Treatment interruption (TI) has been considered as a potentially useful strategy for limiting exposure to antiretroviral therapy (ART) and thus decreases some of the disadvantages of long-term treatment. Results of clinical studies performed to test the utility of TI have offered controversial results. Several controlled and uncontrolled studies have shown that ART might be safely interrupted for varying periods of time among healthy patients with controlled viremia and high CD4+ cell counts.1-3 Recent reports, however, have found that patients undergoing TI were at significantly higher risk of severe clinical events and death than those continuing ART.4-6
The most unexpected finding to date of clinical trials of CD4-guided TI has been the excess occurrence and death rate associated with nonopportunistic complications in the large Stragegies for Management of Antiretroviral Theraphy (SMART) trial.4 Overall rate of cardiovascular, renal, or hepatic disease in patients who interrupted therapy exceeded any prediction because these events are usually considered to be drug-related adverse events, and patients who discontinue therapy should be relatively protected from them.
The higher rate of cardiovascular disease in patients on TI was particularly intriguing. Previous reports have found a relationship between dyslipidemia associated with antiretroviral drugs and the occurrence of cardiovascular events.7,8 No data are provided in the SMART trial regarding the evolution of lipids in patients who interrupted therapy. We present herein the results of a small, randomized clinical trial of TI in which measurements of lipid fractions were carried out to evaluate lipid disorders in patients who interrupt therapy.
MATERIALS AND METHODS
This is a prospective, randomized study that was designed to assess the immunological, virological, and clinical outcomes of TIs in patients with chronically suppressed human immunodeficiency virus (HIV-1) infection. The study has been carried out at 4 Spanish hospitals. In addition to the main results of the study, the study aimed to evaluate the impact of TIs on the toxicity of ART, including the evolution of lipids. Written informed consent was obtained from all patients, and the Ethics Committee of the participating hospitals approved the study.
Inclusion criteria of an HIV-infected patient to participate in the study were age older than 18 years, stable highly active antiretroviral therapy (HAART), CD4+ >600/μL, and HIV RNA load <50 copies/mL for at least 6 months before discontinuing therapy. Patients were excluded if they had prior acquired immunodeficiency syndrome, prior ART failure, poor adherence to HAART (<90%), or if they were pregnant or had been splenectomized. Participants were randomized to interrupt all drugs simultaneously or to continue the ongoing antiretroviral treatment (groups TI and control, respectively). The randomization was centralized in the study coordinating center, based on a 1:1 random number table.
Patients were clinically evaluated monthly during the first 3 months and every 2 months thereafter. Reintroduction of therapy in patients allocated to the TI group was decided when severe symptoms or laboratory abnormalities related with HIV developed or when CD4 cell count reached value <350 cells/μL in 2 serial determinations within 1 month. In these situations, the prior HAART was reinitiated.
Multiplex suspension bead array immunoassay was performed using the Luminex 100 analyzer (Luminex Corporation, Austin, TX) to identify protein expression in culture supernatants according to the manufacturers' specifications. We assayed monocyte chemotactic protein-1 (MCP-1), tumor necrosis factor (TNF)-α, IL-6, IL-8, adiponectin, total plasminogen activator inhibitor-1 (PAI-1), leptin, and apoproteins [Apo-A1, Apo-B100 (Apo-B), and Apo-E] in plasma of patients according to the user manual. A minimum of 100 events (beads) were collected for each of the analyte protein, and median fluorescence intensities were obtained. Analyte protein concentrations were automatically calculated based on standard curve data using MasterPlex QT Analysis version 2 (MiraiBio, Inc, Alameda, CA). A 5-parameter regression formula was used to calculate the sample concentrations from the standard curves. The coefficients of variation (coefficient of variation = SD/mean) for measurements of either serum sample were <8%. The sensitivity was 0.2-1.5 pg/mL (PAI-1t, IL-6, IL-8, MCP-1, and TNF-α), 85.4 pg/mL (leptin), 145.4 pg/mL (adiponectin), and 0.1-3 pg/mL (Apo-A1, Apo-B, and Apo-E).
We measured T-cell subsets, viral load (VL), and plasmatic lipids. Plasma levels of cholesterol and triglycerides were determined by enzymatic methods that used commercially available reagents (CHOD-PAP reagent; Roche) by means of a Hitachi 747 automatic analyzer. All samples were obtained after overnight fast, and triglycerides and cholesterol were analyzed before 4 hours. T-lymphocyte CD4 cell counts were determined in whole blood by flow cytometry. Plasma HIV-1 RNA was quantified using the Amplicor Ultransensitive Assay (Roche Amplicor HIV-1 Monitor assay, version 1.5) with a limit of detection of 50 copies/mL.
The analysis of variance test was used to compare the means of the 2 groups. The Fisher exact test was used for all other comparisons between groups. Statistical differences between baseline and 12 months of follow-up were studied with nonparametric tests (Wilcoxon for continuous variables or Sign for dichotomous variables). We also analyzed the differences in evolution among the 2 groups of patients. The differences between the end of study and baseline were analyzed using Mann-Whitney U test. All P values were 2 tailed, and the threshold of significance was set at 0.05. All statistical analyses were performed with SPSS 12.0 software (SPSS Inc, Chicago, IL).
Table 1 shows the clinical, immunological, and virological characteristics of HIV-infected patients on TI and on HAART. We did not find any statistical differences between the groups. Both groups were well balanced in all the characteristics recorded. After ART interruption, no patient showed signs of clinical progression. One patient developed an antiretroviral acute syndrome, and 2 patients presented thrombocytopenia for which they had to restart HAART.
As a summary of main results, patients who continued on HAART (control group) maintained undetectable HIV RNA and stable CD4+ above 800/μL during the follow-up (Fig. 1A). Three patients in this group changed therapy (2 for virological failure and 1 for toxicity). Moreover, we did not find any statistical differences on cholesterol and triglyceride values at the baseline and 12 months of follow-up (Fig. 1B). In the TI group, HIV RNA returned to pretreatment values within the first month after interruption of treatment and remained at the same mean level during the follow-up. CD4+ cell count had a significant decrease during the first month (P < 0.05); mirroring plasma HIV RNA had a significant increase during the first month (P < 0.05) (Fig. 1C). Furthermore, we only found a significant decrease in cholesterol levels after 12 months of follow-up (P < 0.001) (Fig. 1D).
We evaluated lipids and some related parameters in the 2 groups and compared the evolution after 12 months (Table 2). In patients who continued on HAART, there were no significant changes in any of the measured parameters, including total cholesterol; triglycerides; apoproteins A1, B, and E; leptin; adiponectin; PAI-1; MCP-1; IL-6; IL-8; and TNF-α. In the TI group, median cholesterol levels decreased (from 215 to 166 mg/dL, P < 0.001), although median triglyceride levels remained unchanged. Of note, we observed a decrease in Apo-A1 (P = 0.048) and Apo-B levels (P < 0.001) and an increase in TNF-α levels (P = 0.034). Given the greater decrease in Apo-B, the ratio Apo-A1/Apo-B increased after 12 months of TI (from 3.4 to 5.1, P = 0.008). We did not find significant variations in leptin or adiponectin levels.
When changes in lipid and lipid-related parameters from baseline to 12 months were compared, patients in the TI group had a higher decrease of Apo-A (P = 0.025) and Apo-B (P = 0.012) than patients in the control group. All the other changes were not significant when the 2 groups were compared.
Moreover, we analyzed the correlation between lipids/lipoproteins and other factors involved in the inflammatory response of all HIV patients at baseline. PAI-1 level has a positive correlation with Apo-B (r = 0.499; P < 0.01) and Apo-E (r = 0.454; P < 0.05) and negative correlation with Apo-A1/Apo-B ratio (r = −0.490; P < 0.01). MCP-1 has a positive correlation with triglyceride levels (r = 0.342; P < 0.05) and a negative correlation with Apo-A1/Apo-B ratio values (r = −0.499; P < 0.01).
This study shows that the lipid profile and apoproteins levels change toward a less atherogenic profile after TI, although no significant changes occur in patients who continue therapy during 12 months of follow-up. We have observed a significant decrease in total cholesterol, Apo-A1, and Apo-B but a significant increase in Apo-A1/Apo-B ratio. These findings argue against an increased cardiovascular risk in patients who undergo TIs.
Before the results of the SMART trial were reported,4 TI had been considered as a safe alternative to continued therapy in patients with controlled viremia and good immunological status.9 In addition, interruption of ART was expected to enhance the quality of life and limit drug-related adverse events.10 Among these events, changes in body fat (lipodystrophy) and increased cardiovascular risk associated to drug-induced dyslipidemia are mostly feared by patients and physicians, and any intervention that could decrease their incidence is certainly needed. TI was viewed as a logical, potential solution, and consequently, the finding of an increased cardiovascular risk in patients interrupting therapy in the SMART trial has been surprising and difficult to explain.
In an attempt to show a relation between lipid variations in patients who interrupt therapy and the increased cardiovascular risk shown in the SMART trial, we have analyzed stored samples from patients included in a small clinical trial of TI. In addition to total cholesterol and triglycerides, we have measured apoproteins and some adipocytokines. Plasma levels of apoproteins have been shown to be reliable predictors of atherogenesis and cardiovascular risk, with some authors advocating for using them in routine clinical practice. In particular, Apo-A1 and Apo-B are proposed as the best markers for coronary risk. Apo-A1 levels are related with high-density lipoprotein levels, and their decrease is associated with increased cardiovascular risk,11 whereas Apo-B is associated with low-density lipoprotein, and high levels are predictive of increased cardiovascular risk.12 The Apo-A1/Apo-B ratio seems to be the best predictor, low values indicating higher risk.11 Patients in the TI group in our study had a significant decrease in both Apo-A1 and Apo-B but with a highly significant increase in the Apo-A1/Apo-B ratio, thus meaning an overall decrease in the atherogenic profile and cardiovascular risk as it is described in other pathologies associated to cardiovascular risk. Moreover, the drop in Apo-A1 may reflect 2 events: stopping the HAART that raises Apo-A by suppressing viral replication and the effect of HIV that lowers Apo-A.13 Accordingly, with previous reports and assumptions, our results suggest that some alterations in lipid metabolism in the setting of HAART reverse with drug withdrawal and that the decrease of plasmatic lipid levels could be of help in decreasing the cardiovascular risk.
The adipose tissue has been recognized as an important source of metabolically active secretory products (adipocytokines) that include TNF-α, IL-6, IL-8, MCP-1, PAI-1, adiponectin, and leptin. They seem to play important regulatory roles in a variety of complex processes, including fat metabolism, but none is without controversy regarding its respective mechanism and scope of action.12,14 These bioactive molecules have been shown to be related to elevated lipolysis and insulin resistance in HIV-infected patients.15,16 No significant changes have been found in several parameters, with the exception of TNF-α, which is most likely related with viral replications and inflammatory activation. Many of the changes in lipoproteins during infection/inflammation help protect the host from harmful effects of the stimuli. In cases of chronic infection, however, these cytokine-induced changes in the structure and function of lipoproteins could be deleterious and may contribute to the development of atherosclerosis. So during infection and inflammation, high-density lipoprotein loses its protective properties.13 Although no significant changes were detected during the follow in any of the 2 groups, we have found high PAI-1 and MCP-1 levels, which were associated with a low Apo-A1/Apo-B ratio. Circulating PAI-1 levels are elevated in patients with coronary heart disease and may play an important role in the development of atherothrombosis.12 In our study, we found elevated PAI-1 in patients with low Apo-A1/Apo-B ratio as in previous report.17 Moreover, hypercholesterolemia severely impairs monocyte function in hypercholesterolemic coronary artery disease patients. Monocyte dysfunction is probably connected to impaired collateral artery growth. The high levels of MCP-1 were described to be associated cardiovascular risk.12 In our study, we found that elevated MCP-1 associated a decreased Apo-A1/Apo-B ratio (high cardiovascular risk).
In summary, the evolution of lipids and apoproteins in patients who interrupt therapy would favor an improvement in the cardiovascular risk. These findings are in full agreement with previous reports showing a high incidence of dyslipidemia in patients who receive antiretroviral drugs and the subsequent increase in cardiovascular risk but are contrary to the increased risk of cardiovascular events and deaths found in some trials. It has to be assumed, then, as proposed by some authors, that this excess risk is not lipid mediated but secondary to the contribution of uncontrolled HIV replication to short-term cardiovascular risk.18
We want to thank Jose Ma Bellón Cano for his statistical assistance.
1. Maggiolo F, Ripamonti D, Gregis G, et al. Effect of prolonged discontinuation of successful antiretroviral therapy on CD4 T cells: a controlled, prospective trial. AIDS
2. Cardiello PG, Hassink E, Ananworanich J, et al. A prospective, randomized trial of structured treatment interruption for patients with chronic HIV type 1 infection. Clin Infect Dis
3. Krolewiecki AJ, Zala C, Vanzulli C, et al. Safe treatment interruptions in patients with nadir CD4 counts of more than 350 cells/microL: a randomized trial. J Acquir Immune Defic Syndr
4. El-Sadr WM, Lundgren JD, Neaton JD, et al. CD4+ count-guided interruption of antiretroviral treatment. N Engl J Med
5. Ananworanich J, Gayet-Ageron A, Le Braz M, et al. CD4-guided scheduled treatment interruptions compared with continuous therapy for patients infected with HIV-1: results of the Staccato randomised trial. Lancet
6. Danel C, Moh R, Minga A, et al. CD4-guided structured antiretroviral treatment interruption strategy in HIV-infected adults in west Africa (Trivacan ANRS 1269 trial): a randomised trial. Lancet
7. Friis-Moller N, Sabin CA, Weber R, et al. Combination antiretroviral therapy and the risk of myocardial infarction. N Engl J Med
8. Sudano I, Spieker LE, Noll G, et al. Cardiovascular disease in HIV infection. Am Heart J
9. Hogg RS, Havlir D, Miller V, et al. To stop or not to stop: that is the question, but what is the answer? AIDS
10. Ruiz L, Paredes R, Gomez G, et al. Antiretroviral therapy interruption guided by CD4 cell counts and plasma HIV-1 RNA levels in chronically HIV-1-infected patients. AIDS
11. Badiou S, Merle De Boever C, Dupuy AM, et al. Fenofibrate improves the atherogenic lipid profile and enhances LDL resistance to oxidation in HIV-positive adults. Atherosclerosis
12. Guzik TJ, Mangalat D, Korbut R. Adipocytokines-novel link between inflammation and vascular function? J Physiol Pharmacol
13. Khovidhunkit W, Kim MS, Memon RA, et al. Effects of infection and inflammation on lipid and lipoprotein metabolism: mechanisms and consequences to the host. J Lipid Res
14. Fantuzzi G. Adipose tissue, adipokines, and inflammation. J Allergy Clin Immunol
. 2005;115:911-919; quiz 920.
15. Vigouroux C, Maachi M, Nguyen TH, et al. Serum adipocytokines are related to lipodystrophy and metabolic disorders in HIV-infected men under antiretroviral therapy. AIDS
16. Gougeon ML, Penicaud L, Fromenty B, et al. Adipocytes targets and actors in the pathogenesis of HIV-associated lipodystrophy and metabolic alterations. Antivir Ther
17. Matsuzawa Y. Therapy insight: adipocytokines in metabolic syndrome and related cardiovascular disease. Nat Clin Pract Cardiovasc Med
18. Wolf K, Tsakiris DA, Weber R, et al. Antiretroviral therapy reduces markers of endothelial and coagulation activation in patients infected with human immunodeficiency virus type 1. J Infect Dis