagnosis of mixed hyperlipidaemia (with or without lipodystrophy) of at least 6-month duration and unresponsive to a hypolipidaemic diet and regular exercise of 3 or more months. Exclusion criteria were prior treatment with any NNRTI, prior single or dual NRTI therapy, concomitant drug or alcohol abuse, history of genetic hyperlipidaemia, coronary heart disease, acute or chronic pancreatitis, major psychiatric illness, diabetes mellitus (fasting glucose levels > 126 mg/dl), dysthyroidism, Cushing's syndrome, acute or chronic liver or kidney disease, myopathy, pregnancy, or concurrent treatment with lipid-lowering agents, corticosteroids, androgens, oestrogens, thiazide diuretics, beta-blockers, thyroid preparations, or anticoagulants. Patients with severe hypertriglyceridaemia (fasting plasma triglyceride levels > 750 mg/dl) or severe hypercholesterolaemia (fasting plasma total cholesterol levels > 350 mg/dl) were also excluded from the study. Written informed consent was obtained from all eligible patients before randomization.
All the enrolled subjects were randomly assigned to switch from PI to nevirapine (arm A) or efavirenz (arm B) in their current antiretroviral regimen while continuing to take their NRTIs, or to receive oral hypolipidaemic treatment with 20 mg daily of pravastain (arm C) or 400 mg daily of bezafibrate (arm D) with their unchanged HAART regimen, according to a computer-generated randomization list. Lipid-lowering drugs were administered at initial dosage recommended for the management of PI-associated dyslipidaemia [13]. All of these enrolled patients were prospectively followed-up at 3-month intervals for at least 12 months, in order to assess both efficacy and tolerability of substitution therapy and lipid-lowering treatment, according to some epidemiological, clinical, and laboratory features of HIV infection.
Eventual clinical adverse events and adherence to drug therapies were carefully checked on monthly outpatient visits, while plasma fasting lipid levels (triglyceride and total-LDL-HDL cholesterol concentrations), haematology, CD4 lymphocyte count, plasma HIV viral load, complete liver and kidney function tests, coagulation profile, serum creatine-phosphokinase (CPK), aldolase, amylase, lipase, and urinalysis, were performed at day 0 and subsequently at months 3, 6, 9, and 12. Plasma HIV viral load was evaluated using the bDNA Quantiplex HIV-RNA-3 assay (Chiron Corporation, Emeryville, California, USA), with a lower limit of detection placed at 50 bDNA copies/ml.
The primary study endpoint was decrease in triglyceride and total cholesterol levels at month 12 compared with respective baseline values in the four evaluated arms, while secondary endpoints included variation in high-density lipoprotein (HDL) and LDL cholesterol concentrations, HIV viral load, and CD4 cell count, in association with the evaluation of clinical and laboratory adverse events. Moreover, we compared the efficacy in lowering plasma lipid levels of the two 'switch' arms (A-B) versus the two 'add-on' arms (C-D). Mantel-Haenszel chi-square test and Fisher's exact test were used for comparisons of proportions. Student's t-test was used for comparisons of quantitative variables, with significance levels placed at P < 0.05. All analyses were performed using an intent-to-treat approach.
|
|
| |
| |
|
|
|
