Lipid profiles for antiretroviral-naive patients starting PI- and NNRTI-based therapy in the Swiss HIV Cohort Study

Transcription

1 Research article Antiviral Therapy 1: Lipid profiles for antiretroviral-naive patients starting PI- and NNRTI-based therapy in the Swiss HIV Cohort Study Jim Young 1 *, Rainer Weber 2, Martin Rickenbach, Hansjakob Furrer, Enos Bernasconi 5, Bernard Hirschel 6, Philip E Tarr 7, Pietro Vernazza 8, Manuel Battegay 9, Heiner C Bucher 1,9 and the Swiss HIV Cohort Study 1 Institut für klinische Epidemiologie, Universitätsspital Basel, Basel, Switzerland 2 Abteilung für Infektionskrankheiten und Spitalhygiene, Universitätsspital Zürich, Zürich, Switzerland Swiss HIV Cohort Study Co-ordination Centre, Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland Klinik und Poliklinik für Infektiologie, Universitätsspital Bern, Bern, Switzerland 5 Servizio di Malattie Infettive, Ospedale Regionale di Lugano, Lugano, Switzerland 6 Division des Maladies Infectieuses, Hôpital Universitaire de Genève, Genève, Switzerland 7 Service des Maladies Infectieuses, Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland 8 Departement Innere Medizin, Kantonsspital St Gallen, St Gallen, Switzerland 9 Klinik für Infektiologie, Universitätsspital Basel, Basel, Switzerland *Corresponding author: Tel: ; Fax: ; jyoung@uhbs.ch Background: Blood lipid abnormalities in patients on highly active antiretroviral therapy (HAART) have been associated with exposure to protease inhibitors (PIs), particularly ritonavir. First therapy with a non-nucleoside reverse transcriptase inhibitor (NNRTI) leads to relatively favourable lipid profiles. We report on medium-term lipid profiles (up to 5 years) for antiretroviral-naive patients starting NNRTI- and PI-based HAART in the Swiss HIV Cohort Study. Methods: Since April 2, blood samples taken at visits scheduled every 6 months have been analysed for cholesterol and triglyceride concentrations. For 165 antiretroviral-naive patients starting HAART after April 2, we estimated changes in concentration over time using multivariate linear regression with adjustment for baseline covariates, use of lipid-lowering drugs and whether the sample was taken in a fasting state. Results: Non-high density lipoprotein (HDL) cholesterol levels increase with increasing exposure to either PI- or NNRTI-based therapy, HDL cholesterol levels increase and triglyceride levels decrease with increasing exposure to NNRTI-based therapy, whereas triglyceride levels increase with increasing exposure to PI-based therapy. Between NNRTI-based therapies, there is a slight difference in triglyceride levels, which tend to increase with increasing exposure to efavirenz and to decrease with increasing exposure to nevirapine. Of the three common PI-based therapies, nelfinavir appears to have a relatively favourable lipid profile, with little change with increasing exposure. Of the other two PI therapies, lopinavir with ritonavir has a more favourable profile than indinavir with ritonavir, with smaller increases in both non-hdl cholesterol and triglycerides and an increase in HDL cholesterol. Increasing exposure to abacavir is associated with a decrease in the level of triglycerides. Conclusion: In general, NNRTI-based therapy is associated with a more favourable lipid profile than PI-based therapy, but different PI-based therapies are associated with very different lipid profiles. Nelfinavir appears to have a relatively favourable lipid profile. Of the two boosted PI therapies, lopinavir appears to have a more favourable lipid profile than indinavir. Introduction Changes in blood lipids occur naturally during the course of HIV infection, with decreases early in the infection in both total cholesterol and high density lipoprotein (HDL) cholesterol and increases later in triglycerides [1]. However, highly active antiretroviral therapy (HAART) also leads to lipid changes with increases in both total cholesterol [2] and triglycerides []. While increases in total cholesterol on therapy may 25 International Medical Press

2 J Young et al. represent a return to pre-infection levels to some degree [2], HAART is associated with an increased risk of myocardial infarction []. Changes in blood lipids have been associated with exposure to protease inhibitors (PIs) particularly ritonavir (RTV) [5,6]. Switching studies suggest that replacing the PI component of therapy with a nonnucleoside reverse transcriptase inhibitor (NNRTI) improves lipid profiles [7,8]. Additionally, in antiretroviral-naive patients, first therapy with either nevirapine (NVP) [9] or efavirenz (EFV) [1] leads to increases in HDL cholesterol. Yet studies of lipid profiles for first therapy with an NNRTI are either of short duration (less than a year) or based on a small number of patients (less than 1). We report on medium-term lipid profiles (up to 5 years) for antiretroviral-naive patients on PI- and NNRTI-based HAART in the Swiss HIV Cohort Study (SHCS). Methods Patients The SHCS is a prospective cohort with continuing enrolment of HIV-infected adults. Since 1 April 2, a cardiovascular risk assessment has been part of followup visits scheduled every 6 months. Blood is analysed for cholesterol, triglyceride, glucose and lactate concentrations; fat loss or gain, weight, hip and waist measurements are also recorded. Our population of interest comprised all antiretroviral-naive patients in the SHCS starting HAART after 1 April 2 with at least one subsequent measurement of a lipid concentration prior to any switch from PI- to NNRTI-based therapy or the reverse. Our sample from this population is those patients with weight, CD cell count and plasma HIV RNA (viral load) measured between 6 months before and months after starting HAART. We define HAART as the combination of at least two nucleoside reverse transcriptase inhibitors (NRTIs) with at least one PI or NNRTI. Statistical analyses We used SAS version 8.2 for all analyses (SAS Institute Inc, Cary, NC, USA). For an unadjusted descriptive analysis, we present non-parametric regression curves showing lipid profiles over time on HAART. We used a local regression procedure with iterative re-weighting (LOESS) so that the fitted curve is relatively robust to outliers [11]. For an adjusted confirmative analysis, we fitted multivariate linear regression models with the level of a blood lipid as the response and with covariates and exposure to antiretrovirals as predictors. For covariates we used: CD cell count, viral load and body mass index (BMI) at baseline, age, gender, family history of cardiovascular disease and intravenous (iv) drug use as the likely mode of transmission all time-independent, and use of lipid-lowering drugs and whether the sample was taken in a fasting state both time-dependent. Exposure to a specific drug class or drug is timedependent and was calculated at each lipid measurement as time to date in units of years. We estimated model parameters using generalized estimating equations, to allow for correlation between repeated measurements from the same patient. We didn t use the baseline measurement of the lipid as a covariate, because there was no reason to expect that the correlation between this measurement and others would be different from the correlation between other measurement pairs [12]. We assumed an independent correlation structure: although other correlation structures may be more efficient, this structure ensures unbiased estimates for time-dependent predictors [1]. Two types of statistic are given for each model. The nature of any association between a blood lipid and exposure to therapy is shown by an estimate of the mean change in the lipid level per year of exposure; 95% confidence intervals are given for each estimate, calculated using the empirical standard error. The evidence for different associations under different therapies is assessed by a generalized score test [1]. For each blood lipid, we considered a sequence of three models. Firstly, for patients starting HAART, we compared PI- and NNRTI-based therapy prior to any switch from one therapy to the other. Secondly, for patients starting NNRTI-based therapy, we compared EFV with NVP prior to any switch from one to the other or to a PI. Finally, for patients starting a common PI-based therapy, we compared nelfinavir (NFV), lopinavir (LPV) with RTV and indinavir (IDV) with RTV prior to any switch to another PI or to an NNRTI. We also adjusted for concomitant NRTI use for those NRTIs that have been linked to changes in blood lipids [15 18]. Firstly, we adjusted separately for time on abacavir (ABC) and time on stavudine (dt). Then, as a sensitivity analysis, we adjusted separately for time on either ABC or tenofovir (TDF) and for time on either dt or zidovudine (AZT) (because AZT and dt have similar effects in vitro [19]). We did not consider TDF separately from ABC because TDF was not yet in common use in this cohort. Results Patients As of 1 January 25, 1565 antiretroviral-naive patients in the SHCS had started HAART after April 2. Of these, 178 patients had at least one blood lipid measurement prior to any switch from PI- to NNRTI-based HAART or the reverse; 165 (77%) had International Medical Press

3 Lipid profiles in the SHCS also had weight, CD cell count and viral load measured between 6 months before and months after starting HAART. Of these 165 patients, 2% were female, 71% were Caucasian southern or northern Europeans, 15% were black sub-saharan Africans, 11% had a family history of cardiovascular disease and 15% were probably infected through iv drug use; when starting HAART, 59% were in CDC group A, the median age was 7 and the median BMI was 22. Of these 165 patients, 59 and 71 started PI- and NNRTI-based therapy, respectively, and were followed for a median of 18 [interquartile range (IQR) 6.5 ] and 17 (IQR ) months from starting HAART until either switching from one therapy to the other or until the last lipid measurement to date. These patients contributed a total of 1 blood samples, of which 2% were taken in a fasting state. For patients starting PI- and NNRTI-based therapy, 288 and 166 samples were available, respectively, and of these, 1% and 1% were taken in a fasting state. Antiretroviral therapy The most common PIs were NFV and LPV with RTV, used by 269 (5%) and 292 (9%) patients, respectively, of the 59 using any sort of PI. Ninety-two patients (15%) received other boosted PIs, notably IDV, atazanavir and saquinavir. As first therapy, 55 of these 59 patients (9%) started on either NFV (25, %), LPV with RTV (256, %), or IDV with RTV (5, 8%). Most NNRTI use was of EFV this was used by (9%) of the 71 patients using either of the two NNRTIs. As first therapy, 22 of these 71 patients (9%) started on efavirenz. Common NRTIs were lamivudine and AZT, used by 125 (96%) and 9 (85%) patients, respectively. dt, ABC and TDF were used by 157 (15%), 178 (17%), and 1 (12%) patients, respectively, with median exposures of 11 (IQR.7 2), 1 (IQR.5 26) and 6.2 (IQR ) months. Lipid profiles Using the same smoothing function for each lipid, unadjusted descriptive analyses suggest early changes in both HDL and non-hdl cholesterol (that is, total cholesterol minus HDL cholesterol). The levels of both increased within the first year, with a greater increase in HDL cholesterol for patients on NNRTI-based therapy (Figure 1). Triglyceride levels were higher for those starting PI-based therapy compared with those starting NNRTI-based therapy. Triglyceride levels decreased with time on either therapy a result not confirmed by subsequent adjusted analyses. Adjusted analyses suggest similar increases in non- HDL cholesterol with increasing exposure to either PIor NNRTI-based therapy [.15 (95% CI: 7,.2) and.11 (95% CI: 2,.2) mmol/l per year, respectively; see Table 1]. However, with increasing exposure to NNRTI-based therapy, HDL cholesterol levels increased and triglyceride levels decreased; whereas with increasing exposure to PI-based therapy, triglyceride levels increased. Between NNRTI-based therapies, the only difference was in triglyceride levels, which tended to increase with increasing exposure to EFV and to decrease with increasing exposure to NVP. Between the common PIbased therapies, differences were the rule rather than the exception. In comparison with other PIs and NNRTIs, increasing exposure to NFV was associated with the smallest increase in non-hdl cholesterol and the greatest decrease in triglycerides. HDL cholesterol increased with increasing exposure to LPV with RTV, but otherwise RTV-boosted regimens were associated with the least favourable lipid profiles because the highest increases in both non-hdl cholesterol and triglycerides were seen with increasing exposure to either of the two boosted PIs. Increasing exposure to ABC was consistently associated in all analyses with a decrease in triglyceride levels. Increasing exposure to ABC was also associated with a decrease in non-hdl cholesterol but only within NNRTI-based therapy [.29 (.7,.11) mmol/l per year] and not within PI-based therapy [7 (.16,.1) mmol/l per year]. We carried out three sensitivity analyses. Firstly, the sample included patients with viral load measured up to months after starting HAART, so that fewer patients were excluded from our analyses. However, viral load can change rapidly after starting HAART. If we repeated our analyses without baseline viral load as a covariate, the results in Table 1 were essentially the same. Secondly, only 5 patients (.%) in this sample used lipid-lowering drugs, so estimates of the association between lipid levels and exposure to such drugs are not precise. The results in Table 1 are essentially the same when we remove all observations made after starting a lipid-lowering drug. Thirdly, as an alternative adjustment for concomitant NRTI use, we adjusted separately for time on either ABC or TDF and for time on either dt or AZT. This increases point estimates in Table 1 for the effect of each class and for the effect of each individual PI and NNRTI. This confounding occurs because so many patients received AZT that time on AZT is correlated with time on other components of therapy. However, the ranking of the two different classes and of different drugs within each class is robust to this alternative adjustment. Even crossclass comparisons between specific PIs and NNRTIs seem relatively robust to this alternative adjustment, although one should be cautious about such comparisons as many involve at least one relatively small patient subgroup. Antiviral Therapy 1:5 587

4 J Young et al. Figure 1. Lipid concentration (mmol/l) over time on HAART (years) for those receiving PI-based therapy only ( ) or NNRTIbased therapy only ( -) A 1 8. B 2.5 Total cholesterol 6. HDL cholesterol C D Total cholesterol HDL cholesterol Total cholesterol:hdl cholesterol E 5. Triglycerides. 1. (A) total cholesterol, (B) HDL cholesterol, (C) total cholesterol minus HDL cholesterol, (D) total cholesterol to HDL cholesterol ratio and (E) triglycerides. Discussion Similar increases in non-hdl cholesterol are seen with increasing exposure to either PI- or NNRTI-based therapy. Increases in HDL cholesterol and decreases in triglycerides are seen with increasing exposure to NNRTI-based therapy; whereas increases in triglycerides are seen with increasing exposure to PI-based therapy. Comparing NNRTIs, triglyceride levels are slightly higher with increasing exposure to EFV than with increasing exposure to NVP. Increasing exposure to boosted PI therapies leads to higher levels of both International Medical Press

5 Lipid profiles in the SHCS Table 1. Associations between time on antiretroviral therapy and change in blood lipid concentrations Mean change in lipid concentration per year of use [95% confidence interval] Total cholesterol, HDL cholesterol, Total cholesterol Total cholesterol: Triglycerides, Antiretroviral therapy* mmol/l mmol/l HDL cholesterol HDL cholesterol mmol/l Between classes n=16 P=.5 n=158 P<1 n=155 P=.5 n=155 P=1 n=16 P<1 PI.18 [9,.28] 1 [ 2, ].15 [7,.2].11 [ 7,.].21 [9,.] NNRTI.21 [.11,.].1 [7,.1].11 [2,.2].2 [.8,.11].1 [.2, ] dt.1 [.26, 6] [ 7, 7] 9 [.25, 7]. [.8, 1.5] [.17,.25] ABC.11 [.2, ] 2 [ 2, 6].1 [.22, ].2 [., 5].2 [.7, ] Between NNRTIs n=57 P=.61 n=5 P=.5 n=52 P=.87 n=52 P=.9 n=57 P= EFV.2 [.1,.] 7 [,.11].16 [7,.25].1 [.25, ] 8 [,.19] NVP. [5,.55].11 [2,.19].18 [ 6,.2].2 [.6, 7] 7 [.19, 6] dt 1 [.2,.19] [ 8,.16] 6 [.29,.18] 5 [.9,.9].1 [.16,.] ABC. [.56,.1] [.11, ].29 [.7,.11].25 [.6, ].1 [.25, 1] Between PIs n=5 P<1 n=528 P=1 n=526 P=1 n=526 P=7 n=52 P<1 NFV 7 [,.19] [, 6] 5 [ 6,.16] [.27,.2].12 [.25, 1] LPV/r.1 [.16,.7] 8 [,.1].22 [7,.6].12 [.,.11]. [.17,.6] IDV/r.66 [.2,.99] [ 7, 2].6 [.6,.8].1 [9,.7].69 [.16, 1.22] dt 5 [.26,.16] 2 [.1,.1] [.2,.17] 1. [.68, 2.7].2 [,.7] ABC 8 [.15,.1] 5 [,.1] 7 [.16,.1].18 [.5,.17].22 [.5, ] *For between-class analyses, therapy is represented by four predictors: time on any PI, any NNRTI, stavudine and abacavir. For within-class analyses (between NNRTIs or between PIs), therapy is represented by time on each of the (four or five) predictors listed. All analyses are adjusted for gender, family history of cardiovascular disease, transmission probably by iv drug use; baseline age, BMI, log viral load and CD cell count; measurement known to be made in fasting state; and use of lipid lowering drugs. Excludes measurements made after a patient has taken both PI- and NNRTI-based therapies (P value is the result of a score test for a difference between PI- and NNRTI-based therapies). Excludes measurements made after a patient has taken either a PI-based therapy or both EFV and NVP (P value is the result of a score test for a difference between the two NNRTIs). Excludes measurements made after a patient has taken either an NNRTI-based therapy, any PI therapy other than NFV, LPV/r or IDV/r, or more than one of these PI therapies (P value is the result of a score test for a difference between the three PI therapies). BMI, body mass index; iv, intravenous; NNRTI, non-nucleoside reverse transcriptase inhibitor; PI, protease inhibitor; ABC, abacavir; dt, stavudine; EFV, efavirenz; IDV/r, indinavir plus ritonavir; LPV/r, lopinavir plus ritonavir; NFV, nelfinavir; NVP, nevirapine. non-hdl cholesterol and triglycerides than increasing exposure to NFV. These results are consistent with a recent crosssectional study of lipid profiles in 11 cohorts including the SHCS [2]. The strengths of our study are the longitudinal design and the duration of follow-up, combined with our exclusion of pre-treated patients and of data once patients switch from one type of therapy to another. The main limitation of our study is the potential for confounding inherent in every observational study. While the two NNRTIs have relatively similar lipid profiles, other studies with more power show that NVP has the more favourable profile [2,21] (EFV has the better safety profile [22]). However, different PI therapies are associated with very different lipid profiles. NVP appears to have a relatively favourable lipid profile. In a randomized trial comparing NFV with NVP, the only difference in lipid profiles was a higher increase in HDL cholesterol for patients on NVP [2]. With boosted PIs, there is debate over whether lipid abnormalities are due to RTV or to the other PI or to both [2]. While LPV plasma concentrations have been found to correlate with total cholesterol and triglyceride levels, RTV has been shown to cause lipid abnormalities in healthy volunteers [25]. Few studies compare boosted LPV with boosted IDV [26,27], but what evidence there is suggests that, as in this study, boosted IDV has the less favourable lipid profile. The increase seen in this study in HDL cholesterol with increasing exposure to boosted LPV is consistent with an increase in HDL cholesterol when boosted LPV is used as salvage therapy [28]. ABC is associated with lower triglyceride levels, in both between- and within-class analyses. ABC is also associated with lower non-hdl cholesterol levels but only within NNRTI-based therapy. There could be between-class differences in the effect of ABC on lipid profiles. Improvements in lipid abnormalities are not always seen when dt is replaced by ABC in HAART. In one study where improvements were seen, roughly 8% of patients were on NNRTI-based therapy [17]; whereas in a study where no improvements were seen [18], only around 5% of patients were on NNRTIbased therapy [29]. Antiviral Therapy 1:5 589

6 J Young et al. In conclusion, while in general NNRTI-based therapy is associated with a more favourable lipid profile than PI-based therapy, different PI-based therapies are associated with very different lipid profiles. NFV appears to have a relatively favourable lipid profile. Of the two boosted PI therapies, LPV appears to have a more favourable lipid profile than IDV. Increasing exposure to ABC is associated with a decrease in the level of triglycerides. Acknowledgements This study has been partly funded within the framework of the Swiss HIV Cohort Study, supported by the Swiss National Science Foundation (Grant no ). The Basel Institute for Clinical Epidemiology is supported by grants from santésuisse and from the Gottfried and Julia Bangerter-Rhyner Foundation. An unrestricted grant specifically for this study was received from Bristol Myers Squibb Switzerland. The members of the Swiss HIV Cohort Study are M Battegay, E Bernasconi, J Böni, H Bucher, Ph Bürgisser, S Cattacin, M Cavassini, R Dubs, M Egger, L Elzi, P Erb, K Fantelli, M Fischer, M Flepp, A Fontana, P Francioli (President of the SHCS, Centre Hospitalier Universitaire Vaudois, Lausanne), H Furrer (Chairman of the Clinical and Laboratory Committee), M Gorgievski, H Günthard, B Hirschel, L Kaiser, C Kind, Th Klimkait, U Lauper, B Ledergerber, M Opravil, F Paccaud, G Pantaleo, L Perrin, J-C Piffaretti, M Rickenbach (Head of Data Centre), C Rudin (Chairman of the Mother & Child Substudy), P Schmid, J Schüpbach, R Speck, A Telenti, A Trkola, P Vernazza (Chairman of the Scientific Board), R Weber and S Yerly. References 1. Sellmeyer DE & Grunfeld C. Endocrine and metabolic disturbances in human immunodeficiency virus infection and the acquired immune deficiency syndrome. Endocrine Reviews 1996; 17: Riddler SA, Smit E, Cole SR, Li R, Chmiel JS, Dobs A, Palella F, Visscher B, Evans R & Kingsley LA. Impact of HIV infection and HAART on serum lipids in men. Journal of the American Medical Association 2; 289: Thiebaut R, Dabis F, Malvy D, Jacqmin-Gadda H, Mercie P & ValentinVD. Serum triglycerides, HIV infection, and highly active antiretroviral therapy, Aquitaine Cohort, France, 1996 to Groupe d Epidemiologie Clinique du Sida en Aquitaine (GECSA). Journal of Acquired Immune Deficiency Syndromes 2; 2: Friis-Moller N, Sabin CA, Weber R, d Arminio MA, El Sadr WM, Reiss P, Thiebaut R, Morfeldt L, De Wit S, Pradier C, Calvo G, Law MG, Kirk O, Phillips AN & Lundgren JD. Combination antiretroviral therapy and the risk of myocardial infarction. New England Journal of Medicine 2; 9: Penzak SR & Chuck SK. Hyperlipidemia associated with HIV protease inhibitor use: pathophysiology, prevalence, risk factors and treatment. Scandinavian Journal of Infectious Diseases 2; 2: Rhew DC, Bernal M, Aguilar D, Iloeje U & Goetz MB. Association between protease inhibitor use and increased cardiovascular risk in patients infected with human immunodeficiency virus: a systematic review. Clinical Infectious Diseases 2; 7: Murphy RL & Smith WJ. Switch studies: a review. HIV Medicine 22; : Drechsler H & Powderly WG. Switching effective antiretroviral therapy: a review. Clinical Infectious Diseases 22; 5: van der Valk M, Kastelein JJ, Murphy RL, van Leth F, Katlama C, Horban A, Glesby M, Behrens G, Clotet B, Stellato RK, Molhuizen HO & Reiss P. Nevirapinecontaining antiretroviral therapy in HIV-1 infected patients results in an anti-atherogenic lipid profile. AIDS 21; 15: Tashima KT, Bausserman L, Alt EN, Aznar E & Flanigan TP. Lipid changes in patients initiating efavirenz- and indinavir-based antiretroviral regimens. HIV Clinical Trials 2; : Cleveland WS, Devlin SJ & Grosse E. Regression by local fitting : methods, properties, and computational algorithms. Journal of Econometrics 1988; 7: Carriere I & Bouyer J. Choosing marginal or randomeffects models for longitudinal binary responses: application to self-reported disability among older persons. BMC Medical Research Methodology 22; 2: Pan W & Connett JE. Selecting the working correlation structure in generalized estimating equations with application to the lung health study. Statistica Sinica 22; 12: Rotnitzky A & Jewell NP. Hypothesis testing of regression parameters in semiparametric generalized linear models for cluster correlated data. Biometrika 199; 77: Domingo P, Labarga P, Palacios R, Guerrero MF, Terron JA, Elias MJP, Santos J, Camps MIR, Llibre JM & Moreno S. Improvement of dyslipidemia in patients switching from stavudine to tenofovir: preliminary results. AIDS 2; 18: Gallant JE, Staszewski S, Pozniak AL, DeJesus E, Suleiman JMAH, Miller MD, Coakley DF, Lu B, Toole JJ & Cheng AK. Efficacy and safety of tenofovir DF vs stavuldine in combination therapy in antiretroviral-naive patients a -year randomized trial. Journal of the American Medical Association 2; 292: Garcia-Benayas T, Blanco F, Alcolea A, De la Cruz JJ, Gonzalez-Lahoz J & Soriano V. Benefits in the lipid profile after substitution of abacavir for stavudine: a 8-week prospective study. AIDS Research & Human Retroviruses 2; 2: Martin A, Smith DE, Carr A, Ringland C, Amin J, Emery S, Hoy J, Workman C, Doong N, Freund J & Cooper DA. Reversibility of lipoatrophy in HIV-infected patients 2 years after switching from a thymidine analogue to abacavir: the MITOX Extension Study. AIDS 2; 18: Caron M, Auclair M, Lagathu C, Lombes A, Walker UA, Kornprobst M & Capeau J. The HIV-1 nucleoside reverse transcriptase inhibitors stavudine and zidovudine alter adipocyte functions in vitro. AIDS 2; 18: Fontas E, van Leth F, Sabin CA, Friis-Moller N, Rickenbach M, d Arminio MA, Kirk O, Dupon M, Morfeldt L, Mateu S, Petoumenos K, El Sadr W, De Wit S, Lundgren JD, Pradier C & Reiss P. Lipid profiles in HIVinfected patients receiving combination antiretroviral therapy: are different antiretroviral drugs associated with different lipid profiles? Journal of Infectious Diseases 2; 189: van Leth F, Phanuphak P, Stroes E, Gazzard B, Cahn P, Raffi F, Wood R, Bloch M, Katlama C, Kastelein JJ, Schechter M, Murphy RL, Horban A, Hall DB, Lange JM & Reiss P. Nevirapine and efavirenz elicit different changes in lipid profiles in antiretroviral-therapy-naive patients infected with HIV-1. PLoS Medicine 2; 1:e International Medical Press

7 Lipid profiles in the SHCS 22. van Leth F, Phanuphak P, Ruxrungtham K, Baraldi E, Miller S, Gazzard B, Cahn P, Lalloo UG, van der Westhuizen IP, Malan DR, Johnson MA, Santos BR, Mulcahy F, Wood R, Levi GC, Reboredo G, Squires K, Cassetti I, Petit D, Raffi F, Katlama C, Murphy RL, Horban A, Dam JP, Hassink E, van Leeuwen R, Robinson P, Wit FW & Lange JM. Comparison of first-line antiretroviral therapy with regimens including nevirapine, efavirenz, or both drugs, plus stavudine and lamivudine: a randomised open-label trial, the 2NN Study. Lancet 2; 6: Fisac C, Virgili N, Ferrer E, Barbera MJ, Fumero E, Vilarasau C & Podzamczer D. A comparison of the effects of nevirapine and nelfinavir on metabolism and body habitus in antiretroviral-naive human immunodeficiency virus-infected patients: a randomized controlled study. Journal of Clinical Endocrinology & Metabolism 2; 88: Gutierrez F, Padilla S, Navarro A, Masia M & Hernandez I. Lipid abnormalities in HIV-infected patients and lopinavir plasma concentrations. Journal of Acquired Immune Deficiency Syndromes 2; 6: Purnell JQ, Zambon A, Knopp RH, Pizzuti DJ, Achari R, Leonard JM, Locke C & Brunzell JD. Effect of ritonavir on lipids and post-heparin lipase activities in normal subjects. AIDS 2; 1: Bongiovanni M, Bini T, Chiesa E, Cicconi P, Adorni F & Monforte AA. Lopinavir/ritonavir vs. indinavir/ritonavir in antiretroviral naive HIV-infected patients: immunovirological outcome and side effects. Antiviral Research 2; 62: Antoniou T, Park-Wyllie L & Boyle E. A comparison of dyslipidemias associated with either lopinavir/ritonavir- or indinavir/ritonavir-based antiretroviral therapy. Journal of Acquired Immune Deficiency Syndromes 2; 7: Gutierrez F, Padilla S, Navarro A, Masia M, Hernandez I, Ramos J, Esteban A & Martin-Hidalgo A. Lopinavir plasma concentrations and changes in lipid levels during salvage therapy with lopinavir/ritonavir-containing regimens. Journal of Acquired Immune Deficiency Syndromes 2; : Carr A, Workman C, Smith DE, Hoy J, Hudson J, Doong N, Martin A, Amin J, Freund J, Law M & Cooper DA. Abacavir substitution for nucleoside analogs in patients with HIV lipoatrophy: a randomized trial. Journal of the American Medical Association 22; 288: Received 8 March 25, accepted 2 May 25 Antiviral Therapy 1:5 591