Impact of the “atherosclerotic pabulum” on in‐hospital mortality for SARS‐CoV‐2 infection. Is calcium score able to identify at‐risk patients?

Abstract Background Although the primary cause of death in COVID‐19 infection is respiratory failure, there is evidence that cardiac manifestations may contribute to overall mortality and can even be the primary cause of death. More importantly, it is recognized that COVID‐19 is associated with a high incidence of thrombotic complications. Hypothesis Evaluate if the coronary artery calcium (CAC) score was useful to predict in‐hospital (in‐H) mortality in patients with COVID‐19. Secondary end‐points were needed for mechanical ventilation and intensive care unit admission. Methods Two‐hundred eighty‐four patients (63, 25 years, 67% male) with proven severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) infection who had a noncontrast chest computed tomography were analyzed for CAC score. Clinical and radiological data were retrieved. Results Patients with CAC had a higher inflammatory burden at admission (d‐dimer, p = .002; C‐reactive protein, p = .002; procalcitonin, p = .016) and a higher high‐sensitive cardiac troponin I (HScTnI, p = <.001) at admission and at peak. While there was no association with presence of lung consolidation and ground‐glass opacities, patients with CAC had higher incidence of bilateral infiltration (p = .043) and higher in‐H mortality (p = .048). On the other side, peak HScTnI >200 ng/dl was a better determinant of all outcomes in both univariate (p = <.001) and multivariate analysis (p = <.001). Conclusion The main finding of our research is that CAC was positively related to in‐H mortality, but it did not completely identify all the population at risk of events in the setting of COVID‐19 patients. This raises the possibility that other factors, including the presence of soft, unstable plaques, may have a role in adverse outcomes in SARS‐CoV‐2 infection.


and it was declared a pandemic by World Health
Organization on March 11, 2020. Although the primary cause of death in COVID-19 infection is respiratory failure, there are evidence that cardiac manifestations may contribute to overall mortality and can even be the primary cause of death. 2 More importantly, it is recognized that COVID-19 is associated with a high incidence of thrombotic complications 1 and that the thrombotic diathesis is due to endothelial cell dysfunction. 3 Of note, while there is a strong evidence that known risk factors for coronary artery disease (CAD), such as age, hypertension, and diabetes, are associated with a poorer prognosis, 2-6 it has been shown that patients with reduced ventricular function do not have increased mortality compared to controls. 7 In this context, the coronary artery calcium score (CAC score), an established and validated prognostic indicator of CAD, has been of utmost importance in recognizing patients at high risk of poor outcome. 8,9 Indeed, there are increasing evidence that plaque characteristics are important in defining accurate cardiovascular risk beyond calcifications. 10 Therefore, our hypothesis was to verify if CAC per se is able to identify patients at risk of adverse outcomes and in-hospital (in-H) death in patients with SARS-CoV-2.

| Study population
We conducted a retrospective, post hoc analysis of all patients admitted to Padua University Hospital with a confirmed COVID-19 diagnosis by polymerase chain reaction (PCR) from January 2020 to January 2021. Sample for real-time PCR was obtained by nasal-oral pharyngeal swab. Exclusion criteria were a history of previous percutaneous coronary artery stenting or coronary bypass surgery, as it may interfere with CAC score calculation. We included patients with known previous CAD who were under medical treatment.
Our population consisted of 284 patients who underwent chest computed tomography (CT) scans because of moderate or severe COVID-19 infection, according to World Health Organization guidelines. 11 Baseline demographic, clinical, and laboratory variables (including inflammatory biomarkers) were retrieved from our electronic medical record system. High-sensitivity cardiac troponin I (HScTnI, cutoff value <16 ng/L) was considered suggestive of acute myocardial damage when its value was at least one above the 99th percentile of the upper reference limit. 12

| CT scan protocol
All CT scans were performed with a 64-slice CT system (Aquilion 64; Toshiba) and slice CT system (SOMATOM Sensation; Siemens). A Spiral non-electrocardiogram (ECG) gated technique during a deep inspiratory breath-hold was employed (tube voltage 120 kV, tube current power 50-200 mAs,). Images were reconstructed with the following parameters: slice thickness 3 mm, the field of view 250-300 mm, convolution kernel filtering b30f. CAC score was performed on the workstation (Vitrea FX, version 1.0; Vital Images), using CAC score analysis software (VScore; Vital Images). Coronary calcium was defined as an area of at least three contiguous voxels in the axial plane in the course of the coronary artery, with an attenuation cutoff of ≥100 HU.
Coronary calcium was defined as an area of at least three contiguous voxels in the axial plane in the course of the coronary artery, with an attenuation cut-off of ≥100 HU (corresponding to a minimum lesion area >1 mm 2 ) in the 3.0 mm reconstruction. 8 Although the traditional Agatston method for measuring CAC requires ECG-gated acquisition, a good correlation has been demonstrated between CAC identified on non-gated CT scans and ordinal scores obtained from gated CT scans. 13 Patient with Calcium were further stratified according to validated CAC score thresholds (1-100: mild; 101-400: moderate; >400: severe) 8 and to the cutoff point of 10 (Table 1). 14 We evaluated the occurrence of complications including acute coronary syndrome (ACS), embolic events (cerebral or peripheral), pulmonary embolism, myocarditis, pericarditis, acute heart failure, septic shock, severe acute respiratory distress syndrome, acute kidney injury, and deep vein thrombosis. The primary endpoint was in-H mortality. The secondary endpoint was need for admission to the intensive care unit (ICU) and mechanical ventilation.

| Statistical analysis
Descriptive statistics were reported as I quartile/median/III quartile for continuous data and percentages (absolute numbers) for categorical data.
Univariable and multivariable generalized linear models were estimated to assess the effect of baseline variables on the outcomes of interest using the Aranda link function, which was chosen because it was the parametrization that minimized the Bayesian information criterion. 15 Multivariable model variable selection was made according to the Akaike information criterion. 16 The marginal effect was computed considering the partial derivatives of the marginal expectation. Results were reported as average marginal effect (AME), 95% confidence interval, and p-value.
The AME expresses the change in probability of the event, that is, ICU admission, in-H mortality, mechanical ventilation.
Analyses were performed with R system 17 within rms package. 18

| RESULTS
Two-hundred-eighty-four patients were analysed.
Demographic, clinical, and laboratory features stratified by CAC status are presented in Table 2. Ordinal CAC score was calculated in 284 patients, 46 patients having mild (1-100), 39 moderate (101-400), and 57 severe (>400) CAC scores. However, we used only dichotomic values for statistical analysis (CAC = 0 was present in 142 patients, CAC ≥ 1 was present in 142 patients) as we did not note any increase in the outcomes or in cardiovascular complications with increased CAC values.
As expected, factors associated with CAC were male sex, age, hypertension, diabetes, smoke, and previous CAD. Of note patients with CAC had a higher inflammatory burden at admission (D-dimer, CRP, and procalcitonin) and higher HScTnI at admission and at peak.
While there was no association with the presence of lung consolidations, patients with CAC had a higher incidence of bilateral pulmonary involvement and a trend towards worse GGO.
In-H mortality was associated with CAC. Nevertheless, it did not increase for each point increment in CAC. As expected, in-H mortality was associated with age but also with hypertension, hyperlipidaemia, obesity, and previous CAD. It was indeed related to lung consolidations and with a higher inflammatory response (Table 3A-C). Of note, peak HScTnI >200 ng/dl was positively associated with in-H mortality both at univariable and multivariable analysis.
CAC was not associated with the need of ICU admission and mechanical ventilation (Table 3A-C), whereas it appears that HScTnI >200 ng/L was associated with both.
Older age, hypertension, hyperlipidemia, and smoking were positively associated with in-H mortality, need for ICU, and mechanical ventilation, also when considered as composite outcomes. The same increasing trend across the groups was observed for laboratory data at admission (CRP and HScTnI peak). In particular, CRP and HScTnI >200 ng/L remained positively associated with the composite outcome also in the multivariable model (Table 4).

| DISCUSSION
Data from multiple cohorts shows that CAC effectively stratifies patients for long-term all-cause and cardiovascular mortality better than traditional risk factors. 11,[19][20][21][22] On the contrary, the effects of CAC on in-H mortality due to other causes, like sepsis, have been less explored.
The main finding of our study is the presence of calcium, was related to peak HScTnI. Peak HScTnI was linked with all the endpoints. CAC was associated with a higher rate of cardiovascular complications which was likely related to the increase in mortality.
This association was not observed after correcting for traditional risk

| CONCLUSION
Our findings demonstrated that peak HScTnI is linked with all the endpoints in COVID-19 patients. CAC score was not, per se, the strongest marker for the considered endpoints. This arises the possibility CAC score may slightly underestimate the risk of adverse events. These findings support the conduct of larger trials on cardiovascular disease potentially in other infectious and inflammatory diseases.

| Limitations
The study's inclusion criteria of infected patients who had a chest CT selected a higher-risk population, reflected in the higher mortality

CONFLICTS OF INTEREST
The authors declare no conflicts of interest.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions. Note: Data are percentages for categorical variables and I quartile/median/III quartile for continuous variables. The table also reports the results of the univariable models, as AME, p (p-value), and lower and upper bound of the 95% confidence Interval.
PERGOLA ET AL.