Literature Review—Transthoracic Echocardiography, Computed Tomography Angiography, and Their Value in Clinical Decision Making and Outcome Predictions in Patients with COVID-19 Associated Cardiovascular Complications

The sudden outbreak of the COVID-19 pandemic posed a great threat to the world’s healthcare systems. It resulted in the development of new methods and algorithms for the diagnosis and treatment of both COVID-19 and its complications. Diagnostic imaging played a crucial role in both cases. Among the most widely used examinations are transthoracic echocardiography (TTE) and computed tomography angiography (CTA). Cardiovascular complications in COVID-19 are frequently associated with a severe inflammatory response, which results in acute respiratory failure, further leading to severe complications of the cardiovascular system. Our review aims to discuss the value of TTE and CTA in clinical decision making and outcome prediction in patients with COVID-19-associated cardiovascular complications. Our review revealed the high clinical value of various TTE findings and their association with mortality and the prediction of patients’ clinical outcomes, especially when used with other laboratory parameters. The strongest association between increased mortality and findings in TTE was observed for tachycardia and decreased left ventricular ejection fraction (odds ratio (OR) 24.06) and tricuspid annular plane systolic excursion/pulmonary artery systolic pressure ratio (TAPSE/PASP ratio) < 0.31 mm/mmHg (OR 17.80). CTA is a valuable tool in diagnosing COVID-19-associated pulmonary embolism, but its association with mortality and its predictive role should always be combined with laboratory findings and patients’ medical history. D-dimers > 3000 ng/mL were found as the strongest predictors of pulmonary embolism (PE) (OR 7.494). Our review indicates the necessity for an active search for cardiovascular complications in patients with severe COVID-19, as they are linked with an increased probability of fatal outcomes.


Introduction
The novel severe acute respiratory syndrome coronavirus (SARS-CoV-2) was first discovered in December 2019 in Wuhan, China, when several cases of atypical pneumonia of unknown origin were reported [1]. During the next few months, this new disease, called COVID-19, spread rapidly all around the world, with incidence and mortality rates exceeding any expectations. The outbreak of the pandemic of COVID-19 was an unprecedented challenge for people and healthcare systems all around the world that also caused changes in our current lifestyle by introducing various types of restrictions or minimizing interpersonal contact [2]. In addition to many deaths caused by SARS-CoV-2, some convalescents suffer from various types of complications after being infected with the virus [3]. The physicians and scientists were confronted with the urgent necessity of developing methods and algorithms for the diagnosis and treatment of both COVID-19 itself and its long-term health consequences. The crucial factor in determining the severity of the disease as well as its complications is diagnostic imaging [4]. Among widely used examination methods in patients diagnosed with SARS-CoV-2 infection are transthoracic echocardiography (TTE) and computed tomography angiography (CTA). These methods are particularly valuable in establishing the lung surface affected by the disease or the presence of cardiovascular complications, such as pulmonary embolism or heart failure, which are among the most common outcomes of COVID-19 [5]. Diagnostic imaging, if combined with clinical and laboratory findings, may also facilitate early diagnosis. Chest computed tomography (CT) findings include bilateral, peripheral, and basal predominant consolidation, ground-glass opacity, often of an extensive geographical distribution, or both, along with many other less common abnormalities [5]. These findings can also be noticed before the onset of symptoms, and their extent increases significantly during the first and second weeks of the disease and decreases continuously afterward [6]. The evolution of the CT imaging corresponds with the progression of the SARS-CoV-2 infection, which makes it a useful tool to assess the current stage of a patient's condition, as well as to detect COVID-19 pneumonia in some as yet asymptomatic patients [6]. However, those findings are not specific to COVID-19, so the diagnosis should not be established based on diagnostic imaging with no other clinical symptoms or laboratory tests. Moreover, studies show that chest CT may have a prognostic role as a predictor of a higher risk of a severe COVID-19 outcome or death, after adjustment for clinical risk factors and age. They therefore help to identify patients who may benefit from more aggressive treatment [7]. Finally, chest CT has a crucial role in the diagnosis of one of the most common COVID-19 complications, which is pulmonary embolism (PE), as well as atherosclerosis, myocardial injury, or acute myocarditis [8]. Echocardiography also gives a non-invasive assessment of cardiovascular conditions, which allows the physicians to immediately initiate appropriate treatment and consequently significantly improve the patient's prognosis [9]. Therefore, our review aims to discuss the value of TTE and CTA in clinical decision making and outcome prediction in patients infected with the novel coronavirus with cardiovascular complications.
The main subject of research was the role of conventional echocardiography and CTA in the management and treatment of patients diagnosed with COVID-19. Additionally, we also reviewed the most common echocardiography outcomes in SARS-CoV-2 infection. Medical subject heading (MeSH) terms included 'COVID-19', 'transthoracic echocardiography', and 'computed tomography angiography'. Only original research studies concerning over 100 study subjects were included in the literature review. Non-English publications, systematic reviews, and case reports were excluded from the review.

Cardiovascular Complications of COVID-19
Most COVID-19 cardiovascular manifestations are associated with a systemic inflammatory response, which leads to hypercoagulability and hypercytokinaemia [10]. From the pathophysiological point of view, an increase in proinflammatory cytokines such as interleukins 6 and 2 (IL-6 and IL-2), but also tumor necrosis factor-alpha (TNF-α), causes damage to the endothelium, which further generates coagulation cascade, which is the key cause of disseminated intravascular coagulation (DIC), thromboembolic events, or bleeding [10]. However, acute respiratory failure caused by SARS-CoV-2 pneumonia also leads to impairment of the cardiovascular system, as respiratory and cardiovascular systems cooperate to maintain systemic homeostasis [10].
Myocarditis is an inflammation of the myocardium most often caused by infectious agents [11]. The exact mechanism is yet to be elucidated, but the systemic inflammatory response and replication and dissemination of SARS-CoV-2 are presumed to play an important role in myocardial damage [12]. This results in a reduction in the strength of myocardium contractility, which may finally result in acute heart failure [13]. Inflammation of the myocardium might also be complicated by pericarditis, which leads to pericardial effusion and can further result in cardiac tamponade [14].
Heart failure is one of the most important causes of mortality among COVID-19 patients and occurs as a direct myocardial injury caused by the activity of the coronavirus [15]. Myocardial damage is likely to be caused by viral interaction with spike 1 glycoprotein, which activates serine 2 transmembrane protease, resulting in myocardial dysfunction [15]. Respiratory failure leads to acute respiratory distress syndrome (ARDS), which further results in pulmonary hypertension and dysfunction of the right ventricle (RV), whereas septic shock, renal impairment, and volume increase are rather a cause of impairment of LV function [16]. Moreover, heart failure might also be a complication of myocarditis [15].
Thrombotic manifestations of SARS-CoV-2 occur often due to a severe inflammatory response, which leads to vascular damage [17]. A prothrombotic state leads to various thrombotic complications including deep vein thrombosis (DVT), pulmonary embolism (PE), or arterial thrombosis [18]. Less common COVID-19-associated cardiovascular manifestations include acute myocardial infarction (AMI) [11], arrhythmias [19], or takotsubo cardiomyopathy [19]. AMI and takotsubo cardiomyopathy are likely to be linked with myocardial damage, a generalized inflammatory response, and sympathetic activation caused by viral infection [16]. On the other hand, arrhythmias may occur due to the treatment of suspected viral infection [16]. Therefore, it is extremely important to develop methods for the early diagnosis of cardiovascular complications caused by COVID-19.

Diagnosis of COVID-19 Cardiovascular Complications
On physical examination, symptoms of cardiovascular manifestations might be difficult to distinguish from one other, as a majority of patients present with chest pain, dyspnea, cough, and fever [20], but in some cases, also with syncope, palpitations, chest discomfort, or post-exertional fatigue [21]. Nevertheless, those symptoms should be used to define the severity of myocarditis [22]. The presence of the symptoms mentioned should be further examined with the use of other diagnostic tools, including electrocardiography (ECG), the measurement of cardiac troponin (cTn), and echocardiography [22]. The presence of an elevation of the ST-segment (the interval between depolarization and repolarization of the ventricles) with no reciprocal ST-segment depression, an increased QRS complex (depolarization of ventricles) duration or a diffuse inversion of the T-wave (repolarization of the ventricular myocardium), the elevation of cTn, and abnormalities of ventricular motion in echocardiography indicate a possible inflammatory process of the myocardium [22]. Further steps should include cardiac magnetic resonance (CMR) and a biopsy of the myocardium [22].
The diagnosis of SARS-CoV-2 associated heart failure is similar, as it also requires a physical examination, the measurement of laboratory parameters such as atrial natriuretic peptide (ANP) and B-type natriuretic peptide (BNP), and an examination of heart condition via ECG and echocardiography [16]. In addition, laboratory examinations should include natriuretic peptides, and cardiac function should be further monitored using echocardiography [16]. Moreover, the diagnosis of COVID-19-associated PE is also difficult. COVID-19 pneumonia might diminish PE symptoms, as typically, reported dyspnea and hypoxemia are likely to occur also during viral pneumonia [23]. Clinical probability scores such as the Wells score may be a support tool, but it is important to remember that underestimation of the COVID-19-associated PE probability is possible [24]. D-dimer is also used to assess clinical probability, but it is not clear if its value can definitely "rule out" or "rule in" PE probability [25]. Therefore, CTA should be regarded as a first-choice method due to its availability and accuracy [23].

TTE in COVID-19 Cardiovascular Complications
Many studies are focused on using TTE as a method for COVID-19 complication detection. Therefore, we try to summarize and systematize the current state of knowledge in this field in Table 1 and the detailed descriptions below.  A prospective international study that consisted of 1216 patients [9] showed that in 405 patients, abnormal echocardiographic findings led to a change in patient management. Abnormalities in echocardiograms were found in 667 patients. Of these, 305 were diagnosed with left ventricle (LV) abnormalities and 185 with RV abnormalities, and 174 had abnormalities in both ventricles [9]. In addition, abnormalities were also more common among patients with an ST-segment elevation in their ECG or increased biomarkers of cardiac function [9]. The study, which included 1000 patients, showed that a severe or moderate course of COVID-19 is more associated with significantly impaired echocardiographic parameters than a mild course of the infection [26]. During 3 months' follow-up, echocardiography showed a significant change in LV internal dimension-diastole (LVIDd) (+1 ± 0.3 vs. +0.6 ± 0.3 mm) with a 95% confidence interval (CI) (0.32-1. An international multicenter study of 870 subjects found that approximately 30% of subjects were diagnosed with RV dysfunction and approximately 20% with LV dysfunction [27]. Mild LV dysfunction (LVEF, 40-50%) was noted in 11%, whereas moderate (LVEF, 30-40%) and severe (LVEF < 30%) were noted in 5% and 3%, respectively [27]. In addition, significant differences were also noted, as patients from Asia had better echocardiographic parameters, including LVEF, LV longitudinal strain (LVLS), RV free-wall strain (RVFWS), and RV basal diameter (RVBD), in comparison to subjects from the United States, Europe, and Latin America [27]; these parameters were also linked to mortality. A retrospective analysis of 724 hospitalized patients with a minimum of one echocardiographic examination [28] showed LV diastolic dysfunction in 20% and RV systolic dysfunction defined as RVFWS < 20% in 16% of hospitalized subjects but did not report any significant differences between intensive care unit (ICU) and non-ICU patients [28]. The prediction of fatal outcomes using a receiver operator characteristics (ROC) curve resulted in their increase after the addition of echocardiographic measurements (a rise from AUC = 0.77 (CI 0.70-0.84) to AUC = 0.91 (CI 0.85-0.96)) [28].
Studies that consisted of 301-500 patients reported that enlargement of the RV in TTE was more prevalent in patients with shock (OR 1.81, CI 1.03-3.18), thromboembolic events (OR 2.31, CI 1.37-3.88), the need for renal replacement therapy (RRT) (OR 2.35, CI 1.31-4.21), or a fatal outcome at 60 days (OR 1.93, CI 1.13-3.30), and abnormal RV function was more often linked to shock (OR 1.75, CI 1.21-2.92) [34]. RV dilatation was significantly associated with patients' primary composite outcome (ICU or death) (p = 0.03) [35]. Firstphase EF < 25% was found to be a strong predictor of death due to the increased mortality of subjects with first-phase EF < 25% (35.7%) in comparison to subjects with first-phase EF > 25% (7.8%) [36]. RV dysfunction, LV wall abnormalities, LV global dysfunction, diastolic dysfunction grade II or III, and pericardial effusion were more common in patients with elevated biomarkers of myocardial injury than in subjects without elevation ( [37]. The observational study divided patients into three classes: first, normal RV function (52%), second, dilated RV with mostly preserved systolic function (31%), and third, RV dilatation with systolic impairment (17%), which showed a significant difference in 90-days mortality, outcome, and response to mechanical ventilation: 22%, 42%, and 73%, respectively, p-value (p < 0.001) [38]. RV dysfunction in TTE and troponin levels were also found to be good predictors of pulmonary embolism (PE) (AUC = 0.77) [39]. A retrospective study of 368 subjects found a significant association between LA dilation and LV thrombus in patients with ischemic stroke in comparison to subjects without (48.3% vs. 27.9%, p = 0.04; 4.2% vs. 0.7%, p = 0.03 respectively) [40]. The OR of the COVID-19 ischemic stroke risk, which also consisted of the findings mentioned, was 4.1 (CI 1.40-16.10), and AUC = 0.70 [40].
Some reports indicate that the use of invasive mechanical ventilation (IMV) might influence the results of TTE [60]. In studies that were utilized in our review, there was no information about TTE alterations among patients with advanced ventilatory support, but we found one study describing the possible impact of invasive mechanical ventilation on a lower prevalence of RV abnormalities [61].
To sum up, studies conducted on bigger cohorts indicate the precious clinical value of TTE in the clinical management of patient conditions, a higher prevalence of ventricular abnormalities in COVID-19-associated cardiovascular complications, and continental differences in the severity of cardiovascular complications, but also a less pronounced association between TTE findings and fatal outcomes. On the other hand, studies with smaller cohorts show a strong link between various TTE findings and mortality and the risk for ICU admission and/or ventilatory support, but they also report no association at all. Consistently across all studies, TTE was shown as a strong method that leads to the prediction of patients' conditions, which can be used for the prediction of patients' clinical outcomes, their risk of death, or changes in management. The summary of the studies discussed is presented in Table 1. Associations between TTE findings and endpoint outcomes are also presented in Figure 1.

CTA in COVID-19 Cardiovascular Complications
Another method most often used to observe COVID-19 cardiovascular complications is CTA. We summarize observed CTA changes in the working cardiovascular system in

CTA in COVID-19 Cardiovascular Complications
Another method most often used to observe COVID-19 cardiovascular complications is CTA. We summarize observed CTA changes in the working cardiovascular system in Table 2 and describe the available literature reports in more detail below.  Patients with PE compared with patients without PE were more frequently hospitalized in ICU and required mechanical ventilation CAP-community acquired pneumonia; cTnT-hs-highly sensitive cardiac troponin T; CTPA-computed tomography pulmonary angiogram; CTSS-computed tomography severity score; ED-emergency department; HR-hazard ratio; ICU-intensive care unit; IMV-invasive mechanical ventilation; OR-odds ratio; PAO-pulmonary artery obstruction; PASP-pulmonary artery systolic pressure; PE-pulmonary embolism; TAPSE-tricuspid annular plane systolic excursion; VTE-venous thromboembolism.
A retrospective observational study of 1240 patients found 8.3% of patients were diagnosed with PE, but 77.7% were diagnosed using CTA within the first 48 h after admission [62]. However, another large-scale case-control study of 316 subjects including 158 individuals with confirmed SARS-CoV-2 infection suggested the overutilization of CTA in COVID-19-associated PE, as it was diagnosed in 8.9% of COVID-19 subjects in comparison to 39.9% of non-COVID-19 patients [63]. Another study with 413 subjects found PE among 25% of all subjects-29% of ICU vs. 24% of non-ICU patients-but did not report any significant difference in mortality between patients with and without PE [64]. A retrospective, multicenter, observational study with 399 patients diagnosed 22% with PE, with 32% of cases occurring among SARS-CoV-2 positive individuals, but no significant difference between radiologic, clinical, or laboratory parameters and their impact on outcomes was noted [65]. Another multicenter study detected PE in 33% of cases but did not find any significant difference in mortality [66]. Retrospective comparative research conducted on 300 subjects confirmed PE in 15% of patients, with bilateral involvement in 57%, highly sensitive cardiac troponin T (cTnT-hs) and N-terminal (NT)-pro hormone BNP (a non-active prohormone that is released from the same molecule that produces BNP) and significantly elevated in PE patients; D-dimer was elevated in all patients, with no significant difference between PE and non-PE patients.  [71].
A study with fewer than 200 patients reported an association between computed tomography severity score (CTSS) and the increased probability of ICU admission (OR 1.21, CI 1.10-1.34), and death (OR 1.15, CI 1.03-1.30) [72]. A multicenter study of 169 patients confirmed PE in 15.4%; median D-dimer was significantly higher in patients with PE vs. those without PE (9.84 mg/L vs. 1.64 mg/L) [73]. Another multicenter research study with a similar number of patients detected PE in 44.7%; a significant correlation between pulmonary artery obstruction (PAO) index and D-dimer level (p = 0.002) was reported, but without any significant differences between CTA or laboratory findings and patients' outcomes [74]. CTA research on 100 subjects found 235 with acute PE. PE patients were more frequently hospitalized in ICU (74% vs. 29%) and required mechanical ventilation (65% vs. 25%) [75].
In sum, CTA might be a valuable tool to confirm the diagnosis of COVID-19-associated PE, but some studies also report no association and the overutilization of CTA. Clinical outcomes, the mortality risk, and changes in the management of COVID-19-associated PE should always be analyzed with other clinical parameters, especially D-dimers, troponins, natriuretic peptides, or the patient's medical history and general condition. The predictive factors and endpoints of PE are presented in Figure 2.

Conclusions and Prospects
The results presented in this review indicate that various parameters measured TTE have prognostic value, which may influence the clinical decisions of physician garding patients' management. The most frequently changed parameters detected TTE belong to LV dysfunction measured as decreased values for LVEF, abnorma function, TAPSE, or TAPSE/PASP ratio. These changes are associated with the predi of clinical outcomes and mortality. TAPSE/PASP < 0.31 mm/mmHg, as well as comb tachycardia and decreased LVEF, were reported as the most significant risk factors ciated with increased mortality (OR 17.80, CI 3.70-86.31, and OR 24.06, CI 4.63-12 respectively). Nevertheless, TTE parameters alone or combined with patients' me history and laboratory parameters, including troponin, D-dimers, or natriuretic pep have high value as predictors of patients' clinical outcomes in some cases. CTA is a cTnT-hs-highly sensitive cardiac troponin T; VTE-venous thromboembolism; CTPA-computed tomography pulmonary angiogram; ICU-intensive care unit; PE-pulmonary embolism; *-increased probability when alterations on echocardiographic examinations are present.

Conclusions and Prospects
The results presented in this review indicate that various parameters measured with TTE have prognostic value, which may influence the clinical decisions of physicians regarding patients' management. The most frequently changed parameters detected with TTE belong to LV dysfunction measured as decreased values for LVEF, abnormal RV function, TAPSE, or TAPSE/PASP ratio. These changes are associated with the prediction of clinical outcomes and mortality. TAPSE/PASP < 0.31 mm/mmHg, as well as combined tachycardia and decreased LVEF, were reported as the most significant risk factors associated with increased mortality (OR 17.80, CI 3.70-86.31, and OR 24.06, CI 4.63-125.11, respectively). Nevertheless, TTE parameters alone or combined with patients' medical history and labo-ratory parameters, including troponin, D-dimers, or natriuretic peptides, have high value as predictors of patients' clinical outcomes in some cases. CTA is a valuable tool for making a firm diagnosis of PE, but clinical management and predictive value should be always combined with patients' clinical history and/or laboratory parameters. D-dimers with a value over 3000 ng/mL were found to be the best predictive factor for PE (OR 7.494, CI 3.038-18.484), and IMV was reported as the most significant risk factor for PE (OR 8.07, CI 2.70-23.82). However, results obtained by CTA, such as fatal outcomes and an urgent need for ICU admission, can relate to D-dimers, cTnT-hs and NT-pro BNP, and CTA imaging findings.
It is worth considering whether there is a necessity for an active search for thrombotic or cardiac problems in every patient previously hospitalized due to SARS-CoV-2 or only in patients with specific risk factors, and if so, what those risk factors should be. Another matter for discussion is which examination would be the best for screening for COVID-19 complications-many factors, such as sensitivity and specificity, costs, or the duration of such an examination, should be taken into consideration. For that reason, further research is required to establish whether there is an actual necessity for active screening for a post-COVID-19 complication such as pulmonary embolism, and if the answer to this question is positive, then one would need to decide which group of patients should be considered for such screening and which methods should be used.

Limitations
The study is subject to several limitations, including the many retrospective studies used, which may be prone to selection and observer bias. Furthermore, the study population had varying stages of disease progression, coexisting conditions, and other risk factors, as well as specific conditions during the conduct of the research, especially in relation to TTE, which could have influenced the final outcome.