Original article The effect of tenofovir disoproxil fumarate on whole-body insulin sensitivity, lipids and adipokines in healthy volunteers

Transcription

1 Antiviral Therapy 21 15: (doi: /IMP1518) Original article The effect of tenofovir disoproxil fumarate on whole-body insulin sensitivity, lipids and adipokines in healthy volunteers Paul A Randell 1, Akil G Jackson 1, Lijie Zhong 2, Kitty Yale 2, Graeme J Moyle 1 1 Chelsea and Westminster Hospital, London, UK 2 Gilead Sciences, Foster City, CA, USA Corresponding author paul.randell@chelwest.nhs.uk Background: Certain antiretrovirals are known to affect lipid and glucose homeostasis. The aim of this study was to assess the effect on insulin sensitivity (determined by peripheral glucose uptake using a hyperinsulinaemic euglycaemic clamp) of tenofovir disoproxil fumarate () administration compared with placebo for 2 weeks in HIV-1-seronegative healthy male volunteers. Changes in lipids, adiponectin, leptin, plasminogen activator inhibitor 1 (PAI-1) and the adhesion molecules E-selectin and P-selectin were also assessed. Methods: This was a single-centre, randomized, doubleblinded, placebo-controlled study that used a twosequence, two-period cross-over design. A total of 19 HIV-negative males were recruited to the study and randomized 1:1 to receive either 2 weeks of (3 mg once daily) followed by 2 weeks of placebo or placebo initially followed by tenofovir. Clamps were performed at baseline, after 2 weeks and after 4 weeks. Results: All three clamps were completed by 16 participants. During the euglycaemic clamp, there were no significant changes in insulin sensitivity after 2 weeks of administration compared with placebo or baseline. There was a significant reduction in the mean total cholesterol (9.4%) and low-density lipoprotein (LDL; 8.1%) cholesterol following 2 weeks of compared with placebo. Levels of adiponectin, leptin, PAI-1, P-selectin and E- selectin were not significantly altered. Conclusions: use for 2 weeks does not affect insulin sensitivity, as assessed by the hyperinsulinaemic euglycaemic clamp in HIV-negative male volunteers. use resulted in modest, but statistically significant, reductions in total and LDL cholesterol. Introduction Effective antiretroviral therapy has improved mortality [1] and reduced morbidity associated with HIV infection. As HIV-infected patients are living longer, chronic prevalent diseases of aging, particularly ischaemic heart disease and type 2 diabetes mellitus, are becoming important causes of morbidity and mortality. Specific antiretrovirals have been associated with risk of myocardial infarction [2], diabetes mellitus [3] and the development of metabolic and morphological disturbances including dyslipidaemia, insulin resistance and body-fat changes [4]. These changes differ between drug classes and between drugs in the same class. As the choice of antiretroviral regimen can influence cardiovascular and diabetes mellitus disease risk, it is necessary to understand the metabolic effect of individual drugs to help inform this choice. The hyperinsulinaemic euglycaemic clamp [5] is the gold standard for investigating insulin sensitivity in a research setting and has been used to investigate the effect of different antiretrovirals on glucose homeostasis. A number of protease inhibitors (PIs) [6 11] and nucleoside reverse transcriptase inhibitors (NRTIs) [12,13] have been associated with the development of insulin resistance. Mitochondrial toxicity is implicated in the development of insulin resistance associated with thymidine NRTIs [12,13], whereas with PIs insulin resistance appears to be mediated through the inhibition of the glucose transporter GLUT4 [14]. This study investigated in healthy HIV-negative volunteers the effect of tenofovir disoproxil fumarate () on whole-body insulin sensitivity, as assessed by the hyperinsulinaemic euglycaemic clamp. In addition, 21 International Medical Press (print) (online) 227

2 PA Randell et al. the effects of on lipid metabolism, adipokines (adiponectin, leptin and plasminogen activator inhibitor 1 [PAI-1]) and other markers of endothelial function (E- and P-selectin) were assessed. Methods Study design This was a single-centre, double-blinded, placebocontrolled study using a two-sequence, two-period cross-over design. Eligible participants were randomized 1:1 to receive either 2 weeks of (3 mg orally, once daily) followed by matching placebo for 2 weeks or commenced with placebo followed by active. A hyperinsulinaemic euglycaemic clamp was performed at baseline and then at weeks 2 and 4 (2 weeks after treatment switch). Fasting samples for lipids, adiponectin, leptin, PAI-1, E-selectin and P-selectin were also taken at these visits. Eligibility criteria The study was approved by the Hammersmith research ethic committee and conducted according to the principles of good clinical practice. Participants gave informed consent and were screened up to 28 days prior to their baseline visit. Male participants were included if they were aged between 18 and 55 years, were clinically well, and not receiving any metabolically active or lipidlowering medication, hormonal agents, glucocorticoids, β-blockers, thiazide diuretics, thyroxine, psychotropic agents, anabolic steroids, megestrol acetate, nephrotoxic agents, chemotherapeutic agents or immunomodulating agents. Fasting glucose and lipid parameters were required to be less than the upper limit of normal: for glucose <6. mmol/l (19 mg/dl), triglycerides <2. mmol/l (182 mg/dl) and cholesterol <5. mmol/l (192 mg/dl). Creatinine clearance was assessed using the Cockcroft Gault formula and participants were only included if the calculated creatinine clearance was >1 ml/min. Other relevant exclusion criteria included a waist:hip ratio >.97, a body mass index (BMI) >28 kg/m 2, acute or chronic hepatitis B or C infection, diabetes mellitus or any disease process likely to cause a marked disturbance in glucose and lipid homeostasis. Hyperinsulinaemic euglycaemic clamp Participants were admitted to the research unit after an overnight fast of at least 8 h. An intravenous cannula was inserted into the right antecubital fossa for infusion of insulin and a 2% glucose solution. A retrograde cannula was inserted in the back of the distal third of the left forearm, which was heated to 55 C in a thermoregulated box. This was used for blood sampling for the duration of the 18 min clamp procedure. At t= a continuous infusion of insulin (Actrapid, 1 IU/ ml, Novo Nordisk A/S, Bagsværd, Denmark) was commenced at 8 mu/m 2 /min. At t=5 min, the rate was reduced to 6 mu/m 2 /min until t=1 min when it was reduced to 4 mu/m 2 /min and continued at this rate for the remainder of the clamp. At t=4 min the 2% glucose solution was commenced and from t=1 min the plasma glucose was analysed (using the i-stat Analyser; istat Corporation, East Windsor, NJ, USA) every 5 min. A negative feedback formula [5] was used to adjust the rate of glucose infusion in order to maintain euglycaemia at 4.5 mmol/l. Serial electrocardiograms and whole-blood potassium were monitored during the clamp procedure. Blood samples where taken at t=12 min, 15 min and 18 min to determine average insulin concentration during the last hour of the clamp [5]. Insulin sensitivity was assessed by calculating the ratio M/I (glucose disposal rate [M], during the last hour of the clamp, divided by average insulin concentration [I] over this time). In addition, the homeostatic model assessment index-insulin resistance (HOMA-IR) was calculated using the formula: (fasting insulin fasting glucose)/22.5 was calculated at each visit. Statistical analyses The primary objective of the study was to assess the effect of the administration of compared with placebo for 2 weeks in HIV-1-negative healthy male volunteers using the hyperinsulinaemic euglycaemic clamp. Results were analysed to assess for any period effect. Subsequently, the mean insulin-stimulated glucose disposal (M/I: mg/kg/min per mu/l insulin 1) in the group was compared with the placebo group using the Wilcoxon signed-rank test. Results A total of 19 participants were randomized. Three participants withdrew prior to the baseline visit; the remaining 16 completed all study assessments. Nine participants received placebo followed by and seven received followed by placebo. There were no significant differences in baseline characteristics between the two groups. The participants ranged in age from 19 to 37 years with a mean of 27.4 years. Mean (±sd) BMI was 23.2 kg/m 2 (±1.665); the mean (±sd) for the waist:hip ratio was.88 (±.37). Overall, the mean (±sd) insulin concentrations achieved during the last hour of the clamps were (±15.51) mu/l, (±15.) mu/l and (±15.99) mu/l for the 12, 15 and 18 min time points, respectively. Boxplots of the ratio of insulin concentrations over these time points to the insulin concentration at 12 min are presented in Figure 1, to better illustrate the variability observed. There was no significant effect on whole-body insulin sensitivity, as assessed by the International Medical Press

3 The effect of tenofovir on whole-body insulin sensitivity glucose disposal rate M or M/I during the final hour of the clamp during the and placebo treatment periods. No significant period effect was observed between groups. In the placebo group, two patients experienced marked increases in the M/I ratio compared with baseline (Figure 2). No procedural errors were noted during these time points and these results have been included in the analysis; however, the outcomes are the same whether these individuals are included or excluded. HOMA-IR was also not significantly altered. There was no significant difference between the two groups (see Table 1). Lipid parameters are presented in Figure 3. Mean total cholesterol was reduced by 9.4% compared with baseline (P=.4) after dosing with 2 weeks of. A statistically significant reduction was also seen compared with placebo (P=.17). This reduction was predominantly due to an 8.1% reduction in low-density lipoprotein (LDL) cholesterol (P=.25 for the change from baseline in the treatment period). There were no significant changes in lipids after 2 weeks of placebo. Neither nor placebo had any significant effect on adiponectin, leptin, PAI-1, P-selectin or E- selectin (Figure 4). No serious adverse events were experienced by the participants during the course of the study. There was no development of grade 3 or 4 laboratory results in either group. Adverse events were mild with upper respiratory tract infection the commonest adverse event (three in each group). Other adverse events reported (by individual participants) were headache, nausea and thrombophlebitis in the placebo group and non-cardiac chest pain, fatigue and back injury in the group. Discussion Studies using the hyperinsulinaemic clamp have demonstrated variable effects of antiretrovirals on glucose homeostasis. In healthy volunteers, single doses of lopinavir/ritonavir [6,7,11] have resulted in a reduction in insulin-mediated glucose disposal. A hierarchy of the acute effect PIs have on insulin sensitivity has been proposed: indinavir causes the greatest reduction followed by ritonavir and lopinavir/ritonavir, whereas amprenavir Figure 1. Boxplot of the ratio of insulin concentration at different time points to the insulin concentration at 12 min Treatment arm Placebo Ratio 2. Ratio 2. Ratio Time point, min Time point, min Time point, min Outliers represented by circles (between 1.5 and 3 interquartile range above upper quartile) and asterisks (>3 interquartile range). Antiviral Therapy

4 PA Randell et al. has no effect [7]. Neither atazanavir given for 5 days [1] nor atazanavir coadministered with ritonavir for 1 days [8] significantly affected insulin-mediated glucose disposal. Although an approximate 25% reduction in the glucose disposal rate was seen after 5 days [1] and 1 days [8] of lopinavir/ritonavir, no change was seen after 4 weeks of dosing [15]. Four weeks of indinavir treatment led to a 2% reduction in insulin-mediated glucose disposal rate, independent from changes in lipids and visceral fat [9], compared with a 34% reduction acutely [11]. Both 4 weeks of indinavir and lopinavir/ ritonavir were associated with an increase in adiponectin [16], which might be responsible for ameliorating the initial effect of these PIs on insulin sensitivity. NRTIs have also been implicated in the development of insulin resistance. Stavudine given for 1 month in healthy volunteers resulted in a significant reduction in glucose infusion rates during the clamp. This was associated with a 52% reduction in mitochondrial DNA [13]. In the MEDICLAS substudy, HIV-positive patients in the NRTI arm (lopinavir/ritonavir plus lamivudine plus zidovudine) showed a 25% reduction in insulin- stimulated glucose disposal at 3 months, which was in contrast to the NRTI-sparing arm (lopinavir/ ritonavir plus nevirapine) where glucose metabolism was not affected [12]. These findings are consistent with those from the D:A:D study group, which found that cumulative exposure to stavudine and zidovudine were significantly associated with diabetes mellitus in a large HIV-positive cohort [3]. In our study, 2 weeks of administration did not significantly affect insulin sensitivity during the hyperinsulinaemic euglycaemic clamp, as measured by the glucose disposal rate (M) and the ratio M/I. In addition, there was no significant affect on the HOMA-IR. These findings are consistent with a lack of any Figure 2. Intraindividual changes in the M/I ratio after 2 weeks treatment with or placebo A B M/I, mg/kg/min per mu/l insulin P=.117 Week 2 Week 2 Clamp visit M/I, mg/kg/min per mu/l insulin P=.464 Week 2 Week 2 Clamp visit Error bars are interquartile range. P-values were obtained by using Wilcoxon signed-rank test. (A) Changes in M/I (glucose disposal rate [M] divided by the average insulin concentration [I]) over last 6 min of clamp (12 18 min) in participants treated with tenofovir disoproxil fumarate () for 2 weeks. (B) Changes in M/I in participants treated with placebo for 2 weeks. Table 1. Insulin sensitivity at baseline and after 2 weeks treatment with or placebo (day 1 clamp) After 2 weeks placebo After 2 weeks administration Glucose disposal rate (M), mg/kg/min 9.47 ( ) 9.48 ( ) 8.75 ( ) M/I a, mg/kg/min/mu/l ( ) 16.8 ( ) 16.3 ( ) HOMA-IR.89 ( ).92 (.5 1.8).83 (.7.96) Data are (interquartile range) and were derived from 16 participants. The Wilcoxon signed-rank test was used for comparison to baseline. No statistically significant changes were found. a The glucose disposal rate (M) divided by the average insulin concentration (I) over min. HOMA-IR, homeostatic model assessment index-insulin resistance;, tenofovir disoproxil fumarate International Medical Press

5 The effect of tenofovir on whole-body insulin sensitivity significant mitochondrial toxicity associated with [17]. Furthermore, the absence of an increase in adiponectin indicates a lack of compensatory changes in glucose handling. Changes seen in healthy volunteer studies are not necessarily observed in HIV-infected individuals [12,18,19], but the results of this study are consistent with the lack of a significant effect on insulin sensitivity seen when has been used as part of the antiretroviral combination in patients initiating treatment [18,19]. Dyslipidaemia and insulin resistance are important contributors to cardiovascular risk and can be modified by antiretroviral treatment [4]. Studies have demonstrated a more favourable lipid profile of relative to stavudine [2] and zidovudine [21], and when switching from stavudine [22] or zidovudine to [23,24]. In our study, there was a statistically significant improvement in total cholesterol after 2 weeks of dosing with compared with placebo and baseline. Importantly, this reduction was driven mainly by a reduction in LDL cholesterol. The mechanism of this effect is not known, but has also been observed in studies where has been used as an intensification agent [25]. In our study we evaluated selected biomarkers in an attempt to further assess the metabolic effect of. The adipokine adiponectin has an important role in Figure 3. Measurements of lipid parameters at baseline and after 2 weeks treatment with or placebo Mean, mmol/l P=.17 P=.4 Total cholesterol HDL P=.25 LDL Lipid parameter Placebo Triglycerides Error bars are 95% confidence intervals. P-values obtained using Wilcoxon signedrank test. Non-statistically significant changes not shown. HDL, high-density lipoprotein; LDL, low-density lipoprotein;, tenofovir disoproxil fumarate. Figure 4. Measurements of adipokines and vascular markers at baseline and after 2 weeks treatment with or placebo 1, Adiponectin 1, Leptin 8, 8, 6, 4, pg/ml 6, 4, 2, 2, Placebo Placebo Plasminogen activator inhibitor E-selectin Placebo Placebo Placebo P-selectin In all panels, the value is plotted together with the interquartile range., tenofovir disoproxil fumarate. Antiviral Therapy

6 PA Randell et al. the regulation of energy metabolism. Decreased levels have been found to be associated with insulin resistance [26] and it appears to have a protective role against the development of atherosclerosis [27]. Leptin is also involved in energy metabolism and serves as a marker of energy stores and regulates neuroendocrine, immunological and metabolic function, in energy-deficient states [28 3]. The dysregulation of adiponectin and leptin have been implicated in the pathogenesis of the metabolic syndrome associated with antiretroviral therapy (ART) [31]. PAI-1 has a role in thrombotic vascular disease [32] and increased levels have been seen in patients receiving PI-containing ART [33]. Increased endothelial adhesion molecules [34,35] (including E- selectin) have been associated with HIV and can be used as a marker of endothelial function. In this study, did not appear to affect adipokine output or markers of endothelial function. A limitation of this study is the short period of followup; further investigation to assess the effect of on these biomarkers beyond 2 weeks would be useful. In conclusion, dosing with for 2 weeks did not have a significant effect on whole-body insulin sensitivity, as assessed using the hyperinsulinaemic euglycaemic clamp, compared with placebo and with baseline in HIV-negative volunteers. In addition, there was no significant effect on the adipokines (adiponectin, leptin and PAI-1) or on the adhesion molecules E-selectin and P-selectin. There was a statistically significant decrease in total cholesterol (compared with placebo and baseline) and LDL cholesterol (compared with baseline) after 2 weeks of. Acknowledgements The project was funded by Gilead Sciences, Foster City, CA, USA. Disclosure statement LZ and KY are employed by Gilead Sciences, Foster City, CA, USA. The remaining authors declare no competing interests. References 1. Antiretroviral Therapy Cohort Collaboration. Life expectancy of individuals on combination antiretroviral therapy in high-income countries: a collaborative analysis of 14 cohort studies. Lancet 28; 372: Lundgren J, Reiss P, Worm S, et al. Risk of myocardial infarction with exposure to specific ARV from the PI, NNRTI, and NRTI drug classes: the D:A:D study. Conference on Retroviruses and Opportunistic Infections February 29, Montreal, QC, Canada. Abstract 44LB. 3. De Wit S, Sabin CA, Weber R, et al. Incidence and risk factors for new-onset diabetes in HIV-infected patients the data collection on adverse events of anti-hiv drugs (D:A:D) study. Diabetes Care 28; 31: Grinspoon S, Carr A. Cardiovascular risk and body-fat abnormalities in HIV-infected adults. N Engl J Med 25; 352: DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol 1979; 237:E214 E Lee GA, Lo JC, Aweeka F, et al. Single-dose lopinavirritonavir acutely inhibits insulin-mediated glucose disposal in healthy volunteers. Clin Infect Dis 26; 43: Lee GA, Rao M, Mulligan K, et al. Effects of ritonavir and amprenavir on insulin sensitivity in healthy volunteers. AIDS 27; 21: Noor MA, Flint OP, Maa J-F, Parker RA. Effects of atazanavir/ritonavir and lopinavir/ritonavir on glucose uptake and insulin sensitivity: demonstrable differences in vitro and clinically. AIDS 26; 2: Noor MA, Lo JC, Mulligan K, et al. Metabolic effects of indinavir in healthy HIV-seronegative men. AIDS 21; 15:F11 F Noor MA, Parker RA, O Mara E, et al. The effects of HIV protease inhibitors atazanavir and lopinavir/ritonavir on insulin sensitivity in HIV-seronegative healthy adults. AIDS 24; 18: Noor MA, Seneviratne T, Aweeka FT, et al. Indinavir acutely inhibits insulin-stimulated glucose disposal in humans: a randomized, placebo-controlled study. AIDS 22; 16:F1 F Blumer RME, vanvonderen MGA, Sutinen J, et al. Zidovudine/lamivudine contributes to insulin resistance within 3 months of starting antiretroviral therapy. AIDS 28; 22: Fleischman A, Johnsen S, Systrom DM, et al. Effects of a nucleoside reverse transcriptase inhibitor, stavudine, on glucose disposal and mitochondrial function in muscle of healthy adults. Am J Physiol Endocrinol Metab 27; 292:E1666 E Murata H, Hruz PW, Mueckler M. The mechanism of insulin resistance caused by HIV protease inhibitor therapy. J Biol Chem 2; 275: Lee GA, Seneviratne T, Noor MA, et al. The metabolic effects of lopinavir/ritonavir in HIV negative men. AIDS 24; 18: Lee GA, Mafong DD, Noor MA, et al. HIV protease inhibitors increase adiponectin levels in HIV-negative men. J Acquir Immune Defic Syndr 24; 36: Birkus G, Hitchcock MJM, Cihlar T. Assessment of mitochondrial toxicity in human cells treated with tenofovir: comparison with other nucleoside reverse transcriptase inhibitors. Antimicrob Agents Chemother 22; 46: Jackson A, Patel N, Lo G, Gazzard B, Moyle G. Effects of atazanavir or saquinavir QD with ritonavir 1 mg and tenofovir/emtricitabine as initial therapy for HIV-1 infection on peripheral glucose disposal; a randomised open-label study. Conference on Retroviruses and Opportunistic Infections February 27, Los Angeles, CA, USA. Poster Randell P, Jackson A, Taylor J, Mandalia S, Moyle G. The effects of fosamprenavir or lopinavir with ritonavir, lamivudine and tenofovir, as initial treatment for HIV-1 infection, on insulin sensitivity. 11th European AIDS Conference October 27, Madrid, Spain. Poster LBP92/1. 2. Gallant JE, Staszewski S, Pozniak AL, et al. Efficacy and safety of tenofovir DF vs stavudine in combination therapy in antiretroviral-naive patients. JAMA 24; 292: Arribas JR, Pozniak AL, Gallant JE, et al. Tenofovir disoproxil fumarate, emtricitabine, and efavirenz compared with zidovudine/lamivudine and efavirenz in treatment-naive patients. J Acquir Immune Defic Syndr 28; 47: Moyle GJ, Sabin C, Cartledge J, et al. A randomized comparative trial of tenofovir DF or abacavir as replacement for a thymidine analogue in persons with lipoatrophy. AIDS 26; 2: International Medical Press

7 The effect of tenofovir on whole-body insulin sensitivity 23. Ananworanich J, Nuesch R, Côté HC, et al. Changes in metabolic toxicity after switching from stavudine/ didanosine to tenofovir/lamivudine a Staccato trial substudy. J Antimicrob Chemother 28; 61: Fisher M, Moyle GJ, Shahmanesh M, et al. A randomized comparative trial of continued zidovudine/lamivudine or replacement with tenofovir DF/emtricitabine in efavirenztreated HIV-1 infected individuals. J Acquir Immune Defic Syndr 29; 51: Tungsiripat M, Kitch D, Glesby M, et al. A pilot study to determine the effect on dyslipidemia of the addition of tenofovir to stable background ART in HIV-infected subjects: results from the A526 Study Team. Conference on Retroviruses and Opportunistic Infections February 29, Montreal, QC, Canada. Abstract Yamauchi T, Kamon J, Waki H, et al. The fat-derived hormone adiponectin reverses insulin resistance associated with both lipoatrophy and obesity. Nat Med 21; 7: Kubota N, Terauchi Y, Yamauchi T, et al. Disruption of adiponectin causes insulin resistance and neointimal formation. J Biol Chem 22; 277: Welt CK, Chan JL, Bullen J, et al. Recombinant human leptin in women with hypothalamic amenorrhea. N Engl J Med 24; 351: Chan JL, Matarese G, Shetty GK, et al. Differential regulation of metabolic, neuroendocrine, and immune function by leptin in humans. Proc Natl Acad Sci U S A 26; 13: Ahima RS, Prabakaran D, Mantzoros C, et al. Role of leptin in the neuroendocrine response to fasting. Nature 1996; 382: Jan V, Cervera P, Maachi M, et al. Altered fat differentiation and adipocytokine expression are inter-related and linked to morphological changes and insulin resistance in HIV- 1-infected lipodystrophic patients. Antivir Ther 24; 9: Kohler HP, Grant PJ. Plasminogen-activator inhibitor and coronary heart disease. N Engl J Med 2; 342: Koppel K, Bratt G, Schulman S, Bylund H, Sandström E. Hypofibrinolytic state in HIV-1-infected patients treated with protease inhibitor-containing highly active antiretroviral therapy. J Acquir Immune Defic Syndr 22; 29: Sipsas NV, Sfikakis PP, Touloumi G, Pantazis N, Choremi H, Kordossis T. Elevated serum levels of soluble immune activation markers are associated with increased risk for death in HAART-naive HIV-1-infected patients. AIDS Patient Care STDS 23; 17: Mu H, Chai H, Lin PH, Yao Q, Chen C. Current update on HIV-associated vascular disease and endothelial dysfunction. World J Surg 27; 31: Accepted for publication 17 October 29 Antiviral Therapy