Impact of COVID-19 Infection on Regional Periocoronary Inflammation: An Angio-CT Study of Epicardial Fat Attenuation
Article Category: Original Article
Published Online: Jun 30, 2023
Page range: 47 - 53
DOI: https://doi.org/10.2478/rjc-2023-0011
Keywords
© 2023 Botond Mátyás et al., published by Sciendo
This work is licensed under the Creative Commons Attribution 4.0 International License.
It is currently well-known that inflammation is an essential key factor in all stages of atherosclerosis. While stable plaques exhibit a chronic inflammatory state, vulnerable or ruptured plaques exhibit signs of active inflammation. This can result in the thinning of the fibrous cap and can increase the risk of rupture, which is a major contributor to acute vascular events.[1]
Epicardial fat (EF) and pericoronary adipose tissue (PCAT) have been extensively studied in the last decade as markers and promoters of local coronary inflammation. [2–3] The PCAT is a layer of fatty tissue that encases the coronary arteries. It is made up of three primary constituents, including lipid-laden cells (adipocytes), stromal cells (such as preadipocytes), and interstitial tissue. PCAT is considered part of the vessel, and it has a closely connected anatomic and physiologic association with the arterial wall. [4] EF and PCAT also contribute to systemic inflammation by releasing cytokines into the circulation via paracrine signaling pathways.[5] Quantification of epicardial and pericoronary fat via several non-invasive imaging methods has been proven effective in predicting the risk for acute events across a plethora of cardiovascular disorders, and has also been shown to correlate with markers for systemic inflammation, increased oxidative stress, severity of atherosclerosis, and patient prognosis.[6–7]
One current challenge in research on coronary atherosclerosis is to identify markers that can reflect coronary inflammation, which is directly linked to plaque vulnerability. This would allow for the stratification of patients at high risk for acute cardiovascular events before clinical syndromes develop. In addition, targeting inflammation may be an effective strategy for preventing and treating this disease.[8]
One way to visualize inflammation in the PCAT using coronary computed tomographic angiography (CCTA) is through the use of the fat attenuation index (FAI). The FAI has been introduced by Caristo Diagnostics (Oxford, UK), representing a measure of the CT attenuation gradients in the perivascular space, which is calculated using the attenuation values across the PCAT When there is inflammation in the coronary region, the composition of PCAT changes to become more aqueous and less lipophilic, as indicated by the FAI. In addition, chronic inflammation can result in harmful fibrotic and other changes in the structure of PCAT. These alterations can be identified using sophisticated radiomic texture analysis techniques of the perivascular fat.[9–10] The FAI has several advantages as a measure of vascular inflammation: it is not influenced by the degree of coronary calcification, it is independent of the patient’s systemic inflammation as measured by hs-CRP, and it is not associated with the severity of coronary stenosis.[11–13]
Another indicator of cardiovascular risk, which is derived from FAI, is the FAI score, which is accompanied by vessel-specific nomograms for each coronary area, which offer a personalized evaluation of the level of inflammation.[12] CaRi-Heart score is a score derived by FAI calculation, which indicates the relative risk of a patient developing a major cardiovascular event, incorporating demographic factors such as age and sex, as well as anatomical considerations related to fat distribution.
Recent CT studies showed that atherosclerotic plaques found in the right coronary arteries (RCA) display a greater number of features that increase their susceptibility to destabilization compared to plaques found in left coronary arteries (LCA) This difference in vulnerability characteristics may be attributed to the unique geometric and hemodynamic features of the coronary circulation, as well as the impact of inflammation-related changes to endothelial shear stress caused by the release of inflammatory mediators from surrounding EF.[3]
COVID-19 has a significant impact on cardiovascular health, being associated with increased inflammation and immune system destabilization, both during the acute phase of the infection as well as over the clinical course and in long-COVID syndrome. COVID-19 can mediate systemic but also local coronary inflammation, which can accelerate the destabilization process of preexistent atherosclerotic plaques, thus triggering acute coronary events.[14-17] While the impact of COVID-19 on cardiovascular diseases has been extensively studied, little is known about the impact of SARS-CoV-2 infection on the regional inflammation at the level of coronary circulation.
This study aimed to evaluate the regional differences between local inflammation at the level of left versus right coronary circulation in patients who had CCTA examinations for chest pain in the early stages after COVID-19 infection.
We conducted an observational cross-sectional study, in the Center for Advanced Research in Multimodality Cardiac Imaging, Cardio Med Medical Center, Târgu Mureş, Romania, that included 172 patients with chest pain and low to intermediate clinical likelihood of coronary artery disease (CAD) who underwent 128-slice CCTA. The study population was divided into two main groups based on COVID-19 infection: Group 1 (n = 80) consisted of patients who had a COVID-19 infection a few months prior to their CCTA examination, and Group 2 (n = 92) consisted of patients adjusted for age and gender who did not have a COVID-19 infection prior to their CCTA examination. The demographic and laboratory characteristics, cardiovascular risk factors, and the development of signs and symptoms were all monitored and assessed for each patient before their CCTA examination.
The study employed multiple exclusion criteria, including high coronary calcification, irregular heart rate, inability to maintain heart rate below 65 bpm, morbid obesity, and difficulty complying with breath-hold instructions. Other exclusion criteria included the presence of chest pain due to causes other than CAD, a history of myocardial infarction, high CAD risk, severe symptoms, or acute coronary syndromes requiring percutaneous coronary intervention. Figure 1 provides a detailed PRISMA flowchart illustrating the inclusions and exclusions of the study.
Figure 1
PRISMA flowchart of patient’s recruitment
* Coronary artery calcium (CAC) > 1000; ** Poor quality coronary reconstructions and contained artifacts
The study was conducted in accordance with Good Clinical Practice guidelines and the Declaration of Helsinki and received approval from the ethics committee of the institution where it was conducted. All study participants provided their informed consent.
All subjects participating in the study underwent CCTA examination using a 128-slice Siemens Somatom Definition AS scanner (Siemens Healthcare, Germany) located at the Center for Advanced Research in Multimodality Cardiac Imaging, Cardio Med Medical Center, Târgu Mureş, Romania, to obtain images. The scanning was conducted with retrospective gating at a heart rate below 65 bpm, using a tube voltage of 120 kV, gantry rotation time of 0.33 seconds, and a collimation of 128 × 0.6. In cases where the resting heart rate was above 65 bpm, intravenous or oral beta blockers were administered, while blood pressure was monitored. The acquisition process began with a native scan to evaluate coronary calcium, succeeded by the administration of an iodine- based contrast solution of 80–100 ml, based on the patient’s body weight. A 50 ml saline chase was administered at a flow rate of 5.5–6 ml/s during inspiratory breath-hold. All the acquired scans were stored in a specific electronic imaging database to facilitate offline image post-processing and cloud delivery.
All the acquired images were set to DICOM format, then anonymized, transferred through the cloud and post-processed at the Centre of Caristo Diagnostics (Oxford, United Kingdom). The inclusion criteria for the analysis included the presence of plaques with a stenosis effect of at least 50%, and with at least one vulnerability marker (VM) present (spotty calcifications—SC, positive remodeling—PR, low attenuation plaque—LAP and the napkin ring sign—NRS). The FAI and AI- based FAI scores were determined for each of the major coronary arteries in all patients.
The calculation of the pericoronary adipose tissue FAI (PCAT- FAI) involved the use of AI-enhanced algorithms (CaRi-Heart®, Caristo Diagnostics, Oxford, UK) that offer precise and consistent measurements of attenuation in concentric 3D layers of perivascular tissue surrounding the human arterial wall.[18–20]
After quantification of PCAT-FAI for each coronary artery, all the collected data were sent back to our center and were stored in an electronic Microsoft Excel (Microsoft Corporation, Redmond, USA) database. The statistical analysis was performed using GraphPad Prism 9.5 software (GraphPad Software, Inc., San Diego, USA). The study included a PCAT-FAI analysis of 516 coronary arteries: 172 on the left anterior descending artery (LAD), 172 on the circumflex artery (LCX), and 172 on the right coronary artery (RCA). In addition, the CaRi-Heart® risk and the Duke score were also determined for each individual.[21]
The data was analyzed comparatively between patients with and without previous SARS-CoV-2 infection. Nominal (categorical) variables were reported as integer values (percentages) and compared between groups using the Chi-square test (χ2) and its variables. Numeric data were expressed as mean ± standard deviation and the Mann Whitney or unpaired Student
This study included 172 patients with chest pain and low to intermediate clinical likelihood of CAD, 46.51% (n = 80) with COVID-19 infection a few months prior to the CCTA examination, and 53.49% (n = 92) without COVID-19 infection prior to the CCTA examination. The mean age was 62.43 ± 11.62 years, and 66.86% (n = 115) of the patients were males. There were no statistically significant differences between the two study groups regarding gender; age; or comorbidities such as hypertension, diabetes, and obesity. However, there was a significantly higher rate of hypercholesterolemia in Group 1 compared to Group 2 (58.69% vs. 40.00%,
Baseline characteristics, comorbidities, and risk factors in the study population
| Parameters | Whole study sample (n =172) | Group 1 (COVID-19) (n = 80) | Group 2 (non-COVID-19) (n = 92) | |
|---|---|---|---|---|
| Male gender, n (%) | 115 (66.86%) | 49 (61.25%) | 66 (71.73%) | NS |
| Age at time of scan, mean ± SD | 62.43 ± 11.62 | 60.19 ± 10.10 | 62.84 ± 9.90 | NS |
| Smoking, n (%) | 31 (18.02%) | 10 (12.50%) | 21 (22.82%) | NS |
| Hypertension, n (%) | 146 (84.88%) | 65 (81.25%) | 82 (89.13%) | NS |
| Hypercholesterolemia, n (%) | 86 (50.00%) | 32 (40.00%) | 54 (58.69%) | 0.02 |
| Diabetes, n (%) | 47 (27.32%) | 19 (23.75%) | 28 (30.43%) | NS |
| Obesity, n (%) | 44 (25.58%) | 26 (32.50%) | 19 (20.65%) | 0.06 |
| BMI, mean ± SD | 27.57 ± 4.29 | 28.51 ± 4.21 | 26.93 ± 4.25 | 0.03 |
| PCI after CCTA, n (%) | 75 (43.60%) | 30 (37.50%) | 47 (51.08%) | NS |
| Multi-vessel PCI, n (%) | 25 (14.53%) | 8 (26.66%) | 18 (38.29%) | NS |
| Heart failure, n (%) | 127 (72.67%) | 61 (76.25%) | 70 (76.08%) | NS |
| LVEF (%), mean ± SD | 47.69 ± 5.07 | 48.34 ± 4.18 | 47.12 ± 5.71 | NS |
| Creatinine (mg/dl), mean ± SD | 0.97 ± 0.26 | 0.93 ± 0.23 | 1.00 ± 0.27 | NS |
| Total cholesterol (mg/dl), mean ± SD | 167.3 ± 47.13 | 161.4 ± 43.55 | 171.0 ± 49.24 | 0.07 |
| Triglycerides (mg/dl), mean ± SD | 145.7 ± 69.49 | 134.1 ± 75.04 | 154.0 ± 64.58 | 0.03 |
| COVID-19 vaccine, n (%) | 107 (62.20%) | 46 (57.50%) | 56 (60.86%) | NS |
| Time from COVID-19 to CCTA (days), mean ± SD | 134.2 ± 104.9 | |||
Figure 2 illustrates CCTA images of the three major coronary arteries in a patient who had a COVID-19 infection a few months prior to CCTA examination. Panel A shows a stable atherosclerotic lesion (green arrows); Panel B shows pericoronary fat with the FAI colored mapping around the non-culprit lesion (single green arrow) demonstrating abnormal FAI in the same patient.
Figure 2
Conventional CCTA image of the three major coronary arteries (A), and a colored mapping of an abnormal FAI for the same patient (B)
According to the standard adipose tissue Hounsfield unit range of -190 to -30 HU, the conventional CT attenuation index did not reveal any significant differences in the coronary arteries between the two groups. The FAI score was consistently higher in the non- COVID-19 group, more precisely: FAI score LAD (Group 1: 9.12 ± 6.20 vs. Group 2: 11.87 ± 8.23, p = 0.05), FAI score LCX (Group 1: 10.77 ± 6.13 vs. Group 2: 13.02 ± 6.76, p = 0.05), FAI score LCA (Group 1: 9.78 ± 5.95 vs. Group 2: 12.11 ± 7.21,
AI-based PCAT - FAI Score
| Parameters | Whole study sample (n = 172) | Group 1 (COVID-19) (n = 80) | Group 2 (non-COVID-19) (n = 92) | |
|---|---|---|---|---|
| FAI HU LAD, mean ± SD | −76.08 ± 7.66 | −75.07 ± 7.59 | −76.46 ± 7.74 | NS |
| FAI HU LCX, mean ± SD | −71.32 ± 7.50 | −71.44 ± 7.88 | −71.21 ± 7.16 | NS |
| FAI HU RCA, mean ± SD | −73.11 ± 8.94 | −72.97 ± 9.38 | −73.23 ± 9.61 | NS |
| FAI Score LAD, mean ± SD | 10.39 ± 7.01 | 9.12 ± 6.20 | 11.87 ± 8.23 | 0.05 |
| FAI Score LCX, mean ± SD | 11.24 ± 6.28 | 10.77 ± 6.13 | 13.02 ± 6.76 | 0.05 |
| FAI Score RCA, mean ± SD | 15.10 ± 11.94 | 14.74 ± 12.24 | 15.88 ± 10.36 | NS |
| FAI Score LCA, mean ±SD | 11.02 ± 6.48 | 9.78 ± 5.95 | 12.11 ± 7.21 | 0.005 |
| FAI Score TOTAL, mean ± SD | 11.90 ± 7.93 | 10.55 ± 7.32 | 13.01 ± 8.16 | 0.001 |
When comparing the FAI score between the left and right coronary artery, we found that for the entire study population, the FAI score was significantly higher at the RCA level. This difference was also maintained in the COVID-19 positive group. However, in the COVID-19 negative group, this difference, although similar, did not reach the threshold of statistical significance (Figure 3).
Figure 3
Analysis of FAI score between the RCA and LCA
In both cases, the CaRi Heart® Risk and the Duke Score had significantly higher values for the patients in the COVID-negative group, as shown in Table 3.
CaRI Heart® Coronary Inflammation
| Parameters | Whole study sample (n = 158) | Group 1 (COVID-19) (n = 75) | Group 2 (non-COVID-19) (n = 83) | |
|---|---|---|---|---|
| CaRi Heart® Risk, mean ± SD | 18.34 ± 16.02 | 11.34 ± 8.47 | 23.81 ± 18.31 | < 0.0001 |
| Duke Score, mean ± SD | 2.97 ± 1.36 | 2.54 ± 1.33 | 3.37 ± 1.27 | < 0.0001 |
The coronavirus pandemic has had far-reaching effects on our society, and one of the areas in which this impact was drastic was on cardiovascular health. While this virus was initially known to affect mainly the lungs, it has been found to have a profound effect on the heart as well.[22] The increased prevalence of COVID-19 has presented healthcare professionals with a multitude of challenges, including the examination of PCAT inflammation for cardiovascular involvement. The respiratory syndrome caused by COVID-19 can lead to severe hypoxemia, which can result in multi-organ failure and cardiac injury.[23] SARS-COV-2 increases inflammation by stimulating the release of cytokines and chemokines from respiratory epithelial cells, dendritic cells, and macrophages.[24] It also promotes the accumulation of EF and PCAT, which can cause local vascular inflammation and endothelial dysfunction, leading to plaque formation and deterioration.[25]
SARS-COV-2 infection has the potential to exacerbate underlying cardiovascular diseases, or to increase vulnerability of coronary atherosclerotic plaques due to elevated levels of pro-inflammatory cytokines and chemokines in circulation. These substances can lead to microvascular and vascular thrombosis, coronary vasospasm, shear stress modulation, and platelet activation, which can further contribute to plaque vulnerability.[16] The severity of the cytokine storm, or excessive release of cytokines, is a significant predictor of the clinical progression of extrapulmonary organ failure and mortality in COVID-19 patients.[17]
CCTA has established itself as the first-line test for investigating suspected CAD. However, with the increasing use and adoption of CCTA in various clinical settings, it is important to enhance its sensitivity and specificity, as well as to refine its diagnostic and prognostic value for early detection of coronary atherosclerosis.[26-27]
There are several studies that have examined how artificial intelligence (AI) based algorithms can improve risk stratification based on standard clinical parameters. For instance, Motwani et al. examined the ability of an AI-based algorithm to predict prognosis in a large group of 10,030 patients followed for five years, with mortality as the endpoint. The AI algorithm was more accurate than the Framingham Risk Score (FRS) or CCTA severity risk scores. [28] Van Rosendael et al. used data from the CONFIRM registry to develop a model for risk stratification. The primary endpoint in their analysis was a composite of myocardial infarction and death, and they found that the AI algorithm was able to predict this endpoint with great precision.[29]
Our study demonstrated several hypotheses that have not been addressed so far in the literature. First, we showed that the FAI score, indicating the general severity of cardiovascular diseases, is consistently higher in the non-COVID-19 group. However, applying this score separately for each coronary artery is a more reliable way of measuring coronary inflammation and a more valuable predictor of the risk of cardiac mortality than the previous FAI methods.[9,30]
Second, when we compared the FAI scores of the left and the right coronary arteries, we noticed that for the entire study population, the FAI score was significantly higher at the RCA level, but this difference was particularly pronounced in the COVID-19 positive group. However, in the COVID-19 negative group, this difference, although similar, did not reach the threshold of statistical significance. This may be related to different hematological characteristics of the right coronary flow, as the right coronary artery has larger and fewer branches than the left coronary artery. Our study indicates that due to the hemorheologic characteristics of the right coronary flow, this artery is more prone to plaque vulnerability in post-COVID-19 patients.
Third, our analysis revealed that both the CaRi Heart® Risk, a FAI-derived score which estimates the likelihood of an individual experiencing a fatal cardiac event within eight years,[31] as well as the Duke Score, were significantly higher in the COVID-19 negative patients.
We believe that the findings of this study provide support for using the pericoronary FAI score as a dependable measure of coronary immune-inflammatory activation and its close association with plaque vulnerability. The ability of the FAI score to monitor pericoronary inflammation could have important implications in assessing vascular involvement in conditions beyond those commonly encountered by cardiologists, such as the COVID-19 pandemic.[21] More research is needed to investigate the effect of plaque location on hemodynamic characteristics, as well as the influence of the specific phenotype of vulnerable plaque location on the degree of pericoronary adipose tissue inflammation.
CaRi-Heart® is a novel tool that combines the FAI mapping technique with traditional cardiovascular risk factors and a comprehensive analysis of CCTA coronary plaque. It has been shown to provide significant clinical benefits in two large, independent CCTA-based studies, beyond what traditional risk factors can predict. Incorporating CaRi-Heart® analysis into current clinical practice has the potential to improve the usefulness of CCTA in assessing and managing the risk of CAD, helping clinicians to provide personalized treatment decisions for a patient’s future cardiac care.[32-33]
There are a few limitations to our study. The patients were recruited from a single center and the study did not include a follow-up, nor the quantification of serum inflammatory biomarkers. Secondly, the study excluded lesions with less than 50% stenosis, so it is unclear if pericoronary FAI can detect cases with a high risk of events before significant stenosis. Thirdly, the study did not include patients with non-ST elevation acute coronary syndrome, a population that may present with an increased inflammatory burden, and the 2020 ESC guidelines suggest CCTA as an alternative to invasive angiography to exclude ACS in cases with a low to intermediate likelihood of CAD. Fourthly, it is worth noting that there is a possibility of undiagnosed COVID-19 cases in the study population, as asymptomatic or mild cases may not have been recorded in the medical records. Further research is needed to confirm the clinical value of FAI score in the treatment of coronary inflammation.
The results of the study suggest that COVID-19 infection is associated with a higher risk of inflammation in the pericoronary epicardial fat. In post-COVID-19 patients, this inflammation seems to be more pronounced at the level of the RCA, which indicates a potential role of local hemorheologic factors in the complex process of inflammation-mediated plaque vulnerabilization after COVID.