| |
Coronary atherosclerosis characteristics in HIV-infected patients on long-term antiretroviral therapy
insights from coronary computed tomography-angiography
|
| |
| |
Senoner, Thomasa; Barbieri, Fabiana; Adukauskaite, Agnea; Sarcletti, Mariob; Plank, Fabiana; Beyer, Christophc; Dichtl, Wolfganga; Feuchtner, Gudrun M.c
AIDS: Oct 1 2019
Download the PDF here
Conclusion
HIV-positive individuals on long-term ART display higher CAD burden and more high-risk plaques.
Furthermore, HIV-infected individuals displayed a higher stenosis severity (CAD-RADS) and more ectatic coronary arteries compared with the control group. Whether targeted therapy reduces cardiovascular events in this at-risk population remains to be evaluated.
coronary artery disease (CAD) poses a particular threat to individuals with HIV infection in terms of both survival and morbidity, with an 1.5-2-fold increased risk of cardiovascular events [2,3]. The underlying pathomechanism of CAD in HIV-positive individuals are complex and not fully understood. Apart from the high prevalence of cardiovascular risk factors in HIV-infected patients, HIV-accelerated inflammation also seems to be a promotor of CAD [4,5]. Furthermore, the long-term effect on coronary artery disease of metabolic changes induced by some antiretrovirals has not been fully elucidated. The mean duration of HIV infection was 17.8 ± 9.4 years (minimum 0.19 to maximum 33 years) and patients were on ART for a mean of 13.0 ± 7.3 years (minimum 0 - maximum 24.27 years). The majority of patients were on ART (67 of 69) at the time of CCTA, except for two, in which ART was initiated 1 and 1.3 years after CCTA, respectively.
"CAD prevalence (any plaque) and more than 50% stenosis were both more prevalent in the HIV-positive cohort compared with the control group (P = 0.043 and P = 0.001, respectively). The SSS (severity involvement score) was significantly higher in HIV-positive compared with HIV-negative individuals (P = 0.038). Mixed noncalcified plaque burden (G-score) as well as CCS were significantly higher in the HIV-positive cohort compared with the control group (P = 0.003; P = 0.015, respectively). HIV-positives showed significantly more HRP (number of patients with at least one HRP) compared with the control group (23 vs. 8, P = 0.003) and the total number of HRP was higher (31 vs. 10; P < 0.001). There was a trend towards higher total plaque burden (SIS) in HIV-infected individuals, although this did not reach statistical significance (P = 0.06; Table 2)......Figure 1 shows a patient with long-standing HIV infection (31 years) and who was started on ART about 10 years after diagnosis. This patient has a high plaque burden, more than 70% stenosis in the LAD, as well as a HRP (low-attenuation fibrofatty plaque (52 HU)). Furthermore, the patient's coronary arteries were massively ectatic."
ABSTRACT
Objective: The aim of the study was to assess coronary artery disease (CAD) characteristics by coronary computed tomography-angiography (CCTA) in individuals with HIV infection on long-term antiretroviral therapy (ART)
Design: Retrospective case-controlled matched cohort study.
Methods: Sixty-nine HIV-positive patients who underwent 128-slice dual source CCTA (mean age 54.9 years, 26.1% women) with mean 17.8 ± 9.4 years of HIV infection and a mean duration on ART of 13 ± 7.3 years were propensity score-matched (1 : 1) for age, sex, BMI, and five cardiovascular risk factors with 69 controls. CCTA was evaluated for stenosis severity [according to Coronary Artery Disease - Reporting and Data System (CAD-RADS)], total plaque burden [segment involvement score (SIS) and mixed-noncalcified plaque burden (G-score)]. As inflammatory biomarkers, high-risk plaque (HRP) features (napkin-ring sign, low-attenuation plaque, spotty calcification, positive remodeling), perivascular fat attenuation index (FAI), and ectatic coronary arteries were assessed.
Results: CAD-RADS was higher in HIV-positive participants as compared with controls (2.21 ± 1.4 vs. 1.69 ± 1.5, P = 0.031). A higher prevalence of CAD and G-score (P = 0.043 and P = 0.003) was found. HRP prevalence [23 (34.3%) vs. 8 (12.1%); P = 0.002] and the number of HRP (36 vs. 10, P < 0.001) were higher in HIV-positive individuals. A perivascular FAI greater than -70 Hounsfield units was present in 27.8% of HRP. Ectatic coronary arteries were found in 10 (14.5%) HIV-positive persons vs. 0% in controls (P = 0.003).
Conclusion: Noncalcified and HRP burden in HIV-infected individuals on long-term ART is higher and associated with higher cardiovascular risk. Moreover, HIV-positive individuals displayed a higher stenosis severity (CAD-RADS) and more ectatic coronary arteries compared with the control group.
Introduction
In 2018, roughly 37 million people live with HIV infection worldwide, with the highest prevalence in countries in sub-Saharan Africa. Worldwide, 21.7 million (59%) are receiving antiretroviral treatment (ART) [1], which has translated a life-threatening virus infection to a chronic disease. Patients who have access and adhere to ART have an excellent virological and immunological response and most of them have a life expectancy similar to HIV-negative individuals.
However, coronary artery disease (CAD) poses a particular threat to individuals with HIV infection in terms of both survival and morbidity, with an 1.5-2-fold increased risk of cardiovascular events [2,3]. The underlying pathomechanism of CAD in HIV-positive individuals are complex and not fully understood.
Apart from the high prevalence of cardiovascular risk factors in HIV-infected patients, HIV-accelerated inflammation also seems to be a promotor of CAD [4,5]. Furthermore, the long-term effect on coronary artery disease of metabolic changes induced by some antiretrovirals has not been fully elucidated.
Moreover, studies using computed tomography-angiography (CCTA) for assessment of coronary vessel wall changes in HIV-positive individuals are sparse and contradicting [4,6].
CCTA allows for the differentiation of coronary vessel wall changes indicating atherosclerosis or inflammation. Imaging biomarkers of inflammation represent high-risk plaque (HRP) features and the perivascular fat attenuation index (FAI) [7]. High-risk plaques are defined by CCTA by applying the four major criteria: low-attenuation (lipid-rich) plaque, napkin-ring sing, spotty calcification and positive remodeling. Furthermore, focal ectatic coronary arteries are a sign of increased systemic inflammation [8].
Moreover, CCTA permits quantification of coronary stenosis severity and total plaque burden [9,10], and the differentiation of noncalcified fibroatheroma from calcified plaque.
Thus, the aim of our study was to assess the coronary artery disease profile and signs of plaque inflammation (defined by HRP markers, perivascular FAI, and coronary ectasia) in patients with long-term HIV infection on ART in a retrospective case-controlled matched cohort study.
Discussion
In our study, HIV-positive individuals have higher CAD prevalence, stenosis severity and noncalcified (fibro-fatty) mixed plaque burden. Additionally, an increased number of HRPs was found. High-risk plaque on CCTA have shown a correlation with indirect signs of plaque inflammation, such as thin-cap fibroatheroma with macrophage infiltration by optical coherence tomography [27].
Furthermore, HIV-positive participants showed more indirect signs of perivascular inflammation, such as ectatic coronary segments. Coronary ectasia is an established imaging feature of vasculitis and indicates a chronic inflammatory process. Moreover, we found a high prevalence of a positive perivascular FAI and a declining perivascular fat attenuation, suggesting perivascular edema and inflammation, which was present in one-third of high-risk lesions.
One landmark multicentric trial (CRISP-CT) has recently shown that the perivascular FAI captures inflammation-induced changes in perivascular fat attenuation and also identifies vulnerable plaques, and thus represents a noninvasive biomarker of coronary inflammation measured by traditional CCTA. An FAI greater than -70 HU resulted in a 5-9-fold increased risk of cardiac death in a cohort of 1872 patients [7,28].
The CCS, another predictor of cardiovascular mortality [29], was also measured in our study. In contrast to our study, mean calcium scores were not higher in HIV+ patients in a meta-analysis including 9000 patients [30]. Reasons might be the longer duration of HIV infection (17.3 vs. 9.3 years) and the longer duration on ART (13 vs. 2.3 years) reported in our study, along with good adherence to ART. Hence, a longer exposure to HIV infection might translate into a higher state of inflammation leading to an increase in coronary calcium [31,32]. Additionally, the higher prevalence of statin medication in the HIV cohort might as well influence the higher calcium score in our cohort. The PARADIGM study has shown that statins reduce especially fibro-fatty and necrotic plaque burden by CCTA, while calcified plaque burden increases [33].
However, the limitation of a calcium score alone is the lack of information regarding noncalcified plaques, which may be even more relevant for risk assessment [34], but can only be detected by CCTA.
Studies have shown a higher prevalence of subclinical coronary atherosclerosis and a greater burden of coronary atherosclerotic plaque, particularly noncalcified inflammatory plaques, in HIV-positive men compared with HIV-negative individuals with similar cardiovascular risk factors. Imaging studies using CCTA have shown a higher prevalence of coronary atherosclerosis in HIV-positive men compared with HIV-negative controls (59.0 vs. 34.4%). Similarly, HIV-positive women had a significantly higher prevalence of noncalcified plaques (74 vs. 23%) compared with HIV-negative female controls [35]. There is a strong association between the presence of HRPs and increased immune activation, with plaques that are prone to rupture being composed of a necrotic core with an overlying thin fibrous cap. Vulnerable plaque features (low attenuation, positive remodeling and spotty calcification) are more prevalent in HIV-positive individuals compared with HIV-negative controls [35].
Furthermore, noncalcified coronary artery plaques have been associated with a reduced CD4+ cell count in HIV-positive patients, supporting the notion of systemic inflammatory dysregulation in HIV-positive individuals contributing to CVD. Consequently, systemic inflammation and immune activation in HIV infection contributes to the accelerated atherogenesis seen in HIV-positive individuals [35].
Data using CCTA for assessment of coronary vessel wall changes in HIV-positive individuals are limited and contradicting [4-6].
The Swiss cohort study [4] is the largest series in Europe evaluating coronary artery disease in 428 HIV-positive and 276 HIV-negative individuals, and in both groups, a similar percentage of noncalcified/mixed plaques and HRPs was found, whereas HIV-positive participants had less calcified coronary plaques. This differs from results of the Multicenter AIDS Cohort Study (MACS) evaluating 618 HIV-positive and 383 HIV-negative individuals, which noted a higher prevalence of any plaque and especially of noncalcified plaques in HIV-positive compared with HIV-negative US males, mainly homosexuals [5]. Noncalcified plaques have been shown to correlate with worse long-term clinical outcomes compared with calcified plaques, independent of cardiovascular risk factors and number of diseased coronary arteries [36].
A previous small sample size study on 41 individuals alluded to the inflammatory theory of atherosclerosis by using 18F-FDG-PET and CCTA [6]; however, this has not yet been confirmed in a larger cohort.
Our study results differ from the results of the Swiss HIV cohort study [4]. In our study, HIV-positive participants displayed more HRP, a higher SSS and CCS compared with HIV-negative individuals. In accordance with the Swiss HIV cohort study [4], we found no correlation between ART duration and G-score, duration of HIV infection and G-score, or CD4+ cells and G-score, although a trend could be observed between the duration of HIV infection and higher G-score. The G-score is a measure of a dominant mixed noncalcified plaque burden, weighting fibrofatty atheroma against calcified plaques.
Importantly, total and especially noncalcified plaque burden has been linked with both adverse outcomes (MACE events) [37] and ischemia [38,39]. Even in the absence of severe obstructive CAD (stenosis >70%), a high fibroatheroma and lipid-rich plaque burden is associated with ischemia [40] [INOCA (ischemia and no obstructive coronary artery disease)]. INOCA also explains atypical chest pain complaints and similarly to high-risk plaque, those patients benefit from more intense medical therapy, such as high intense statins or PCSK 9 inhibitors [41,42].
Strengths of our study include the well matched cohorts, which did not differ regarding risk factors as well as age and sex. Furthermore, HIV-positive individuals had a long history of HIV infection (mean 17.8 ± 9.4 years) and were on ART for a mean of 13 ± 7.3 years. Furthermore, only three patients out of 138 (2.17%) were investigated with a 64-slice CT scanner and the rest were all examined with a 128-slice dual source CT scanner (Flash Somatom Siemens). Therefore, the scanner type can be excluded as a source that could potentially influence the results.
Finally, there was no difference in HRP prevalence in those with shorter and longer HIV infection and ART, suggesting that HRPs develop at any stage of HIV infection, though the number of individuals with short-term ART and HIV was low.
Study limitations: We acknowledge the adherent bias related to the retrospective study design. Influencing factors such as the coronary risk profile were minimized by propensity score matching. Only information about statin therapy but not any other medication was included. Furthermore, the inclusion period between HIV-positive individuals and controls differed, with HIV-positive individuals undergoing a CT scan about 1 year prior to controls.
|
|
| |
| |
|
|
|
