Antiviral Therapy 2012; 17: (doi: /IMP2305)

Transcription

1 Antiviral Therapy 12; 17:111 1 (doi: /IMP235) Original article Switching to tenofovir/emtricitabine from abacavir/ lamivudine in HIV-infected adults with raised cholesterol: effect on lipid profiles Georg Behrens 1 *, Renato Maserati 2, Armin Rieger 3, Pere Domingo 4, Florian Abel 5, Hui Wang 6, Gill Pearce 7 1 Hannover Medical School, Department for Clinical Immunology and Rheumatology, Hannover, Germany 2 Foundation IRCCS San Matteo Hospital, Pavia, Italy 3 Medizinische Universität Wien, Klinische Abteilung für Immundermatologie und Infektiöse Hautkrankheiten, Wien, Austria 4 Hospital Santa Creu i Sant Pau, Barcelona, Spain 5 Gilead Sciences GmbH, Martinsried, Germany 6 Gilead Sciences, Inc., Foster City, CA, USA 7 Gilead Sciences Europe, Ltd, Uxbridge, UK *Corresponding author behrens.georg@mh-hannover.de Background: The aim of this study was to investigate the effect on fasting lipid parameters of switching to tenofovir disoproxil fumarate (TDF) plus emtricitabine (FTC) from abacavir (ABC) plus lamivudine (3TC; both fixeddose combinations), while maintaining ritonavir-boosted lopinavir (LPV/r). Methods: This was an open-label randomized two-arm 12-week controlled study in virologically suppressed HIVinfected patients with elevated cholesterol ( 5.2 mmol/l). Patients stable on ABC/3TC plus LPV/r either continued treatment or switched to TDF/FTC plus LPV/r for 12 weeks. Standard efficacy and safety end points (including fasting lipids) were assessed. Results: In total, 85 subjects were treated (n=42 ABC/FTC and n=43 TDF/3TC). A statistically significant decrease in total cholesterol was observed in the TDF/FTC group: from median (IQR) 6.22 mmol/l ( ) at baseline to 5.75 mmol/l ( ) at week 12 (median [IQR] change from baseline -.73 mmol/l [ ]; P<.1). No notable change was observed for the ABC/3TC group. The difference between groups at week 12 was -.82 mmol/l (P<.1). For TDF/FTC (but not for ABC/3TC), statistically significant reductions (P<.5) from baseline were observed in total, low-density lipoprotein, high-density lipoprotein (HDL)- and non-hdl cholesterol (at weeks 4 and 12). Statistically significant decreases were observed in median estimated creatinine clearance (Cockcroft Gault) from baseline to week 12 for patients who switched to TDF/FTC (-5.47 ml/ min) versus the ABC/3TC group (-2.15 ml/min; P=.16 between groups). Virological suppression was maintained in both groups. No new safety issues were identified. Conclusions: Switching to TDF/FTC from ABC/3TC was associated with rapid improvements in fasting lipid parameters and continued virological control in patients receiving LPV/r as the third component of antiretroviral therapy. The effect of these changes on clinical end points remains unclear and would need to be evaluated in a longer-term study. Introduction HAART regimens have markedly improved the prognosis for patients with HIV infection, reducing the number of patients progressing to full-blown AIDS and death [1,2]. However, long-term treatment can be associated with side effects of some of the agents administered during HAART. Dyslipidaemia has emerged as an important issue in patients who are receiving HAART and, if not addressed, it can represent a significant cardiovascular risk [3 5]. The D:A:D study showed an increase in the risk of myocardial infarction in HIV-infected patients receiving HAART [6] and indicated that cumulative use of protease inhibitors (PIs) may be associated with an increased risk of myocardial infarction [7], for which dyslipidaemia was a possible contributing factor. However, the strength of the association between risk of myocardial infarction and PIs was reduced when controlling for exposure to nucleoside reverse transcriptase inhibitors (NRTIs) [7]. 12 International Medical Press (print) -58 (online) 111

2 G Behrens et al. More recent analyses from this cohort implied that an unexpected, elevated risk of myocardial infarction was present for patients receiving (or having recently received) the NRTIs abacavir (ABC) or didanosine [8]. Although a similar association was observed in the STEAL study, which was conducted in 3 patients over 96-weeks and compared ABC plus lamivudine (3TC) with tenofovir disoproxil fumarate (TDF) plus emtricitabine (FTC) [9], other cohort analyses failed to confirm an ABC effect on the incidence of myocardial infarctions [1 12]. A number of studies have focused on plasma lipid improvements associated with switching from thymidine analogues to newer NRTIs. In switch studies that analysed the effects of changing an NRTI to either ABC or TDF, results suggested that ABC tended to maintain the dyslipidaemic changes of previous NRTIs or to have a less positive effect than TDF. A more marked and consistent effect was indeed associated with TDF in many studies [13 16]. The RAVE study compared switching a thymidine nucleoside analogue to either ABC or TDF [14]; no restrictions were placed on other components of the participants regimens. This 48-week study enrolled virologically suppressed patients with no eligibility criteria for fasting lipid levels. Patients who switched to TDF experienced reductions in mean total cholesterol and mean low-density lipoprotein (LDL) cholesterol, whereas lipid levels in those who switched to ABC remained unchanged (statistically significant differences between the two treatment groups for total cholesterol [P=.3] and LDL [P=.4]). The BICOMBO study compared switching virologically suppressed patients from a 3TCcontaining regimen to ABC plus 3TC or TDF plus FTC, both administered with any PIs or non-nrti. After 48 weeks, reductions were observed in median total cholesterol and LDL-cholesterol for TDF/FTC compared with increases (or no change) for ABC/3TC; the differences between the treatments were statistically significant [16]. Even addition of TDF to an effective regimen led to some improvement in lipid parameters [17]. The current study was the first to recruit patients with elevated total cholesterol who were already receiving ABC/3TC to investigate whether switching the NRTI component of HAART to TDF/FTC while maintaining the PI component had an effect on blood lipid parameters after 12 weeks. The study also intended to investigate whether an improvement in lipid profile had a beneficial effect on the overall cardiovascular risk profile for patients receiving HAART. Methods Study population The study enrolled adults ( 18 years) with plasma HIV-1 RNA<5 copies/ml for 12 weeks prior to and at screening who had been receiving stable HAART consisting of ABC, 3TC and ritonavir-boosted lopinavir (LPV/r) for 24 weeks prior to screening. Patients were to have fasted total cholesterol 5.2 mmol/l ( mg/ dl) at screening and at their two previous tests conducted 4 weeks apart. Patients could take lipid- regulating therapy during the study; however, any patient was required to be on a stable dose and frequency of dosing for 12 weeks before screening, which was expected to remain unchanged throughout the study. Study design This was an open-label randomized two-group 12-week controlled study. The study was conducted from 29 September 8 to 19 October 9 at 1 centres in Italy, 1 centres in Spain, 6 centres in Germany and 3 centres in Austria (ClinicalTrials.gov number NCT77292). All patients provided written informed consent. The study was conducted in accordance with the International Conference on Harmonization Guideline for Good Clinical Practice and the Declaration of Helsinki. Ethics committees approved the study in all participating countries. Patients attended the site for randomization and baseline assessments within 4 weeks of screening, followed by visits at weeks 4 and 12. Patients were randomized to either switch to a fixed-dose combination tablet of TDF 3 mg/ftc mg (Truvada, provided by Gilead Sciences, Ltd, Blackrock, Ireland) oncedaily plus the unchanged fixed-dose combination tablet containing lopinavir mg/ritonavir 5 mg (Kaletra, Aesica Queenborough, Ltd, Queenborough, UK) as prescribed, or to continue taking a fixed-dose combination tablet containing ABC mg/3tc 3 mg (Kivexa, Glaxo Operations UK, Ltd, Ware, UK) once-daily plus a fixed-dose combination tablet containing lopinavir mg/ritonavir 5 mg (Kaletra ) as prescribed. Patients were told to take TDF/FTC with food and to continue taking ABC/3TC (if so randomized) and LPV/r in the same manner as previously. Initiation of the switch treatment of TDF/FTC was to take place within 24 h of the baseline visit. In both groups, treatment with LPV/r was unchanged. Adherence to TDF/FTC was assessed by pill counting. Patients were requested to continue their current lifestyle, in terms of diet and physical activity, and to attend study visits in a fasted state. All blood and urine specimens were analysed by a central laboratory. Electronic case report forms were completed for each patient. Assessments of lipid parameters and antiretroviral efficacy Change from baseline in fasting total cholesterol at week 12 was the primary efficacy end point. Changes in fasting triglycerides, low density lipoprotein (LDL) cholesterol (by direct measurement), high-density International Medical Press

3 Switching from ABC/3TC to TDF/FTC lipoprotein (HDL) cholesterol, non-hdl cholesterol and ratio of fasting total cholesterol to HDL cholesterol were assessed as secondary end points. Antiretroviral efficacy end points included the proportion of patients with HIV-1 RNA<5 copies/ml and changes in CD4 + T-cell count. Lipid parameters, plasma HIV-1 RNA and CD4 + T-cell counts were determined according to standard procedures by a central lab. Assessment of outcome measure The Framingham 1-year risk for coronary heart disease (CHD) was calculated for each patient [18]. Safety assessments Safety and tolerability were assessed using adverse events (coded using the Medical Dictionary for Regulatory Activities Version 12. [19], MedDRA is a registered trademark of the International Federation of Pharmaceutical Manufacturers and Associations [IFPMA]), physical examinations (including vital signs) and assessment of standard laboratory end points (haematology, clinical chemistry and urinalysis) and fasting glucose at each study visit. Creatinine clearance was estimated using the Cockcroft Gault equation and the modification of diet in renal disease (MDRD) formula. Investigators graded adverse event intensity and relationship to study drug. Statistical methods The original protocol assumed an estimated mean ±sd difference of.5 ±1.1 mmol/l between the two treatment groups. A sample size of 8 patients per group would provide approximately 8% power to detect a treatment difference. The study was prematurely terminated due to enrolment issues at a total of 85 subjects. The lower sample size increases the risk of a false negative decision for the hypothesis being tested. The intent-to-treat (ITT) population included patients who were randomized, received 1 dose of study drug, and had 1 post-baseline observation; patients were grouped by assigned treatment. The ITT population was used for the analyses of disposition, fasting lipid parameters, HIV-1 RNA, and CD4 + T-cell end points. The treated population included all patients who received 1 dose of study drug; patients were grouped by treatment received. The treated population was used for the analysis of lipid parameters and safety data. The ITT population, which grouped patients by assigned treatment, was the protocol-defined primary analysis set for evaluating fasting total cholesterol (primary end point) and other lipid parameters. However, one patient who was randomized to ABC/3TC was mistakenly dosed with TDF/FTC. Therefore, results of lipid parameters presented in this manuscript focus on the treated analysis set, which grouped patients by treatment received. The only lipid parameter for which data are presented from the ITT set is fasting total cholesterol. The null hypothesis was that no effect on fasting total cholesterol would be observed by switching treatment. The alternative hypothesis was that switching to TDF/ FTC would have an effect (fasting total cholesterol change from baseline to week 12 in the TDF/FTC group was not equal to total cholesterol change from baseline in the ABC/3TC group). The Wilcoxon rank sum test was used for the statistical comparisons between treatment groups, and 95% CIs were constructed for the mean treatment differences. Within-group comparisons of total cholesterol at week 12 versus baseline were analysed using the Wilcoxon signed rank test. Primary results presented for fasting lipid parameters are from the treated analysis set using observed data. Sensitivity analyses were conducted using the following methods: treated analysis set using the last observation carried forward (LOCF) method (missing values at week 12 were imputed by the last observed post-baseline result [prior to week 12]), ITT analysis set using the LOCF method and ITT analysis set using the LOCF method, excluding data for one subject in the TDF/FTC group who started/modified lipid-lowering medication during the study. Number and percentage of patients with plasma HIV-1 RNA<5 copies/ml were used to summarize the viral load end points; missing values were considered as excluded. Comparisons between groups of percentage of patients with viral load suppression were conducted using Fisher s exact test, with a 95% CI for the between-group difference based on an unconditional exact method using inverted two one-sided tests. For CD4 + T-cell count, differences between treatment groups in change from baseline were tested using the Wilcoxon rank sum test at week 12 (missing values were excluded). Within-group comparisons of the change in Framingham 1-year risk of CHD at week 12 were analysed using the Wilcoxon signed rank test; between-group comparisons were analysed using the Wilcoxon rank sum test. Patients were categorised according to their level of Framingham risk at baseline and week 12 as follows: <1%, 1 % and >%. Safety data collected up to 3 days after the last dose of study drug were summarized; no statistical testing was performed. SAS (SAS Institute Inc., Cary, NC, USA) StatXact (Cytel, Cambridge, MA, USA) and nquery Advisor (Version 6.; Statistical Solutions, Cork, Ireland) software were used for statistical analyses. Results Results are presented for the treated population unless otherwise specified. Antiviral Therapy

4 G Behrens et al. Figure 1. Flow chart of patient disposition and analysis populations n=11 Screened n=85 Randomized and treated a n=15 Screening failures n=1 Passed screening but withdrew consent and was not randomized n=42 TDF/FTC+LPV/r n=41 Completed study n=1 Discontinued treatment prematurely (n=1 Adverse event) n=43 ABC/3TC+LPV/r n= Completed study n=3 Discontinued treatment prematurely (n=1 Withdrew consent; n=2 protocol violation) a One patient was randomized to the abacavir (ABC)/lamivudine (3TC) group but received tenofovir disoproxil fumarate (TDF)/emtricitabine (FTC); this patient was included in the ABC/3TC group in this figure and in the randomized and intent-to-treat populations. LPV/r, lopinavir/ritonavir. Patient disposition and baseline characteristics A total of 85 patients were randomized: 42 to receive TDF/FTC and 43 to receive ABC/3TC (ITT population; Figure 1). One patient randomized to ABC/3TC received TDF/FTC, therefore, the treated population comprised 43 patients for TDF/FTC and 42 patients for ABC/3TC. Four patients in the ITT population discontinued the study prematurely: one patient in the TDF/FTC group (adverse event) and three patients in the ABC/3TC group (one withdrew consent and two were protocol violators). Groups were well balanced in terms of baseline characteristics, including baseline use of lipid-regulating therapies (Table 1). All patients had plasma HIV-1 RNA levels <5 copies/ml at screening, although a total of seven patients had levels 5 copies/ml at the subsequent baseline visit. All of these patients were retested and found to have levels <5 copies/ml. One patient in the TDF/FTC group and none in the ABC/3TC group started a lipid-regulating agent during the study (pravastatin). Excluding this patient for a sensitivity analysis of the changes in total cholesterol did not change the P-value of the results. Fasting lipid parameters Total cholesterol in the TDF/FTC group declined significantly after adjusting for changes in the ABC/3TC group (mean change -.82, 95% CI -1.22, -.43; P<.1; ITT population, LOCF). Within each group, median (IQR) total cholesterol levels in the TDF/FTC group declined from 6.22 mmol/l ( ) at baseline to 5.75 mmol/l ( ) at week 12 (ITT population, LOCF); this change was statistically significant (median [IQR] change from baseline -.73 mmol/l [ ]; P<.1). No statistically significant change in total cholesterol was observed for ABC/3TC: median (IQR) values for total cholesterol were 6.42 mmol/l ( ) at baseline and 6.53 mmol/l ( ) at week 12 (median [IQR] change from baseline -.1 mmol/l [ ]; ITT population, LOCF). These findings were confirmed by all three sensitivity analyses (data not shown). Patients who switched to TDF/FTC had decreases in all lipid parameters: total cholesterol, triglycerides, LDL cholesterol, HDL-cholesterol, non-hdl cholesterol and the ratio of total cholesterol to HDL cholesterol (Table 2). These changes were statistically significant (P<.5; observed data, treated population) at week 4 and week 12 for total cholesterol, LDL cholesterol, HDL cholesterol and non-hdl cholesterol (Table 2). Lipid parameters generally remained unchanged for patients who stayed on ABC/3TC; no statistically significant change was observed through week 12. Differences between TDF/FTC and ABC/3TC were statistically significant at week 4 and week 12 in change from baseline in total cholesterol, LDL cholesterol, HDL cholesterol and non-hdl cholesterol (Table 2). Using National Cholesterol Education Program (NCEP) categories, the number of patients who switched to TDF/FTC with fasting total cholesterol levels <5.2 mmol/l (< mg/dl; NCEP category desirable ) International Medical Press

5 Switching from ABC/3TC to TDF/FTC Table 1. Characteristics of patients at baseline, intent-to-treat dataset Characteristic TDF/FTC (n=42) ABC/3TC (n=43) Total (n=85) Male sex, n (%) 35 (83) 31 (72) 66 (78) Median age, years (IQR) 46. (38 51) 43. (38 48) 44. (38 49) Median time since diagnosis, years (IQR) 7 (4 13) 6 (4 15) 7 (4 14) HIV status Asymptomatic, n (%) 25 () 27 (63) 52 (61) Symptomatic HIV infection, n (%) 6 (14) 2 (5) 8 (9) AIDS, n (%) 11 (26) 14 (33) 25 (29) HIV-1 RNA<5 copies/ml, n (%) 37 (88) 41 (95) 78 (92) Median CD4 + T-cell count, cells/mm 3 (IQR) 57 ( ) 525 ( ) 513 ( ) Median weight, kg (IQR) 72.5 (67. 8.) 77. ( ) 74. ( ) Median BMI, kg/m 2 (IQR) 25.3 ( ) 25.2 ( ) 25.2 ( ) Dyslipidaemia a Triglycerides 2.26 mmol/l, n (%) 25 () 25 (58) 5 (59) LDL cholesterol 4.1 mmol/l, n (%) 18 (43) 19 (44) 37 (44) Lipids Median total cholesterol, mmol/l (IQR) 6.22 ( ) 6.42 ( ) 6.24 ( ) LDL cholesterol, mmol/l (IQR) 3.86 ( ) 4.7 ( ) 3.97 ( ) HDL cholesterol, mmol/l (IQR) 1.33 ( ) 1.25 ( ) 1.27 ( ) Triglycerides, mmol/l (IQR) 2.62 ( ) 2.65 ( ) 2.65 ( ) Other CV risk factors Smoker, n (%) 22 (52) 21 (49) 43 (51) Diabetes, n (%) Median Framingham 1-year risk score (IQR) 8 (4 13) 7 (3 1) 7 (4 1) Most frequent lipid modifying agents, n (%) b 1 (23) 1 (24) (24) Pravastatin, n (%) 3 (7) 6 (14) 9 (11) Fish oil, n (%) 3 (7) 2 (5) 5 (6) a Assessed on a treated basis and therefore patients were n=43 for tenofovir disoproxil fumarate (TDF)/emtricitabine (FTC) and n=42 for abacavir (ABC)/lamivudine (3TC). b Only agents used by >5% of total patients presented. Other agents were atorvastatin, omega-3 fatty acids (both total n=3), ezetimibe and fenofibrate (both n=1). BMI, body mass index; CV, cardiovascular; HDL, high-density lipoprotein; LDL, low-density lipoprotein. increased from 3 (7.%) at baseline to 11 (28.9%) at week 12, compared with 3 (7.1%) and 2 (5.%) patients, respectively, for ABC/3TC. The number of patients with fasting LDL cholesterol >4.1 (>1 mg/dl; NCEP category high ) declined from 19 (44.2%) at baseline to 5 (13.2%) at week 12 for TDF/FTC, compared with 18 (42.9%) and 16 (.%) patients, respectively, for ABC/3TC. The number of patients with fasting HDL cholesterol <1.3 mmol/l (< mg/dl; NCEP category low ) remained almost stable at 8 (18.6%) at baseline compared to 9 (23.7%) at week 12 for TDF/FTC, similar to 7 (16.7%) and 9 (22.5%) patients, respectively, for ABC/3TC (Figure 2). Virological and immunological efficacy No statistically significant differences were observed between the TDF/FTC and ABC/3TC groups at week 12 in the proportions of patients with plasma HIV-1 RNA<5 copies/ml (TDF/FTC 34/38 [9%] and ABC/3TC 37/39 [95%]; P=.43; patients missing ITT data were excluded). In general, increases of plasma HIV-1 RNA>5 copies/ml were isolated occasions and levels returned to <5 copies/ml. CD4 + T-cell counts at week 12 were similar to those at baseline for both groups; median (IQR) change from baseline was 17 cells/mm 3 (-53 65) for the TDF/FTC group, and 7 cells/mm 3 ( ) for the ABC/3TC group. Safety results A total of 19 (44%) patients who switched to TDF/ FTC and 11 (26%) patients who continued to receive ABC/3TC experienced treatment-emergent adverse events. There were no clinically relevant differences between groups in the profiles of adverse events. Events reported by >1 patient in either treatment group were nasopharyngitis (TDF/FTC, three patients and ABC/3TC, four patients); headache (TDF/FTC, two patients and ABC/3TC, one patient); conjunctivitis (TDF/FTC, one patient and ABC/3TC, two patients) and diarrhoea (TDF/FTC, one patient and ABC/3TC, two patients). No ABC-related rash was observed (or expected) as eligible patients had been taking ABC for 24 weeks at the start of the study. One patient who received ABC/3TC (moderate increased blood cholesterol) and two patients who received TDF/FTC (moderate gastritis in one patient and mild renal pain and Antiviral Therapy

6 G Behrens et al. pruritus in another) experienced adverse events judged by the investigator to be related to study drug. Two patients experienced serious adverse events. A 43-year-old woman who switched to TDF/FTC was diagnosed with retroperitoneal sarcoma and was withdrawn from the study. A 38-year-old woman who continued to receive ABC/3TC was diagnosed with fibroadenoma of the right breast (benign on histological examination). The patient continued in the study. Neither of these two serious adverse events was considered related to study drug. Median estimated creatinine clearance values were similar for the two treatment groups at baseline and at week 12 (Table 3). Statistically significant decreases were observed in median estimated creatinine clearance from baseline to week 12 for patients who switched to TDF/FTC; there were no statistically significant changes in estimated creatinine clearance for the ABC/3TC group. Shifts from normal renal function (MDRD formula estimated creatinine clearance 9 ml/min) to mild renal impairment (MDRD formula estimated creatinine clearance 89 ml/min) occurred in 9 (.9%) patients in the TDF/FTC group and in 7 (16.7%) patients in the ABC/3TC group. One patient in the TDF/FTC group shifted from mild to moderate renal impairment (MDRD formula estimated creatinine clearance 3 59 ml/min); however, the estimated creatinine clearance of this patient had been 55.4 ml/min at screening,.3 ml/ min at baseline and 55.3 ml/min at week 12. All other patients remained within their baseline categories for MDRD formula estimated creatinine clearance. When using the Cockcroft Gault equation for calculating estimated creatinine clearance two patients in the ABC/3TC group shifted from ml/min to < ml/ min and none in the TDF/FTC group shifted impairment level. There were no notable changes in mean body weight over the course of the study. Median (IQR) adherence to TDF/FTC during the study was 1% (99 1%); adherence was 95% for the majority (88%) of patients. Only small changes were observed between groups in 12-week mean (sd) change from baseline in Framingham 1-year risk score for CHD (TDF/FTC -.7% [2.42%]; P=.34; ABC/3TC.5% [2.51%]; P=.33; between-group P=.27). Discussion Dyslipidaemia has been shown to be a complication of HAART regimens [8]. In the current study, patients with increased total cholesterol who switched to the TDF/FTC combination tablet from ABC/3TC, while maintaining LPV/r, experienced rapid reductions in fasting lipid parameters. Statistically significant reductions were observed already from week 4 onwards. By contrast, patients who stayed on ABC/3TC did not Table 2. Changes from baseline in fasting lipid parameters TDF/FTC ABC/3TC within- TDF/FTC versus ABC/3TC Parameter TDF/FTC (n=43) within-group P-value a ABC/3TC (n=42) group P-value a between group P-value b Total cholesterol at baseline 6.23 ( ) 6.36 ( ) Change at week (-1..6) P<.1.8 ( ) P=.38 P<.1 Change at week ( ) P< ( ) P=.75 P<.1 Triglycerides at baseline 2.66 ( ) 2.63 ( ) Change at week ( ) P=.1.7 ( ) P=.25 P=.55 Change at week ( ) P= ( ) P=.8 P=.21 LDL cholesterol at baseline 3.88 ( ) 4.5 ( ) Change at week (-.76.) P=.2.6 (-..37) P=1. P=.35 Change at week ( ) P< ( ) P=.91 P=.5 HDL cholesterol at baseline 1.32 ( ) 1.24 ( ) Change at week (-.24.2) P<.1.1 (-.1.9) P=.91 P=.11 Change at week (-.19.) P< ( ) P=.8 P=.26 Non-HDL cholesterol at baseline 4.94 ( ) 4.95 ( ) Change at week ( ) P<.1.17 ( ) P=.31 P<.1 Change at week ( ) P< ( ) P=.73 P<.1 Total cholesterol:hdl cholesterol 4.9 ( ) 5.1 (3.7 6.) ratio at baseline Change at week (-.5.4) P=.61. (-.4.6) P=.53 P=.39 Change at week (-.6.1) P= (-.5.7) P=.72 P=.18 Data on triglycerides are median (IQR) and units are mmol/l and represent the treated dataset, with missing data excluded. a Wilcoxon signed-rank test for withingroup comparison of change from baseline. b Wilcoxon rank-sum test for comparison between groups. ABC, abacavir; FTC, emtricitabine; HDL, high-density lipoprotein; LDL, low-density lipoprotein; TDF, tenofovir disoproxil fumarate; 3TC, lamivudine International Medical Press

7 Switching from ABC/3TC to TDF/FTC experience any notable changes. The observed changes in total cholesterol were accompanied by changes in LDL cholesterol, HDL cholesterol and non-hdl cholesterol. The proportion of patients achieving the total cholesterol concentration considered desirable by the National Cholesterol Education Program Adult Panel III (<5.2 mmol/l) [] increased only in patients who switched to TDF/FTC. Similar shifts in the desirable direction were also seen for LDL cholesterol. High levels of HDL cholesterol can be beneficial in terms of cardiovascular risk. HDL cholesterol levels were also reduced in patients who switched to TDF/FTC. However, the number of patients with HDL cholesterol levels considered a risk factor according to NCEP categories (<1.3 mmol/l) remained almost stable in both arms. The impact of these changes on clinical end points remains unclear and would need to be evaluated in a longer-term study. Potential risks or benefits for patients would probably be needed to be assessed on an individual patient basis taking into account the lipid parameters and the wide range of additional risk factors, including renal impairment. There are some aspects of the study design that should be considered when assessing the results of Figure 2. Categorical changes of treated patients in lipid parameters ABC/3TC TDF/FTC Patients, % Patients, % Patients, % A C E <5.2 <2.6 Total cholesterol baseline 5.2 < < Total cholesterol, mmol/l LDL baseline 2.6 < < LDL, mmol/l HDL baseline Patients, % Patients, % Patients, % B D F <5.2 <2.6 Total cholesterol week < < Total cholesterol, mmol/l LDL week < < LDL, mmol/l HDL week 12 < <1.55 HDL, mmol/l 1.55 < < HDL, mmol/l Results are displayed using National Cholesterol Education Program categories. Total fasting cholesterol (A) at baseline and (B) at week 12. Fasting low-density lipoprotein (LDL) cholesterol (C) at baseline and (D) at week 12. Fasting high-density lipoprotein (HDL) cholesterol (E) at baseline and (F) at week 12. ABC, abacavir; FTC, emtricitabine; TDF, tenofovir disoproxil fumarate; 3TC, lamivudine. Antiviral Therapy

8 G Behrens et al. Table 3. Estimated creatinine clearance, treated dataset TDF/FTC versus ABC/3TC Parameter TDF/FTC (n=43) ABC/3TC (n=42) between group P-value a Cockcroft Gault equation (observed weight) Baseline, ml/min ( ) ( ) Week 12, ml/min ( ) ( ) Change from baseline at week 12, ml/min ( ) ( ) P=.16 Within-group P-value b P=.4 P=.56 Cockcroft Gault equation (ideal weight) Baseline, ml/min ( ) 15.5 ( ) Week 12, ml/min 91.9 ( ) 14. ( ) Change from baseline at week 12, ml/min -.26 ( ) -.28 ( ) P=.33 Within-group P-value b P=.9 P=.39 Modification of diet in renal disease formula Baseline, ml/min/1.73 m ( ) 98.1 ( ) Week 12, ml/min/1.73 m ( ) 97.7 ( ) Change from baseline at week 12, -.15 ( ) -.1 ( ) P=.17 ml/min/1.73 m 2 Within-group P-value b P=.6 P=.37 Data are median (IQR). a Wilcoxon rank-sum test for comparison between groups at week 12. b Wilcoxon signed-rank test for within-group comparison of change from baseline. ABC, abacavir; FTC, emtricitabine; TDF, tenofovir disoproxil fumarate; 3TC, lamivudine. the study. This was an open-label study. However, the study end points reported here, including the primary end point, were objective measures assessed by a central laboratory. The current study was 12 weeks in duration and considered sufficient to detect any effect of switching from ABC/3TC to TDF/FTC on elevated total cholesterol levels and lipid profiles. A previous study that evaluated changes in lipid levels on switching from zidovudine/3tc to TDF/FTC detected notable decreases in total cholesterol and triglycerides after only 4 weeks of treatment in patients who switched to TDF/FTC; between-group comparisons achieved statistical significance from week 4 until week 36 [21]. A potential limitation was the recruitment of only 85 patients, when the original plan was to study 1 patients. Recruitment was slow because the number of patients taking ABC/3TC at our study centres was already declining at the time of patient recruitment. We believe this was a consequence largely of the published data on potential ABC cardiovascular risks [8] that were described in both the EU and the US treatment guidelines at that time [5,22]. A number of strategies were attempted to increase patient numbers: additional sites were added, the recruitment period was extended, the feasibility of including additional countries in the study was investigated but discounted and the potential inclusion of patients receiving any PI rather than just those receiving LPV/r was also investigated, but this was discounted as it would have required stratification of the patient population. The statistical power to detect treatment differences was less than optimum as a consequence of the smaller than planned sample size. However, both the observed magnitude of difference between groups in this study was bigger than that originally assumed (.8 mmol/l observed versus.5 mmol/l in the protocol), and the observed sd in the study was smaller than that originally assumed (.9 mmol/l observed versus 1.1 mmol/l in the protocol). These two factors together could have contributed to achieving statistical significance when comparing changes in lipid concentrations for the TDF/ FTC and the ABC/3TC groups. The findings of the current study are similar to those of the TOTEM study in which dyslipidaemic patients who switched two NRTIs to TDF/FTC improved their lipid parameters [23]. The TOTEM study differed from the current study in that the comparator group comprised patients receiving HAART that could include any NRTI rather than ABC/3TC and the third component of the regimen could be a PI or an NNRTI. In the TOTEM study, patients who switched to TDF/FTCcontaining HAART demonstrated statistically significant decreases in triglycerides and LDL cholesterol after 12 weeks of treatment. The observation of a slight, but statistically significant, reduction in HDL cholesterol in our study is consistent with the TOTEM study [23] and has also been reported previously [24 27]. The reduction in HDL cholesterol explained why the total cholesterol to HDL cholesterol ratio did not change significantly during the current study even though statistically significant reduction was achieved in total cholesterol. No new safety issues were identified in the current study and virological control was maintained after switching to TDF/FTC, in line with previous findings International Medical Press

9 Switching from ABC/3TC to TDF/FTC Statistically significant reductions from baseline in estimated creatinine clearance were observed at week 12 in the TDF/FTC arm, however, the observed 12-week median values were above the threshold marking a normal level of estimated glomerular filtration rate (>9 ml/ min/1.73m 2 ), according to The Renal Association criteria [28]. Even the categorical changes were not considerably different between groups. Reductions in estimated creatinine clearance have been reported in other prospective studies of TDF-containing regimens [29 32], which, after an initial decline, tend to stabilize over time. More recent observational cohort data showed an increasing incidence of chronic kidney disease (estimated glomerular filtration rate ml/min/1.73m 2 for 3 months) with increasing cumulative exposure to TDF [33] and we wish to emphasize that more careful monitoring of kidney function is recommended when starting TDF-containing therapies. In conclusion, statistically significant reductions were observed in fasting total cholesterol, LDL cholesterol, HDL cholesterol and non-hdl cholesterol for patients who switched the NRTI component of their HAART regimen to TDF/FTC from ABC/3TC, while maintaining LPV/r as the third component of antiretroviral therapy. Our findings indicate that treatment with TDF/FTC can contribute to the reduction of elevated lipid concentrations in dyslipidaemic patients who will require many years of antiretroviral therapy. We propose the switch to TDF/FTC from ABC/3TC as an effective strategy to improve dyslipidaemia in patients in which LPV/r therapy cannot be changed because of resistance mutation or tolerability issues. Acknowledgements The authors would like to thank all the patients and study staff that participated in this study (ClinTrials. gov identifier NCT77292). Disclosure statement This clinical trial was sponsored by Gilead Sciences. The study design was developed in cooperation with the lead author GB who had access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. GB has received financial support (grants, speaker/advisor honoraria and travel) from Abbott, Boehringer Ingelheim, Bristol Myers Squibb, Gilead Sciences, GlaxoSmithKline, Merck, Pfizer, Teratechnology and Tibotec. RM has received financial support (grants, speaker/ advisor honoraria and travel) from Abbott, Boehringer Ingelheim, Bristol Myers Squibb, Gilead Sciences, GlaxoSmithKline, Merck, Roche, Pfizer and Tibotec. AR has received financial support (grants, speaker/ advisor honoraria and travel) from Abbott, Boehringer Ingelheim, Bristol Myers Squibb, Gilead Sciences, GlaxoSmithKline, Merck, Pfizer, Roche and Tibotec. PD has received financial support (grants, speaker/ advisor honoraria and travel) from Boehringer Ingelheim, Bristol Myers Squibb, Janssen-Cilag, Gilead Sciences, GlaxoSmithKline, Merck and Pfizer. HW, FA and GP are full-time employees of Gilead Sciences. All listed coauthors meet criteria for authorship. Specifically, the following contributions were made by the authors listed in parentheses: conception and design (GB, FA, GP and HW), country coordinating investigators and acquisition of data (GB, RM, AR and PD), drafting of the manuscript (GB, FA, GP and HW), critical revision of manuscript (all authors), statistical analysis (HW). We thank Justin Cook of Niche Science and Technology, Ltd for editorial assistance with the manuscript. References 1. Palella FJ, Delanay KM, Moorman AC, et al. Declining morbidity and mortality among patients with advanced HIV infection. N Engl J Med 1998; 338: Mocroft A, Vella S, Benfield TL, et al. Changing patterns of mortality across Europe in patients infected with HIV. Lancet 1998; 352: Dubé MP, Stein JH, Aberg JA, et al. Guidelines for the evaluation and management of dyslipidaemia in human immunodeficiency virus (HIV)-infected adults receiving antiretroviral therapy: recommendations of the HIV Medical Association of the Infectious Disease Society of America and the Adult AIDS Clinical Trials Group. Clin Infect Dis 3; 37: Grover SA, Coupal L, Gilmore N, et al. Impact of dyslipidaemia associated with highly active antiretroviral therapy (HAART) on cardiovascular risk and life expectancy. Am J Cardiol 5; 95: European Guidelines for treatment of HIV infected adults in Europe, Version 6. (Updated October 11. Accessed 31 May 12.) Available from europeanaidsclinicalsociety.org/images/stories/eacs-pdf/ EACSGuidelines-v6.-English.pdf 6. D:A:D Study Group. Combination antiretroviral therapy and the risk of myocardial infarction. N Engl J Med 3; 349: D:A:D Study Group. Class of antiretroviral drugs and the risk of myocardial infarction. N Engl J Med 7; 356: D:A:D Study Group. Use of nucleoside reverse transcriptase inhibitors and risk of myocardial infarction in HIV-infected patients enrolled in the D:A:D study: a multi-cohort collaboration. Lancet 8; 371: Cooper DA, Bloch M, Humphries A, et al. Simplification with fixed-dose tenofovir-emtricitabine or abacavirlamivudine in adults with suppressed HIV replication (the steal study): a randomized, open-label, 96-week, non-inferiority trial. Conference on Retroviruses and Opportunistic Infections February 9, Montreal, QC, Canada. Poster Behrens GM, Reiss P. Abacavir and cardiovascular risk. Curr Opin Infect Dis 1; 23: Lang S, Mary-Krause M, Cotte L, et al. Impact of individual antiretroviral drugs on the risk of myocardial infarction in human immunodeficiency virus-infected patients: a casecontrol study nested within the French Hospital Database on HIV ANRS cohort CO4. Arch Intern Med 1; 17: Antiviral Therapy

10 G Behrens et al. 12. Ribaudo HJ, Benson CA, Zheng Y, et al. No risk of myocardial infarction associated with initial antiretroviral treatment containing abacavir: short and long-term results from ACTG A51/ALLRT. Clin Infect Dis 11; 52: Fisher M, Moyle GJ, Shahmanesh M, et al. A randomized comparative trial of continued zidovudine/lamivudine or replacement with tenofovir disoproxil fumarate/ emtricitabine in efavirenz-treated HIV-1-infected individuals. J Acquir Immune Defic Syndr 9; 51: Moyle GJ, Sabin CA, Cartledge J, et al. A randomized comparative trial of tenofovir DF or abacavir as replacement for a thymidine analogue in persons with lipoatrophy. AIDS 6; : Martínez E, Arranz JA, Podzamczer D, et al. A simplification trial switching from nucleoside reverse transcriptase inhibitors to once-daily fixed-dose abacavir/lamivudine or tenofovir/ emtricitabine in HIV-1-infected patients with virological suppression. J Acquir Immune Defic Syndr 9; 51: Martínez E, Larrousse M, Podzamczer D, et al. Abacavir-based therapy does not affect biological mechanisms associated with cardiovascular dysfunction. AIDS 1; 24:F1 F Tungsiripat M, Kitch D, Glesby MJ, et al. A pilot study to determine the impact on dyslipidemia of adding tenofovir to stable background antiretroviral therapy: ACTG 56. AIDS 1; 24: Framingham Heart Study. Coronary heart disease (1-year risk). (Updated 23 April 12. Accessed 5 May 12.) Available from coronary.html 19. Medical Dictionary for Regulatory Acitvities Version 12.. (Updated 28 June 12. Accessed 9 July 12.) Available from Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive summary of the third report of the National Cholesterol Education Program (NCEP) Expert Panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III). JAMA 1; 285: Fisher M, Moyle GJ, Shahmanesh M, et al. A randomised comparative trial of continued zidovudine/lamivudine or replacement with tenofovir disoproxil fumarate/ emtricitabine in efavirenz-treated HIV-1-infected individuals. J Acquir Immune Defic Syndr 9; 51: Panel on Antiretroviral Guidelines for Adults and Adolescents. Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents. Department of Health and Human Services. (Updated 3 November 8. Accessed 25 January 1.) Available from aidsinfo.nih.gov/contentfiles/adultandadolescentgl.pdf 23. Valantin MA, Bittar R, de Truchis P, et al. Switching the nucleoside reverse transcriptase inhibitor backbone to tenofovir disoproxil fumarate + emtricitabine promptly improves triglycerides and low-density lipoprotein cholesterol in dyslipidaemic patients. J Antimicrob Chemother 1; 65: Lundgren JD, Battegay M, Behrens G, et al. European AIDS Clinical Society (EACS) guidelines on the prevention and management of metabolic diseases in HIV. HIV Med 8; 9: Moyle GJ, Sabin CA, Cartledge J, et al. A randomized comparative trial of tenofovir DF or abacavir as replacement for a thymidine analogue in persons with lipoatrophy. AIDS 6; : Domingo P, Labarga P, Palacios R, et al. Improvement of dyslipidaemia in patients switching from stavudine to tenofovir: preliminary results. AIDS 4; 18: Llibre JM, Domingo P, Palacios R, et al. Sustained improvement of dyslipidaemia in HAART-treated patients replacing stavudine with tenofovir. AIDS 6; : The Renal Association. The UK eckd Guide. (Updated January 9. Accessed 5 May 12.) Available from CKDeGUIDE.aspx 29. Gallant JE, Parish MA, Keruly JC, et al. Changes in renal function associated with tenofovir disoproxil fumarate, compared with nucleoside reverse transcriptase inhibitor treatment. Clin Infect Dis 5; : Gallant JE, DeJesus E, Arribas JR, et al. Tenofovir DF, emtricitabine, and efavirenz vs zidovudine, lamivudine, and efavirenz for HIV. N Engl J Med 6; 354: DeJesus E, Ruane P, McDonals C, et al. Impact of switching virologically-suppressed, HIV-1-infected patients from twice-daily fixed-dose zidovudine/lamivudine to once-daily fixed-dose tenofovir disoproxil fumarate/emtricitabine. HIV Clin Trials 8; 9: Cassetti I, Madruga JVR, Suleiman JMAH, et al. The safety and efficacy of tenofovir DF in combination with lamivudine and efavirenz through 6 years in antiretroviral naive HIV-1 infected patients. HIV Clin Trials 7; 8: Mocroft A, Kirk O, Reiss P, et al. Estimated glomerular filtration rate, chronic kidney disease and antiretroviral drug use in HIV-positive patients. AIDS 1; 24: Accepted 17 January 12; published online 15 August International Medical Press