Impact of pericardial and paracardiac fat on the cardiovascular structure and function in HIV infected persons: a cardiovascular magnetic resonance Study

Abstract

Introduction: Cardiovascular disease (CVD) is the major contributor to morbidity and mortality in people living with human immunodeficiency virus (PLHIV). Approximately 8.2 million South Africans live with the human immunodeficiency (HIV) infection. The widescale availability and use of ART has dramatically decreased the number of HIV/acquired immunodeficiency syndrome (AIDS)-related deaths and improved quality of life. However, ART is associated with the development of CVD leading to heart failure with preserved ejection fraction, myocardial inflammation, myocardial fibrosis, and myocardial steatosis are common phenotypes in ARTtreated PLHIV. Increased visceral and pericardiac adipose tissue is associated with poor cardiovascular outcomes, including elevated risk of incident heart failure, atrial fibrillation, myocardial infarction, and sudden cardiac death. While the pathophysiology of HIV-associated CVD is complex and multifactorial, including varying evidence for direct viral effect of HIV, metabolic changes induced by antiretroviral therapy (ART), opportunistic infections, systemic inflammation and subclinical immune activation, mitochondrial dysfunction, and endothelial dysfunction, the role of pericardial and paracardiac adiposity to development of HIV-associated CVD is unknown. Cardiovascular magnetic resonance (CMR), through its high spatial and temporal resolution, is the ideal technique for quantification of pericardial adipose tissue (PAT) and paracardiac adipose tissue (ParaAT). The current study hypothesised that fat around the heart has detrimental effects on the myocardium in PLHIV on ART. Therefore, this study is aimed to investigate the effect of pericardial and paracardiac fat deposits on cardiovascular structure and function in PLHIV using CMR. Methods: This was a retrospective, cross-sectional study design including 3 groups of participants: (1) PLHIV on ART; (2) untreated PLHIV; and (3) matched HIV uninfected controls. CMR was performed on a Siemens Magnetom Skyra 3Tesla between February 2017 and March 2020. Images were postprocessed and analysed with proprietary software from circle Cardiovascular Imaging (CVI)42®. PAT and ParaAT volumes were obtained manually using Simpson’s method, by contouring the borders of PAT and the ParaAT with regions of interest at end-diastole, from base to the apex of the heart. Late gadolinium enhancement (LGE) images were acquired with a phasesensitive inversion-recovery sequence. PAT and ParaAT volumes from different short-axis slices were summed to obtain whole-heart PAT and ParaAT volumes. Left (LV) and right (RV) ventricular volumes, LV mass (LVM), and function, T2 short Tau inversion-recovery imaging, strain and strain rates, T1 and T2 mapping, LGE, and extracellular volume (ECV) were also assessed. Results: 198 participants were included: PLHIV on ART (n=131), PLHIV naïve to ART (n=19), and matched HIV uninfected controls (n=48). PLHIV on ART were older (44.9±8.3 years) than untreated PLHIV (36.4±8.2 years) and uninfected controls (41.7±11.5 years), respectively (p<0.001). A large distribution of our study cohort was female. PLHIV on ART were on antiretroviral combination therapy treatment including nucleoside reverse transcriptase inhibitors, non-nucleoside reverse transcriptase inhibitor and protease inhibitor. PLHIV on ART had greater whole-heart PAT volumes 43.1cm3 (30.7-54.8 IQR) compared to untreated PLHIV 32.1 cm3 (22.8- 52.8 IQR) or uninfected controls 22.99 cm3 (17.2-31.7 IQR), p<0.001. PLHIV on ART had higher whole-heart ParaAT volume 9.0 cm3 (3.8-13.4 IQR) compared to untreated PLHIV 0 cm3 (0.0-10.8 IQR) or uninfected controls 0 cm3 (0.0-5.2 IQR), p<0.001 (Figure 2). PLHIV on ART (100.9±28.5 g) and untreated PLHIV (103.2±29.7 g) had elevated LVM compared to uninfected controls (87.6 ± 19.3g), p<0.01 (Table 1). Untreated PLHIV had higher T1, T2 times and ECV (1301±58 ms; 42±4 ms; 32±5%) compared to PLHIV on ART (1251±47 ms; 39±3 ms; 31±5%) and untreated controls (1224±48 ms; 39±2 ms; 29±9%), respectively. Similarly, untreated PLHIV had the largest burden of LGE (Figure 3) and the greatest impairments in myocardial strain and strain rates. On univariate regression analysis in the pooled population, ParaAT demonstrated weak negative correlation with peak global circumferential strain (r=-0.23, p<0.001) and longitudinal strain (r=- 0.21, p<0.01), while PAT showed a weak negative correlation with peak circumferential strain (r=- 0.18, p<0.01) and moderate positive correlation with LVM (r=0.45, p<0.001) (Figure 4). Conclusion: The study suggests that pericardial and paracardial fat can drive cardiometabolic risk factors and directly promote cardiac dysfunction in HIV patients. CMR offers a non-invasive means of assessing ventricular function, anatomical structure, tissue characterisation and the presence of LGE in asymptomatic HIV-infected patients by assessing myocardial structure, inflammation, fibrosis, deformation, and weak association between the burden of pericardial and paracardiac fat, and tissue characteristics in PLHIV.