Epicardial adipose tissue and atherogenesis: EAT your heart out EDITORIAL
Download the PDF here
Download the PDF here
"HIV-infected men had greater EAT than HIV-uninfected men (P = 0.001). EAT was positively associated with duration of antiretroviral therapy (P = 0.02), specifically azidothymidine (P < 0.05). EAT was associated with presence of any coronary artery plaque (P = 0.006) and noncalcified plaque (P = 0.001), adjusting for age, race, serostatus, and cardiovascular risk factors. Among men with CAC, EAT was associated with CAC extent (P = 0.006). HIV serostatus did not modify associations between EAT and either CAC extent or presence of plaque.
Greater epicardial fat volume in HIV-infected men and its association with coronary plaque and antiretroviral therapy duration suggest potential mechanisms that might lead to increased risk for cardiovascular disease in HIV.........In summary, among men with or at risk for HIV infection, increased epicardial fat deposition is associated with HIV infection, duration of treatment with HAART, and more plaque in the coronary arteries. Further studies are needed to investigate which additional aspects of HIV infection or its treatment are responsible for increased epicardial adiposity, potential therapies to reduce epicardial fat, and whether greater EAT increases risk for cardiovascular events in HIV patients."
"Epicardial adipose tissue (EAT) is a visceral adipose tissue surrounding the heart and the coronary arteries. Because of its endocrine and paracrine activity, secreting pro-inflammatory and anti-inflammatory cytokines and chemokines, it has been suggested to influence coronary atherosclerosis development (1,2,3,4,5). EAT is associated with cardiovascular risk factors (6,7), coronary atherosclerosis (8,9,10), and prevalent coronary artery disease (11,12,13). In addition, a case-control study, drawn from the MESA (Multi-Ethnic Study of Atherosclerosis), suggested a role of increased EAT volume for coronary event manifestation (14). However, to date, large-population-based longitudinal data on the prognostic value of EAT fat for prediction of hard coronary events are lacking....Cardiac computed tomography (CT) imaging of the heart is the gold standard for EAT quantification with non-contrast enhanced cardiac CT enabling risk assessment through coronary artery calcification (CAC) quantification (15,16,17)......In this study, we examined the association of epicardial fat volume with traditional cardiovascular risk factors, CAC scores, and incident fatal and nonfatal coronary events in the population-based Heinz Nixdorf Recall cohort study. We found a significant association of EAT volume with cardiovascular risk factors in univariable and multivariable analyses. By contrast, although CAC scores increased with quartiles of EAT volume, this effect did not hold when adjusting for traditional cardiovascular risk factors, indicating that the association of EAT with CAC is ultimately explained by shared risk factors. EAT volume was significantly associated with coronary events, independent of traditional cardiovascular risk factors. This effect remained unchanged after further adjustment for the CAC score, which itself is a strong predictor for coronary events. Notably, the association of EAT volume with coronary events was more distinct in subjects with no or mild CAC, which-together with the lack of association of EAT with CAC-may support the hypothesis that EAT may be linked with future coronary events through a mechanistic pathway different from that of coronary calcification, such as early and noncalcified plaque burden. Our finding of increased pericoronary fat at the coronary vessel causing a coronary event during follow-up may suggest that local visceral adipose tissue supports the development of atherosclerosis in the underlying coronary vasculature. [Association of Epicardial Fat With Cardiovascular Risk Factors and Incident Myocardial Infarction in the General Population" Journal of American College of Cardiology April 2013 http://content.onlinejacc.org/article.aspx?articleid=1654988]
Epicardial adipose tissue and atherogenesis: EAT your heart out EDITORIAL
AIDS July 17 2014
MGH Program in Nutritional Metabolism, Harvard Medical School, Boston, Massachusetts, USA.
Cardiovascular disease (CVD) is increasingly recognized among HIV-infected patients, and carefully performed cohort studies show increased relative risk ratios of 1.5-2.0 [1,2]. Moreover, these studies suggest that traditional risk factors, while accounting for some degree of this increase in relative risk, do not account for all or even a major portion of the increased risk . Insights into the mechanisms of this disease have been forthcoming from imaging studies, in which increased noncalcified plaque has been shown among HIV-infected men and also in HIV-infected women compared with age and BMI-matched controls [3,4].
This observation is significant because noncalcified plaque may be more vulnerable and prone to rupture. In contrast, although calcium score is itself a marker for increased CVD risk, it may represent a process by which plaque becomes more fibrotic and mechanically more stable, and thus, less prone to rupture. Recent studies assessing detailed measures of plaque morphology take these observations one step further and suggest that HIV-infected patients (men in this instance), demonstrate increased features of high-risk morphology plaque, including positive remodeling and low attenuation, fatty lesions . In prior studies, traditional risk factors, such as age, hypertension, cholesterol and Framingham score, were shown to segregate more with calcified plaque, and non-traditional risk factors, such as increased immune activation indices (sCD163 and sCD14), with noncalcified plaque and high-risk morphology plaque [3,5]. Moreover, recent studies suggest increased arterial inflammation in HIV-infected patients, a reflection of macrophage infiltration, which may provide fertile ground for the development of high risk morphology plaque and increased propensity for plaque rupture . Taken together, these studies and many others suggest that immune activation and other as yet unknown factors may contribute to the development of increased noncalcified plaque, even among those with well treated HIV. The presence of noncalcified plaque with high-risk morphology may contribute to the recent observation of increased rates of sudden cardiac death in HIV-infected patients .
In recent years, focus has turned to the epicardial adipose tissue (EAT) as a depot of ectopic adipose tissue, which may share the same embryonic origin as abdominal visceral adipose tissue (VAT). Via secretion of cytokines and other paracrine factors, EAT may contribute to atherogenesis in juxtaposed coronary artery segments. Alternatively, common factors may contribute to the development of this adipose tissue depot and coronary artery disease. This is an active area of investigation and the article in this issue of AIDS by Brener et al. is timely and advances the field. This study takes advantage of coronary computed tomographic angiography (CCTA) imaging performed among patients in the MACS cohort. Using CCTA, the investigators evaluated coronary segments for presence or absence of any plaque and further characterized the plaque as calcified, noncalcified or mixed plaque. The Agatston calcium score was obtained as well.
The study included over 900 patients, and is thus the largest CT imaging study to date in the HIV population. Importantly, the investigators measured EAT volume, along with visceral adipose tissue. Of note, there were small but significant differences in age between the HIV and non-HIV groups in the study by Brener et al., but these differences were accounted for in adjusted analyses comparing EAT between HIV and controls.
One of the major findings of the study is that noncalcified plaque was increased in the HIV-infected vs. non-HIV-infected patients. This observation confirms and extends the observations from smaller prior studies. Noncalcified plaque was increased to a greater degree vs. non-HIV controls than was calcified plaque. Indeed, if anything, the prevalence of calcified plaque was less in HIV-infected patients vs. controls, although this difference did not reach significance. The study of Brener et al.  was limited to men, but recent studies in women show the same dichotomy, with even more striking differences in noncalcified plaque vs. non-HIV-infected controls . Assessment of specific morphological features including low attenuation and positive remodeling indices was not performed by Brener et al.
A second major finding by Brener et al. relates to EAT, which was increased in the HIV group after adjustment for age and other factors. Prior studies, matching on age, race and BMI, have demonstrated increased EAT in HIV vs. non-HIV-infected patients . Moreover, prior studies have demonstrated a relationship between EAT and carotid IMT , coronary artery calcium (CAC) score [11,12], CAC progression  and prior CVD events  in HIV-infected patients. In novel data, Brener et al. now demonstrate that EAT is significantly associated with increased prevalence of noncalcified plaque, and this relationship remains significant after adjustment for CVD risk factors among HIV-infected patients. The magnitude of this effect was such that an approximate 8% increase in EAT was associated with a 5-6% increase in prevalence of noncalcified plaque. Notably, the relationship fell out after adjustment for visceral fat in the HIV group but not the non-HIV group. EAT is significantly related to VAT, and thus adjusting for VAT affects this relationship in multivariate modeling among HIV-infected patients. Among those with any plaque present, EAT was most highly related to the extent of CAC among the HIV-infected patients. EAT was significantly related to a number of metabolic indices in addition to VAT, including glucose, insulin, and triglyceride concentrations, and duration ART.
What are we to make of these new data? First, they strongly suggest that HIV-infected men have a unique pattern of plaque on CTA, characterized by increases in noncalcified plaque. Second, they demonstrate increased EAT in HIV-infected men, which may relate to increased plaque as well as critical metabolic parameters. As the study by Brener et al. is cross-sectional, we do not know the sequence of development of EAT, nor how the development of EAT relates temporally to metabolic abnormalities or plaque changes. However, the study raises a number of tantalizing questions for the field. Is the development of EAT, an ectopic adipose depot related to VAT, contributing to atherogenesis in the adjacent coronary vasculature? Does EAT relate specifically to high-risk morphology plaque or more to calcified plaque? If so, what is the mechanism of this effect? Will therapies aimed at reducing related VAT depots work to reduce EAT? Will strategies aimed at reducing atherogenesis, including statin therapy, affect EAT? Until longitudinal, randomized studies are completed, we will not know the answers to these important questions, but the study by Brener et al., by extending our knowledge of this critical area, adds a new piece to the puzzle.
Epicardial fat is associated with duration of antiretroviral therapy and coronary atherosclerosis
Brener, Michaela, ; Ketlogetswe, Kerunnea, ; Budoff, Matthewb; Jacobson, Lisa P.c; Li, Xiuhongc; Rezaeian, Pantehab; Razipour, Aryabodb; Palella, Frank J.d; Kingsley, Lawrencee; Witt, Mallory D.b; George, Richard T.a; Post, Wendy S.a,c
aJohns Hopkins University School of Medicine, Baltimore, MarylandbLos Angeles Biomedical Research Institute at Harbor-UCLA Medical Center, Torrance, CaliforniacJohns Hopkins Bloomberg School of Public Health, Baltimore, MarylanddNorthwestern University, Chicago, IllinoiseUniversity of Pittsburgh, Pittsburgh, Pennsylvania, USA. Michael Brener and Kerunne Ketlogetswe contributed equally to this work.
In a well characterized cohort of men with or at risk for HIV infection, we found that HIV infection was associated with increased epicardial adiposity. Duration of HAART use and treatment with azidothymidine were associated with increased epicardial fat among HIV-infected men, whereas other markers of HIV disease stage and control and other ART drugs showed no association.
Additionally, epicardial tissue volume was associated with subclinical coronary atherosclerosis, which was independent of other measures of adiposity, including BMI and AVF volume. Epicardial fat was associated with the presence of any plaque and of noncalcified plaque and the extent of CAC, and these associations remained after adjusting for cardiovascular risk factors. The associations of epicardial fat with coronary plaque did not significantly differ by HIV serostatus. The greater volume of epicardial fat in HIV-infected men and its association with coronary plaque and duration of HAART use may suggest potential mechanisms that might lead to increased risk for cardiovascular disease in patients with HIV.
In our study, we confirmed the independent association of epicardial fat with numerous cardiovascular risk factors and elements of the metabolic syndrome. The difference in epicardial fat between HIV-infected and uninfected men persisted even after adjustment for these well established risk factors, suggesting an additional mechanism beyond traditional risk factors to explain the difference.
Strengths of our study include the large sample size and use of CCTA evaluation that allowed detailed assessment of plaque morphology among HIV-infected and uninfected men. In addition, we measured the full epicardial fat volume, which is a more reproducible and superior measure than epicardial fat thickness and we defined epicardial fat separately from pericardial fat by identifying the pericardial sac . Furthermore, our study addresses many of the limitations of the two previous studies conducted on this subject. For example, although Guaraldi et al. studied a large group of HIV-infected participants, they did not include an HIV-uninfected control group . In contrast, the study by Lo et al. incorporated HIV-uninfected controls, but was limited by a smaller sample size . Our study utilized the strengths of a large cohort of HIV-infected and HIV-uninfected men with similar environmental exposures (all are MSM).
The use of CCTA in addition to calcium scoring further distinguishes our study and adds to our ability to characterize subclinical atherosclerotic lesions.
This modality allowed examination of both calcified and noncalcified plaques, which is of interest as noncalcified plaques are earlier atherosclerotic lesions that may have a greater necrotic lipid core. Identification of these noncalcified plaques, in addition to calcified plaque seen with noncontrast CT scans, allows more accurate characterization of a participant's total plaque burden . We and others have demonstrated that HIV-infected men are more likely to have noncalcified plaque than calcified plaque [28,29]. Studies have demonstrated that noncalcified plaques are more prone to rupture than calcified plaques, leading potentially to acute coronary syndromes [30,31]. Thus, from a design perspective, our study combines the best features of the previous investigations and attends to their shortcomings with a large sample size, an adequate control population, and advanced coronary atherosclerosis imaging technology.
Studies have demonstrated an independent association between EAT and coronary atherosclerosis, and suggested EAT as a risk factor for accelerated progression of subclinical coronary atherosclerosis in the general population [32-35].
However, results were conflicting in patients with HIV [10,13]. Our study adds to the growing evidence that HIV infection may affect vascular health. HIV infection and/or its treatment may cause an alteration in the distribution of visceral adipose tissue, including the epicardium. When adipose tissue envelops the coronary arteries, it may exert a damaging paracrine effect possibly via a milieu of cytokines that promotes inflammation, vessel damage, and leads to resultant atherosclerosis . Our study profiles the early stages of this effect by examining subclinical atherosclerosis parameters like CAC and plaque characteristics, which have been shown in the general population to predict cardiovascular events .
Despite these strengths, our study has some limitations. This is a cross-sectional analysis and, thus, no causality can be inferred from our findings. Given that most of the HIV-infected men were on treatment, we were unable to segregate the effects of HIV infection and ART in our analyses. Therefore, we cannot claim with certitude which of the two is largely responsible for the findings of increased epicardial adiposity. Additionally, this study was conducted only in men and it is unknown whether these findings can be generalized to women.
In summary, among men with or at risk for HIV infection, increased epicardial fat deposition is associated with HIV infection, duration of treatment with HAART, and more plaque in the coronary arteries. Further studies are needed to investigate which additional aspects of HIV infection or its treatment are responsible for increased epicardial adiposity, potential therapies to reduce epicardial fat, and whether greater EAT increases risk for cardiovascular events in HIV patients.
Cytokines released by epicardial fat are implicated in the pathogenesis of atherosclerosis. HIV infection and antiretroviral therapy have been associated with changes in body fat distribution and coronary artery disease. We sought to determine whether HIV infection is associated with greater epicardial fat and whether epicardial fat is associated with subclinical coronary atherosclerosis.
We studied 579 HIV-infected and 353 HIV-uninfected men aged 40-70 years with noncontrast computed tomography to measure epicardial adipose tissue (EAT) volume and coronary artery calcium (CAC). Total plaque score (TPS) and plaque subtypes (noncalcified, calcified, and mixed) were measured by coronary computed tomography angiography in 706 men.
We evaluated the association between EAT and HIV serostatus, and the association of EAT with subclinical atherosclerosis, adjusting for age, race, and serostatus and with additional cardiovascular risk factors and tested for modifying effects of HIV serostatus.
HIV-infected men had greater EAT than HIV-uninfected men (P = 0.001). EAT was positively associated with duration of antiretroviral therapy (P = 0.02), specifically azidothymidine (P < 0.05). EAT was associated with presence of any coronary artery plaque (P = 0.006) and noncalcified plaque (P = 0.001), adjusting for age, race, serostatus, and cardiovascular risk factors. Among men with CAC, EAT was associated with CAC extent (P = 0.006). HIV serostatus did not modify associations between EAT and either CAC extent or presence of plaque.
Greater epicardial fat volume in HIV-infected men and its association with coronary plaque and antiretroviral therapy duration suggest potential mechanisms that might lead to increased risk for cardiovascular disease in HIV.
Infection with HIV and treatment with antiretroviral therapy (ART) have been implicated in the pathogenesis of coronary heart disease (CHD) [1-5]. Questions remain as to the mechanisms by which HIV infection or its treatments might lead to CHD. The use of ART is accompanied by changes in fat distribution and metabolic abnormalities including insulin resistance, and proatherogenic serum lipid changes [6,7]. Expanding or altered visceral fat depots may play a role in facilitating atherosclerosis. These visceral fat depots are metabolically active and harbor an inflammatory milieu that promotes atherosclerosis. Epicardial fat, in particular, may play a unique role in atherosclerosis because of its close proximity to the coronary vessels, thus serving as a local source of proinflammatory cytokines. An association between increased epicardial fat and incident CHD and coronary atherosclerosis has been demonstrated in the general population[8,9].
A few studies have examined epicardial fat and CHD in the HIV-infected population and have generated conflicting findings. In a study of 110 participants, Lo et al.[10,11] found increased epicardial fat in HIV-infected participants compared with HIV-uninfected controls, but found no correlation between epicardial fat and coronary plaque volume, segments with plaque or with coronary calcium score measured by computed tomography (CT). Iacobellis et al. demonstrated an association between echocardiographic measures of epicardial fat thickness and carotid intima-media thickness in 103 HIV patients on HAART with the metabolic syndrome. More recently, Guaraldi et al.  conducted a larger study and confirmed the presence of greater epicardial fat depots in HIV-infected participants, and also demonstrated an association between epicardial fat and increased coronary artery calcium (CAC). The conflicting conclusions from these initial studies require further study.
To investigate the relationship between epicardial fat and HIV infection and the association between epicardial fat and subclinical coronary atherosclerosis and plaque composition, we measured CAC from noncontrast CT scans and coronary artery plaque extent and composition with contrast-enhanced coronary CT angiography (CCTA) in participants from the Multicenter AIDS Cohort Study (MACS), a large, prospective multi-ethnic cohort of both HIV-infected and HIV-uninfected MSM. CAC is known to be a potent predictor of coronary events in both symptomatic and asymptomatic populations [14,15]. CCTA provides more detailed characterization and measurement of plaque burden beyond calcium scoring and allows identification of plaque subtypes that may carry differential risks for adverse cardiovascular events [16-19]. We hypothesized that HIV-infected men have more epicardial fat than HIV-uninfected controls, and that epicardial fat is associated with subclinical coronary artery atherosclerosis.
The MACS is an ongoing prospective observational study that enrolled MSM in four major United States cities: Baltimore, Maryland/Washington, DC, Chicago, Illinois, Los Angeles, California, and Pittsburgh, Pennsylvania. Active MACS participants between the ages of 40 and 70, without a history of prior cardiac surgery, percutaneous transluminal coronary angioplasty or stent placement, and who weighed less than 300 pounds were invited to undergo noncontrast CT scanning. Coronary CT angiography was performed at the time of scanning on men without contrast allergy, atrial fibrillation, or impaired renal function (estimated glomerular filtration rate less than 60 ml/min per 1.73 m2 within 30 days of scanning or during any previous MACS examination). All participants gave informed consent for the participation. This study was approved by the Institutional Review Board of each institution.
Noncontrast CT and CCTA were performed using multidetector scanners at each site as previously described . Noncontrast cardiac CT scans and a single imaging slice of the abdomen at the level of the fourth lumbar vertebra were performed to measure CAC and adipose tissue volumes, including that of epicardial adipose tissue (EAT), pericardial adipose tissue (PAT), intrathoracic adipose tissue (IAT), and abdominal visceral fat (AVF).
CAC scans included a minimum of 40 slices, spaced 2.5-3.0 mm apart, starting from 1 cm below the carina. CAC scores were computed using the Agatston method . Adipose tissue measurements were performed on axial slices of noncontrast images of the heart using GE Advantage 4.6 Workstations (GE Healthcare, Waukesha, Wisconsin, USA). EAT was defined as the adipose tissue between the surface of the heart and visceral pericardium and was measured on axial images from 10 mm above the superior extent of the left main ostium to the last slice containing part of pericardial sac. EAT was measured by manually tracing the pericardium every two to three slices below the start point and software automatically tracing out the segments in between selected slices. PAT extended from outside the visceral pericardium to the chest wall. IAT measurement was performed using the same superior boundary, the diaphragm as the inferior boundary, the chest wall as the anterior boundary and the aorta, bronchi and esophagus defined the posterior boundary. Adipose tissue present in the posterior mediastinum and para-aortic adipose tissue were not included in the IAT measurements. Volume analysis software was used to discern fat from other tissues using a threshold of -190 to -30 Hounsfield Units (HUs). AVF was measured from an axial slice of the abdominal noncontrast CT scan images. The reader traced the parietal peritoneum border at the level of L4-L5 interspace and fat was defined as the intraperitoneal area of the abdominal cavity within -190 to -30 HUs. Adipose tissue measurements were performed by three well trained readers, blinded to participant characteristics. Interobserver and intraobserver agreement of EAT and IAT measurement was highly reproducible .
Eligible participants underwent CCTA using radiation dose reduction techniques. ß-Blocker therapy could be given prior to scanning to reduce heart rates and improve image quality, and sublingual nitroglycerin was given immediately before contrast injection unless contraindicated. The images were reviewed on a three-dimensional image analysis workstation (GE Advantage Workstation; GE Healthcare) at the central reading location (Torrance, California) by two experienced observers unaware of the participant's clinical history. Plaque grading was performed according to the American Heart Association's 15-segment coronary artery classification grading system . The segments were evaluated for the presence or absence of coronary plaque using axial and curved multiplanar reconstruction with one coronary plaque type assigned per segment. Plaques were defined as structures more than 1 mm2 within and/or adjacent to the vessel lumen that could be clearly distinguished from the lumen and surrounding pericardial tissue. Noncalcified plaque was defined as any structure associated with the coronary wall with CT attenuation below that of the lumen, but above that of the surrounding connective tissue and epicardial fat. Calcified plaque was defined as any structure greater than 130 HU visualized separately from the coronary lumen. Within each segment, plaques with calcified tissue comprising more than 50% of the plaque area were classified as calcified, plaques with some but less than 50% calcium were considered mixed, and plaques without any calcium were classified as noncalcified lesions. Plaque size was scored as none (0), mild (1), moderate (2), or severe (3). Semiquantitative measures of overall coronary artery plaque burden were derived. The total calcified plaque score (CPS), mixed plaque score (MPS), and noncalcified plaque score (NCPS) were the sum of the scores of all identified calcified, mixed, and noncalcified plaques, respectively. The total plaque score (TPS) was the sum of the NCPS, MPS, and CPS, and was the summary measure of overall plaque burden .
Metabolic, biochemical, and immunologic parameters
MACS participants underwent semi-annual follow-up with a history and physical examination. From these visits, demographic and clinical data from the visit closest to the scan date were obtained (generally within 6 months of the CT visit), including measures of HIV disease activity in HIV-infected participants, including plasma HIV RNA levels (peak and present), CD4+ T-cell count (present and nadir), history of AIDS, and duration of HAART use. Fasting blood samples were obtained to measure serum levels of glucose, insulin, total cholesterol, high-density lipoprotein (HDL) cholesterol, and triglycerides. Low-density lipoprotein (LDL) cholesterol was calculated using the Friedewald equation or was measured directly in persons with triglyceride levels greater than 400 mg/dl or a nonfasting sample. Creatinine was measured within 30 days of the CT scan with contrast for the calculation of estimated glomerular filtration rate using the Modification of Diet in Renal Disease equation .
Exploratory data analysis was performed to compare the distribution of potential confounders and mediators in HIV-infected and HIV-uninfected men using the Wilcoxon rank-sum test or χ2 test, where appropriate. Because of a high prevalence of zero scores for many plaque variables, the first analysis dichotomized the study population by presence or absence of plaque, with plaque presence defined as a score greater than zero. Separate multiple logistic regressions were used to evaluate the association between epicardial fat and plaque presence for CAC, TPS, CPS, MPS, and NCPS, adjusting for age, race, and HIV serostatus (minimally adjusted models). Nested multivariable analyses were performed with sequential additional adjustment for established cardiovascular risk factors (SBP, history of hypertension, diabetes mellitus, fasting glucose levels, triglyceride levels, use of lipid-lowering medications, and smoking).
The cardiovascular risk factors included in the multivariate model were selected from variables which when individually added to the minimal model had a P value ≤ 0.1 and when added to a multivariate model maintained a P value < 0.1. To assess whether any association between epicardial fat volume and plaque was independent of other measures of adiposity (BMI and/or abdominal visceral adipose tissue volume by CT scan), the multivariate models were further adjusted for BMI or AVF. HIV modification of each association was assessed by inclusion of appropriate interaction terms. Missing values were imputed five times based on the distribution of covariates using a Markov Chain Monte Carlo method assuming multivariate normality. Values for the following number of men were missing and imputed for multiple regression analyses: hypertension medications (three), BMI (17), diabetes medications (3), smoking pack-years (2), lipid medications (11), triglyceride (40), SBP (33), and fasting glucose (37).
Linear regression was used to assess the relationship of epicardial fat with plaque extent (burden) among individuals with plaque present (i.e., plaque score greater than zero) for each of CAC, TPS, CPS, MPS, and NCPS after adjusting for age, race, and HIV serostatus. Plaque scores were natural log-transformed. Multivariable linear regression was performed using serial models as described for plaque presence.
We used robust regression to analyze epicardial fat volumes and traditional CHD risk factors, and associations with HIV parameters. All statistical analyses were performed using SAS 9.2 (SAS Institute, Cary, North Carolina, USA). Statistical significance was established at a P value <0.05.
There were 932 men with complete noncontrast CT imaging who were included in this analysis. Seven hundred and six of these men (75.8%; 422 HIV-infected and 284 HIV-uninfected) underwent CCTA. The HIV-infected participants were younger, had a lower mean BMI than their seronegative counterparts, and were less likely to be white (all P <0.001) (Table 1). In unadjusted analyses, current tobacco use was more prevalent among HIV-infected individuals (P = 0.004), whereas HIV-uninfected men were more likely to be former tobacco users. Compared with HIV-uninfected men, HIV-infected participants were more likely to use lipid-lowering medications (36.3 vs. 29.7%, P = 0.04). HIV-infected participants had significantly higher fasting triglyceride levels and lower HDL (both P < 0.001) and LDL cholesterol levels (P = 0.002) compared with HIV-uninfected men. Fasting glucose levels and prevalence of diabetes were similar by serostatus, but HIV-infected men had higher fasting insulin levels (P < 0.001). Differences in age, serum creatinine, fasting glucose, diabetes, insulin, LDL-C, and hypertension medication use were apparent when comparing individuals who underwent CCTA to persons who did not.
Among HIV-infected men, 95.9% had initiated HAART before undergoing CCTA with a median duration of 12.5 (interquartile range, IQR 8.8-14.1) years (Table 1). A history of an AIDS illness was present in 13.8% of men. HIV RNA level was undetectable (<50 copies/ml) in 80.9% and the median HIV RNA among men with a detectable level was 565 (IQR 150-9070) copies/ml. The median CD4+ T-cell count was 599 (IQR 423-749) cells/μl and the median nadir CD4+ T-cell count was 243 (IQR 133-331) cells/μl.
Epicardial fat and associations with HIV serostatus, HAART, cardiovascular risk factors, and metabolic parameters
HIV-infected men had a significantly greater volume of EAT than HIV-uninfected men after adjusting for age and race [median values: 125 (IQR 93-131) vs. 115 (IQR 81-120) cm3, P = 0.001]. In separate multivariate models adjusted for age, race, and HIV serostatus, increased epicardial fat volume was associated with greater BMI, antihypertensive medication use, use of lipid-lowering medication, higher fasting glucose, higher insulin levels, and use of diabetes medications (all P < 0.001). Among men not taking lipid medications, higher triglyceride levels (P = 0.0003) and lower HDL cholesterol (P = 0.02) were associated with more epicardial fat after adjusting for age, race, and serostatus. There were no associations with cumulative tobacco use (P = 0.18) or total (P = 0.10) or LDL cholesterol (P = -0.35). Among men not taking hypertension medications, SBP (P = 0.004) and DBP (P = 0.006) were associated with more epicardial fat. When the significant covariates (P < 0.05) were included in a single multivariable model to predict epicardial fat, there were significant associations with BMI, use of lipid medications (both P < 0.001), and triglycerides (P = 0.04). There were borderline independent associations with cumulative tobacco use (P = 0.09) and no independent associations with use of diabetes or hypertension medications, fasting glucose or insulin, or SBP (all P > 0.10).
Among HIV-infected men, after adjustment for age, race, and cardiovascular risk factors, EAT volume was positively associated with the duration of HAART (estimate 1.15 per year, P = 0.02) (Table 2). The duration of use of azidothymidine therapy was associated with greater EAT volume after adjustment for age, race, and cardiovascular risk factors (estimate 0.94 per year, P = 0.048). In contrast, there were no associations between duration of use of either protease inhibitors, stavudine, or abacavir therapy.
Epicardial fat and plaque burden
Epicardial fat volume was moderately correlated with pericardial fat and intraabdominal visceral fat volumes (r = 0.67 and r = 0.67, respectively, P < 0.0001) and strongly correlated with total intrathoracic fat volume (r = 0.88, P < 0.0001). Generally, associations between epicardial fat and coronary plaque were similar, but somewhat stronger than with PAT or IAT, and therefore, results are only presented for EAT.
Coronary calcium was present in 53.0% of men. Among the 706 men who underwent CCTA, 77.2% had coronary plaque present (TPS > 0), with an unadjusted prevalence of 78.2% among HIV-infected and 75.7% among HIV-uninfected men. Noncalcified plaques were present in 59.9% overall. The HIV-infected men had a greater prevalence (P = 0.004) and extent (P = 0.001) of noncalcified plaque in unadjusted analyses (Table 3).Adjusted associations between EAT and the presence and extent of plaque (CAC, NCPS, CPS, MPS, and TPS) are presented in Tables 4 and 5. There was an increase in the odds of CAC [odds ratio (OR) 1.04 per 10 cm3 increase in EAT, P = 0.003], any plaque being present on CCTA (OR 1.09, P < 0.001), noncalcified plaque (OR 1.06, P < 0.001), and calcified plaque (OR 1.03, P = 0.037) with increasing EAT, even after adjustments for age, race, and HIV serostatus. These associations remained statistically significant after further adjustment for cardiovascular risk factors for any plaque on CCTA (P = 0.006) and noncalcified plaque (P = 0.001; as listed in the tables). Additional adjustment for BMI or AVF did not significantly attenuate these associations. There were no interactions by HIV serostatus for any of these associations (P > 0.40). Associations between increasing EAT and presence of plaque stratified by HIV serostatus are presented in Table 4A and 4B in the online supplementary material ( http://links.lww.com/QAD/A535).
In linear regression models restricted to men with any plaque present (i.e., plaque scores >0), after adjustment for age, race, and HIV serostatus, increasing EAT was associated with the extent of CAC (ß = coefficient = 0.048 log CAC score per 10 cm3 increase in EAT, P < 0.001), and extent of any plaque on CCTA (ß = 0.021, P = 0.001) (Table 5). There was also a trend with extent of noncalcified plaque (ß = 0.011, P = 0.06). After additional adjustment for cardiovascular risk factors, including BMI or AVF, increasing EAT remained significantly associated with extent of CAC. The association between extent of any plaque and EAT was significant after adjustments for age, HIV, and serostatus, but lost significance after including adjustments for cardiovascular risk factors (Table 5). There were no significant interactions by HIV serostatus for any of these associations (P > 0.10). The fully adjusted associations between EAT and extent of CAC were significant in HIV-infected men (ß = 0.067, P < 0.001) but not in HIV-uninfected men (ß = 0.009, P = 0.78), but the HIV interaction P value was not significant (P = 0.23). Associations between increasing EAT and extent of plaque stratified by HIV serostatus are presented in Table 5A and 5B in the online supplementary material ( http://links.lww.com/QAD/A535).