Boyd
21/11/07
17:16
Page 1265
Antiviral Therapy 12:1265–1271
Relationships between drug exposure, changes in metabolic parameters and body fat in HIV-infected patients switched to a nucleoside sparing regimen Reshma S Autar1,2†, Mark A Boyd1,3,4* † Ferdinand WMN Wit1,2, Kiat Ruxrungtham1,5, Jongkol Sankote1, Joep MA Lange1,2, David A Cooper1,3,4, Praphan Phanuphak1,5, David M Burger 6 and Peter Reiss2 1
The HIV-Netherlands Australia Thailand Research Collaboration - Thai Red Cross AIDS Research Center, Bangkok, Thailand Center for Infection and Immunity Amsterdam & Center for Poverty-related Communicable Diseases, Academic Medical Center, University of Amsterdam, and International Antiviral Therapy Evaluation Center, Amsterdam, The Netherlands 3 National Centre in HIV Epidemiology and Clinical Research, Sydney, Australia 4 University of New South Wales, Sydney, Australia 5 King Chulalongkorn University, Bangkok, Thailand 6 Radboud University Medical Center, Nijmegen, The Netherlands 2
*Corresponding author: Tel: +61 2 9385 0900; Fax: +61 2 9385 0920; E-mail:
[email protected] † These authors contributed equally to this work
Background: The pathogenesis of metabolic disturbances in treated HIV infection is incompletely understood. Methods: Relationships between fasted metabolic parameters, body composition, and drug plasma concentrations were investigated in 59 patients who switched from failed nucleoside analogue treatment to ritonavir-boosted indinavir and efavirenz therapy. Metabolic parameters, peripheral fat, visceral adipose tissue (VAT) and drug plasma concentrations were measured prospectively. Results: Ritonavir exposure was found to be negatively correlated with high-density lipoprotein cholesterol (HDL-c) changes, with a 2.4% decrease in HDL-c for each unit increase in ritonavir concentration ratio. Significant associations between indinavir or efavirenz concentrations and metabolic disturbances were not observed. Total
cholesterol (TC) correlated positively with high body mass index (BMI) and negatively with baseline limb fat mass: each unit increase in BMI and each kilogram reduction in baseline limb fat corresponded with a TC increase of 2.4% and 4.1%, respectively. Baseline triglyceride levels were lower in those patients with relatively greater limb fat mass: each kilogram reduction of total limb fat mass was associated with a 15.7% increase in triglyceride concentration. Changes in VAT were positively correlated with TC: for every unit TC increase a 0.3% VAT increase was observed (over 48 weeks). Conclusions: Reduced limb fat mass at the start of the study treatment, increases in VAT mass, and higher plasma concentrations of ritonavir on study treatment were each – to varying degrees – associated with various metabolic disturbances.
Introduction HIV infection and antiretroviral therapy (ART) are associated with dyslipidaemia [1]. Changes in body fat distribution (lipodystrophy, characterized by peripheral lipoatrophy and/or visceral lipoaccumulation) have been observed since the advent of combination ART, but the pathophysiology is only partially understood [2,3]. Lipoatrophy and lipoaccumulation were thought to be associated and to have a common pathophysiology, but recent research suggests that they may to some degree be independent phenomena, and might even have separate pathophysiological mechanisms. Lipoatrophy appears to be associated with the use of nucleoside © 2007 International Medical Press 1359-6535
reverse transcriptase inhibitors (NRTIs), whereas lipoaccumulation and dyslipidaemia are associated with the use of protease inhibitors (PIs) [4–9]. There appears to be little or no association of the nonnucleoside reverse transcriptase inhibitors (NNRTIs) with changes in body fat distribution, although they are associated with changes in plasma lipids [10]. In the HIV Netherlands Australia Thailand Research Collaboration 009 study (HIV-NAT 009), a group of PI and NNRTI treatment-naive HIVinfected patients failing combination NRTI therapy were switched to ritonavir-boosted indinavir 800/100 mg twice daily and efavirenz 600 mg once 1265
Boyd
21/11/07
17:16
Page 1266
RS Autar et al.
daily; this regimen was observed to have a favourable virological treatment response [11]. After 96 weeks, 69% of patients had plasma HIV RNA 18 years, a documented positive HIV test, adequate use of contraception, treatment failure defined as a composite endpoint with either: (i) plasma HIV-1 RNA >5,000 copies/ml; (ii) plasma HIV-1 RNA >1,000 copies/ml accompanied by a CD4+ T-cell count 5 mmol/l, n HDL-c, mmol/l HDL-c 2 mmol/l, n
59 36:23 32 ±6.6 56.4 ±9.60 21.5 ±2.9 50 ±13 16 (27) 40 (68) 16 ±18 24 (14) 57 (33) 19 (11) 176 ±127 4.6 ±4.7 4.4 ±0.96 18 (31) 0.9 ±0.24 33 (56) 1.6 ±0.85 16 (27)
Previous treatment consisted of various combinations of zidovudine (ZDV), d4T, didanosine (ddI), lamivudine (3TC) and abacavir (ABC). The mean (SD) NRTI exposure was 40 (13) months; 36% of patients had been exposed to two NRTIs, 12% to three NRTIs, 48% to four NRTIs and 3% to five NRTIs. Forty patients had prior d4T exposure with a mean duration of 16 (18) months. Of these 40 patients, 36 had taken d4T with ddI for a mean duration of 15 (18) months. At the time of switch to the NRTI-sparing regimen, 16 (27%) patients were receiving a d4Tcontaining NRTI combination and 42 (70%) were receiving a ZDV-containing NRTI combination.
On-study medication Nine patients had a temporary treatment interruption of study medication. In total 51, 50, 46, 42, 40, 38, 33 and 34 patients were available for analyses at study weeks 4, 12, 24, 36, 48, 60, 72 and 96, respectively. Following the switch to the study regimen, 15 patients in total commenced lipid-lowering therapy over the 96 weeks of study (5 by week 12, 6 by week 24, 8 by week 36, 11 by week 48, 12 by week 60, 14 by week 72, and 15 by week 96). One patient started an oral hypoglycaemic agent at week 72; this patient had commenced treatment with a lipid-lowering agent at week 60.
Plasma concentrations of efavirenz and concentration ratios of indinavir and ritonavir A total of 334 plasma concentrations were available for analysis. The CR of indinavir and ritonavir and the plasma concentrations of efavirenz are summarized as medians and interquartile range (IQR). The median plasma concentration of indinavir was 7.5 (IQR 1.4–12.5) mg/l; the median plasma concentration of ritonavir was 1.7 (IQR 0.8–3.3) mg/l. The median CR for indinavir and ritonavir were 1.3 (IQR 0.8–1.9) and 1.1 (IQR 0.6–2.1), respectively. The median plasma concentration of efavirenz was 3.1 mg/l (IQR 2.3–4.3 mg/l).
Predictors of metabolic parameters and body fat composition Figures 1 A, B, C and D demonstrate the changes in TC, HDL-c, TG and VAT, respectively, over time. Table 2 summarizes the significant predictors of these four variables. During the 96 weeks of treatment, the modelled TC concentrations increased by ~77%. On analysis we found a significant positive association between the changes in TC and the time-updated BMI. An increase Antiviral Therapy 12:8
*Mean ±SD, unless otherwise specified. BMI, body mass index; CDC, Centres for Disease Control; d4T stavudine; HDL-c, high-density lipoprotein cholesterol; NRTI, nucleoside reverse transcriptase inhibitor.
of 1 unit in the BMI was associated with a TC increase of 2.4% (95% confidence interval [CI] 0.63–4.22; P=0.008). A negative correlation was observed between baseline limb fat mass and the change in TC concentration: that is, for every 1 kg reduction in total limb fat mass at baseline, there was an increase in TC of 4.1% (95% CI 1.43–6.72; P=0.003). We observed a 48% modelled increase in HDL-c after 96 weeks of treatment. Ritonavir CR was negatively correlated with the change in HDL-c: that is, a 2.4% (95% CI 0.67–4.18; P=0.007) decrease in HDL-c was observed for each unit increase in ritonavir CR. Significant correlations between HDL-c and indinavir and efavirenz plasma concentrations were not observed. A 226% modelled increase in TG from baseline was seen after 96 weeks of study. TG increases were not significantly correlated with the drug plasma concentrations; however, a strong negative correlation was observed with the baseline limb fat mass. For every 1 kg reduction of total limb fat mass at baseline, there was an increase of 15.7% (95% CI 4.79–26.75; P=0.006) in TG concentration. VAT mass increased by 20% over the 96 weeks of study. TC concentrations were positively correlated with increases in VAT: for every mmol/l TC increase we found a 0.3% VAT increase over 48 weeks (95% 1267
21/11/07
17:16
Page 1268
RS Autar et al.
Figure 1. Lipid parameters and visceral adipose tissue changes over 96 weeks
A
All subjects Low baseline limb fat High baseline limb fat
B
All subjects Low ritonavir levels High ritonavir levels
10
2 HDL-c, mmol/l
Cholesterol, mmol/l
8 6 4
1
2 0
0 4 12
n= 5951 50
24 36 48 60 72 Time since start of study, weeks 46
42
40
C
38
33
0
96
8 VAT, cm2
4 2 0 0 4 12 n= 5951 50
24 36 48 60 72 Time since start of study, weeks 46
42
40
38
33
96 34
24 36 48 60 72 Time since start of study, weeks 46
42
40
38
D
10
6
0 4 12
n= 59 51 50
34
All subjects Low baseline limb fat High baseline limb fat
Triglycerides, mmol/l
Boyd
33
96 34
All subjects Low cholesterol High cholesterol 180 160 140 120 100 80 60 40 20 0
0 4 12
n= 59 51 50
24 36 48 60 72 Time since start of study, weeks 46
42
40
38
33
96 34
Changes in mean (SD) total cholesterol (A) high-density lipoprotein cholesterol (HDL-c) (B) and triglycerides (C) over time. Changes in mean visceral adipose tissue (VAT) (D). The two additional traces in each figure illustrate the profiles over time for patients stratified above and below the mean for the expressed variable.
CI 0.10–0.49; P=0.001). Significant correlations between VAT and TG were not observed.
Discussion In the HIV-NAT 009 study we observed a substantial deterioration in fasted metabolic parameters over 96 weeks of ritonavir-boosted indinavir plus efavirenz treatment. All patients in the study had been extensively pretreated with thymidinecontaining combination NRTI therapies. Several studies have shown that in patients who discontinue the use of thymidine analogues (in particular d4T) and switch to non-thymidine analogues (in particular tenofovir), dyslipidaemia can improve [13,14]. 1268
However, despite the switch to an NRTI-sparing regimen the metabolic profile of the patients in this cohort deteriorated, while the subcutaneous and visceral fat compartments increased in size. Data from other studies confirm that antiretroviral regimens consisting of NNRTIs and PIs are associated with substantial lipid abnormalities in HIV-1infected patients. In the ACTG 5125s study, switch to efavirenz and lopinavir resulted in increases in TC, specifically non-LDL-c, compared with efavirenz with two NRTIs [15]. The present analysis was conducted to investigate possible determinants of the resultant metabolic profile and the increase in VAT and total limb fat in this patient population. We found that patients with © 2007 International Medical Press
Boyd
21/11/07
17:16
Page 1269
Metabolic toxicity and drug plasma concentrations
Table 2. Results of multivariate linear regression analyses Parameter Total cholesterol Body mass index Baseline limb fat mass High-density lipoprotein cholesterol Concentration ratio of ritonavir Triglycerides Baseline limb fat mass Change in visceral adipose tissue* Mean total cholesterol
Change, %
95% CI
P-value
2.4/kg/m2 4.1/kg
0.63–4.22 -6.72–-1.43
0.008 0.003
-2.4/unit
-4.18–-0.67
0.007
-15.7/kg
-26.75–-4.79
0.006
0.29/mmol/l
0.10–0.49
0.005
All analyses were adjusted for the baseline value of the dependent variable. *Mean concentrations of first 48 weeks were analysed. CI, confidence interval.
reduced limb fat mass at study baseline had proportionately greater increases in TC following treatment switch compared to those with greater total limb fat mass at baseline. The patients who sustained the greatest VAT accumulation on study also demonstrated greater TC disturbances. Patients with a higher BMI at time of switch also developed higher TC. Patients with higher TG at baseline had lower total limb fat mass at study entry. We found that patients with relatively less baseline limb fat had a more pronounced increase in TC. This is consistent with the findings of a randomized study of a switch from a PI to either abacavir, nevirapine or efavirenz, in which those patients with relatively greater baseline peripheral lipoatrophy demonstrated the least improvement in dyslipidaemia [16]. This apparent inverse relationship should be taken into account in the context of changing ART in patients with lipoatrophy in the clinic. With respect to the relationship between drug plasma concentrations and metabolic abnormalities, we found that higher ritonavir plasma concentrations were statistically significantly associated with lower HDL-c. This appears to be in agreement with studies performed in healthy volunteers who received highand low-dose ritonavir [17]. Recent studies suggest that the addition of low-dose ritonavir to atazanavir is associated with elevations of TC and TG as well as reductions in HDL-c in healthy volunteers, while atazanavir as single PI does not have this association [18]. We had anticipated that there might exist a relationship between PI concentration, in particular indinavir, and VAT but did not find one [7]. This is consistent with data from recent studies, which suggest that boosted PIs might not be associated with VAT accumulation [19]. Furthermore, we expected that increases in TG would be associated with visceral fat accumulation. However, we found that higher TC, but not TG, was significantly associated with VAT increases. It is important to consider, Antiviral Therapy 12:8
however, that several counteracting processes might influence the changes in TG and TC in these patients, thereby rendering it difficult to determine any straightforward relationships between the plasma concentration of any individual drug and lipid changes. Exposure to ritonavir even at pharmacokinetic ‘booster’ doses, but not unboosted indinavir, is known to be associated with rapid increases in TG and TC [20]. In contrast, withdrawal of thymidine analogue NRTIs, and d4T in particular, is known to result in rapid declines of TG and TC concentration [21]. The latter observation might be explained by a reversal of the hampered clearance of TG from the bloodstream in HIV-infected patients, with lipoatrophy occurring before any measurable improvement in total peripheral fat mass [22]. There are several weaknesses of this study. Firstly, we retrospectively correlated the results of the various HIV-NAT 009 substudies, which were not prospectively designed to answer the questions posed here. Secondly, the drug concentrations were randomly obtained and correction for time was therefore necessary, which resulted in the concentration ratios. By using the concentration ratio it was possible to minimalize any variation caused by timing of the samples; however, a better approach would be to use actual measured drug concentrations. Thirdly, the study design does not allow for randomized comparisons because of the lack of a control group. Lastly, the sample size was small and therefore our analyses might have little power to detect clinically significant associations. As multiple comparisons were made in this study, we elected to reduce the significance level to an α-level of 0.01 in order to reduce the possibility of finding chance associations; however, this strategy cannot exclude the possibility that some relationships might be confounded. In summary, we found that reduced limb fat mass at the start of the study treatment, increases in VAT mass while on study treatment, and higher plasma 1269
Boyd
21/11/07
17:16
Page 1270
RS Autar et al.
concentrations of ritonavir were each – to varying degrees – associated with metabolic disturbances characterized by higher TC, lower HDL-c and higher TG levels. Patients with less limb fat might have a higher risk of developing an unfavourable lipid profile after switch to a ritonavir-boosted-PI and NNRTI regimen. Whether the deterioration of lipid profile poses a risk of developing cardiovascular diseases needs to be studied further.
3.
Brinkman K, Smeitink JA, Romijn JA, Reiss P. Mitochondrial toxicity induced by nucleoside-analogue reverse-transcriptase inhibitors is a key factor in the pathogenesis of antiretroviral-therapy-related lipodystrophy. Lancet 1999; 354:1112–1115.
4.
Palella FJ, Jr., Cole SR, Chmiel JS, et al. Anthropometrics and examiner-reported body habitus abnormalities in the multicenter AIDS cohort study. Clin Infect Dis 2004; 38:903–907.
5.
Wand H, Law M, Emery S, Cooper DA, Carr A. Increase in limb fat after nucleoside analogue cessation is not associated with decreased visceral fat and has different risk factors. 7th International Workshop on Adverse Drug Reactions and Lipodystrophy in HIV. Antivir Ther 2005; 10 Suppl 3: L1–L67.
6.
Mallal SA, John M, Moore CB, James IR, McKinnon EJ. Contribution of nucleoside analogue reverse transcriptase inhibitors to subcutaneous fat wasting in patients with HIV infection. AIDS 2000; 14:1309–1316.
7.
Miller KD, Jones E, Yanovski JA, Shankar R, Feuerstein I, Falloon J. Visceral abdominal-fat accumulation associated with use of indinavi. Lancet 1998; 351:871–875.
8.
van der Valk M, Gisolf EH, Reiss P, et al. Increased risk of lipodystrophy when nucleoside analogue reverse transcriptase inhibitors are included with protease inhibitors in the treatment of HIV-1 infection. AIDS 2001; 15:847–855.
9.
Bacchetti P, Gripshover B, Grunfeld C, et al. Fat distribution in men with HIV infection. J Acquir Immune Defic Syndr 2005; 40:121–131.
Acknowledgements We thank the patients, HIV-NAT staff, Dr Anchalee Avihingsanon for supervising the ongoing care of the cohort, Sasiwimol Ubolyam, Meena Gorowara for laboratory coordination and analyzing drug plasma concentrations, and Noppong Hirunwadee for data entry.
Conflicts of interest Mark Boyd has received a medical school grant and sponsorship for conference attendance from Merck & Co.; Saskia Autar and Jongkol Sankote did not receive grants or any other sponsorships from Merck & Co.; Ferdinand Wit has received sponsorship for conference attendance from Merck & Co.; Kiat Ruxrungtham, Joep Lange, David Cooper and Praphan Phanuphak received study grants, served as scientific topic speakers and were advisory members for Merck & Co.; David Burger is a paid consultant for and received honoraria from Merck & Co.; Peter Reiss has received medical school grants from Merck & Co.
Funding This study was financially supported by The HIVNetherlands Australia Thailand Research Collaboration (HIV-NAT) – Thai Red Cross AIDS Research Center (TRCARC), The Center for Infection and Immunity Amsterdam (CINIMA) & Center for Poverty-related Communicable Diseases, Academic Medical Center (AMC), University of Amsterdam, and International Antiviral Therapy Evaluation Center (IATEC), National Center in HIV Epidemiology and Clinical Research (NCHECR), the Radboud University Medical Center and the Flinders Medical Centre Foundation of South Australia.
References 1.
2.
1270
Grunfeld C, Pang M, Doerrler W, Shigenaga JK, Jensen P, Feingold KR. Lipids, lipoproteins, triglyceride clearance, and cytokines in human immunodeficiency virus infection and the acquired immunodeficiency syndrome. J Clin Endocrinol Metab 1992; 74:1045–1052. Carr A, Samaras K, Chisholm DJ, Cooper DA. Pathogenesis of HIV-1-protease inhibitor-associated peripheral lipodystrophy, hyperlipidaemia, and insulin resistance. Lancet 1998; 351:1881–1883.
10. Boyd MA, Carr A, Ruxrungtham K, et al. Changes in body composition and mitochondrial nucleic Acid content in patients switched from failed nucleoside analogue therapy to ritonavir-boosted indinavir and efavirenz. J Infect Dis 2006; 194:642–650. 11. Boyd MA, Siangphoe U, Ruxrungtham K, et al. Indinavir/ritonavir 800/100 mg bid and efavirenz 600 mg qd in patients failing treatment with combination nucleoside reverse transcriptase inhibitors: 96-week outcomes of HIV-NAT 009. HIV Med 2005; 6:410–420. 12. Boyd MA, Aarnoutse RE, Ruxrungtham K, et al. The pharmacokinetics of indinavir/ritonavir 800/100 mg in combination with efavirenz 600 mg in HIV-1 infected subjects. J Acquir Immune Defic Syndr 2003; 34:134–139. 13. Carr A, Workman C, Smith DE, et al. Abacavir substitution for nucleoside analogs in patients with HIV lipoatrophy: a randomized trial. JAMA 2002; 288:207–215. 14. 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 2006; 20:2043–2050. 15. Tebas P, Zhang J, Yarasheski K, et al. Switching to a protease inhibitor-containing, nucleoside-sparing regimen (lopinavir/ritonavir plus efavirenz) increases limb fat but raises serum lipid levels: results of a prospective randomized trial (AIDS clinical trial group 5125s). J Acquir Immune Defic Syndr 2007; 45:193–200. 16. Fisac C, Fumero E, Crespo M, et al. Metabolic benefits 24 months after replacing a protease inhibitor with abacavir, efavirenz or nevirapine. AIDS 2005; 19:917–925. 17. Shafran SD, Mashinter LD and Roberts SE. The effect of low-dose ritonavir monotherapy on fasting serum lipid concentrations. HIV Med 2005; 6:421–425. 18. Malan DR, Krantz E, David N, Wirtz V, Hammond J, McGrath D; for the 089 Study Group. Efficacy and safety of atazanavir, with or without ritonavir, as part of once-daily highly active antiretroviral therapy regimens in antiretroviral-naive patients. J Acquir Immune Defic Syndr 2007 Oct 25; [Epub ahead of print]. © 2007 International Medical Press
Boyd
21/11/07
17:16
Page 1271
Metabolic toxicity and drug plasma concentrations
19. Carr A, Zajdenuerg R, Workman C et al. Effects of tipranavir/r (500/200 or 500/100 mg bid) in comparison with lopinavir/r (400/100 mg bid) on changes in body composition and metabolic parameters in ARV-naive patients over 48 weeks. Antivir Ther 2007; 12 Suppl 2:L8. Abstract O-07. 20. Boyd MA, Srasuebkul P, Khongphattanayothin M, et al. Boosted versus unboosted indinavir with zidovudine and lamivudine in nucleoside pre-treated patients: a randomized, open-label trial with 112 weeks of follow-up (HIV-NAT 005). Antivir Ther 2006; 11:223–232.
21. Suleiman JM, Lu B, Enejosa J, Cheng A. Improvement in lipid parameters associated with substitution of stavudine (d4T) to tenofovir DF (TDF) in HIV-infected patients participating in GS 903. Program and abstract of the 44th Annual International Conference on Antimicrobial Agents and Chemotherapy; October 30–November 2, 2004; Washington, USA [Abstract H-158]. 22. van Wijk JP, Cabezas MC, de Koning EJ, Rabelink TJ, van der Geest R, Hoepelman IM. In vivo evidence of impaired peripheral fatty acid trapping in patients with human immunodeficiency virus-associated lipodystrophy. J Clin Endocrinol Metab 2005; 90:3575–3582.
Accepted for publication 19 August 2007
Antiviral Therapy 12:8
1271
Boyd
21/11/07
17:16
Page 1272
