Post-acute Cardiovascular Sequelae of COVID-19: an Overview of Functional and Imaging Insights

Inappropriate sinus tachycardia, atrial tachycardia, postural orthostatic tachycardia, premature ventricular and supraventricular contractions, and some other arrhythmias have been frequently described as post-COVID cardiac complications and can last beyond 6 months after initial infection. Several studies about electrocardiogram (ECG) and ambulatory monitoring by wearable devices have been described, but they are beyond the scope of this review regarding cardiac imaging and CPET [15].

Echocardiography is an inexpensive and widely available tool with several advantages in diagnosing both acute and post-acute cardiovascular COVID manifestations. It has some security benefits for patients such as a lack of ionizing radiation or nephrotoxic contrast media. It has the ability to accurately evaluate the presence and severity of pericardial effusion and/or constriction. It can identify contractility defects, decrease in right or left ventricle function, evaluate chamber dilation, valvular regurgitation/stenosis, the likelihood of pulmonary hypertension, etc.

During the acute phase, bedside echocardiography is the first imaging tool during the initial evaluation. Right ventricle dilation and dysfunction are by far the most frequently found abnormalities probably secondary to hypoxic pulmonary vasoconstriction, direct myocardial injury, ventilator-induced lung injury, thromboembolism, etc. Other abnormalities including left ventricle dysfunction are also present but to a much lesser degree [15]. Nevertheless, after the acute phase (Long-COVID), some studies have shown normal echocardiography findings or slight ventricular remodeling even in patients with persistent symptoms [15]. Another retrospective study by Sechter et al. with a median of 142 days after the acute COVID infection showed abnormal echo findings in up to 32% of cases although most of them were classified as mild: minimal pericardial effusion was the most common finding in 19% followed by left ventricle dysfunction in just 4% [16].

A multicenter, prospective, observational, cohort study carried out by Moody et al. including 79 post-severe-COVID patients 3 months after hospital discharge found abnormal right ventricle systolic function (TAPSE < 17 mm or FAC < 35%) in 14%, whereas left ventricle dysfunction (EF less than 50%) was found in just 9%. Among those with RV dysfunction, 44% had pulmonary embolism previously diagnosed on computed tomography angiography during admission. Almost one-third of patients (29%) had persistent right ventricle involvement (dilation or dysfunction) at 3-month follow-up [17].

On the other hand, several works have failed to show abnormal cardiac findings on echocardiography 3–6 months after COVID-19 besides persistent cardiovascular symptoms and, therefore, lack sufficient evidence supporting echocardiographic parameters and their relationship with symptoms, even in patients with positive biochemical biomarkers (troponins). This finding should not be taken as discouraging with regard to echo use but rather suggest that other more advanced techniques may be needed to rule out whether or not there is underlying cardiac damage that explains the patient’s symptoms [18‐20].

Cardiac CT offers a better image quality than echocardiography which is especially useful in cases with poor acoustic windows and requires less exposure time between healthcare workers and patients. It has also the advantage of being able to evaluate surrounding structures such as lungs and pulmonary vessels, identifying in many cases potential life-threatening conditions during acute COVID: pulmonary embolism, coronary artery disease, parenchyma lung disease (fibrosis, edema, and pleural effusion) and myocardial dysfunction.

Also, in hospitalized COVID patients with no known coronary artery disease, a coronary calcium score (CAC) > 400 HU can identify patients at risk of adverse cardiovascular outcomes such as hospital mortality and ICU admission [21]. Another study identified a positive CAC in 50.7% of patients hospitalized for COVID-19. Of those, 50% experienced noninvasive or invasive mechanical ventilation, ECMO, or death within 30 days after admission compared with 17.5% in patients with negative CAC [22]. These findings suggest that the CAC score could be a useful test for coronary artery disease stratification in this population.

Nevertheless, in both acute COVID-19 and PASC, the main role of cardiac CT is to noninvasively confirm or rule out the presence of cardiovascular disease by using coronary CT angiography which has a negative predictive value of 90.9%, positive predictive value of 87.9%, sensitivity 96.5%, and specificity of 72.4% for detection of coronary artery disease [23].

Cardiac CT can not only confirm the presence of significant coronary obstruction but also characterize such lesions determining if they pose high-risk features for rupture such as the napkin-ring sign, spotty calcifications, and eccentric remodeling [24]. An increased epicardial adipose tissue attenuation in patients with severe COVID as demonstrated by Iacobellis et al. could play a potential role as an inflammatory marker associated with plaque vulnerability [25].

Furthermore, with the implementation of technologies such as dual source CT, it is possible to identify myocardial scar and quantify extracellular volume fraction (ECV) through late iodine enhancement (quadruple rule-out) [26]. However, this technique has still some limitations such as a low contrast-to-noise ratio, the need for highly trained observers, and requires more contrast volume than a normal CT. In consequence, tissue characterization by CT is not widely available yet, and CMR remains the gold standard for the noninvasive characterization of myocardial tissue [27]. In addition, a dual-energy CT (DECT) angiographic study can aid in detecting pulmonary perfusion defects (proximal and distal thrombosis) usually overlooked during acute infection [28].

When evaluating cardiovascular involvement after COVID-19, cardiac magnetic resonance stands as the most studied imaging tool. It has several advantages over other techniques as it is not undermined by a poor acoustic window, does not require ionizing radiation, and has a high signal-to-noise ratio, which is very useful in detecting subtle changes of the myocardial parenchyma associated with persistent inflammation or scarring. Multiple studies have shown features associated with myocardial injury; some of these trials are summarized in Table 1 [29‐33, 36, 38, 40‐43].

Early in 2020, a prospective observational cohort study evaluated 100 patients who recently recovered from COVID-19 (33% previously hospitalized) with a mean of 71 days after diagnosis. Among them, 78% had abnormal CMR findings, increased T1 in 73%, and T2 in 60% (indicating edema and ongoing myocardial inflammation). Endomyocardial biopsy in severe cases showed active lymphocytic inflammation. In addition, COVID patients showed higher ventricular volumes and lower ejection fractions than matched controls. T1 and T2 had the best discriminatory power for the detection of COVID-19-related involvement, against age, sex-matched healthy controls, and risk factors matched patient controls [30].

More recent studies also support the ability of CMR in detecting myocardial damage during the first 3–6 months after COVID-19 infection. A prospective study involving 148 post-severe COVID-19 patients with a mean time of 68 days after infection showed decreased EF in 11%, the persistence of LGE in 54%, myocardial infarction scars in 19%, and criteria compatible with ongoing myocarditis in 30% [31]. Myopericarditis was also identified as probable (41%) or very likely (13%) in 159 previously hospitalized post-COVID patients between 28 and 60 days post-discharge [32]. In 2022, Eiros et al. reported high rates of CMR abnormalities in up to 75% of healthcare workers at 10.4 weeks post-infection: myocarditis 26%, myopericarditis 11%, and pericardial effusion 30% [33].

Beyond the first 3 months after infection, abnormal CMR features tend to be observed in a lesser number of patients and even absent at all after 6 months in some studies [33‐35]. In other studies, however, pathological findings such as decreased LVEF, LGE, or persistently elevated T1 and T2 mapping were observed several months after infection suggesting that ongoing myocardial inflammation could contribute to the pathophysiological mechanisms underlying persistent cardiac symptoms [36, 37].

A prospective, observational cohort study showed cardiac impairment in 26% of 201 patients with a median of 141 days following SARS-CoV-2 infection, notwithstanding most patients (60%) persisted with severe post-COVID symptoms [38]. Another study showed that 19% of Long-COVID patients had abnormal CMR at baseline, including decreased LVEF, increased LV volumes, and increased T1. After a 1-year follow-up, up to 58% of these patients still had these findings [39].

Despite a trend to normalize pathologic CMR findings, many persons still complain of dyspnea, exercise intolerance, chest pain, palpitations, and a myriad of cardiovascular-related conditions. At this point, clinicians should avoid the temptation to assume a psychosomatic cause and keep pursuing the mechanisms responsible for these persistent conditions. Notably, a physiologic evaluation of a patient’s response to exercise by a cardiopulmonary exercise testing can help identify potential mechanisms underlying symptoms.

Some possible PASC mechanisms such as myocardial injury, lung injury, vascular abnormalities, muscular dysfunction, and dysautonomia can be evaluated by cardiopulmonary exercise testing as a complement to imaging, allowing a more physiological approach in these patients, since most of the symptoms reported are related to poor exercise capacity. Continuous monitoring of parameters helps to orientate if such symptoms are associated with ventilatory, circulatory, or gas exchange impairments.

Perhaps the most consistent finding among PASC patients is a reduced peak oxygen uptake (V̇O₂). In a study involving 41 patients with PASC (approx. 8.9 months after infection) and previously normal pulmonary function tests and cardiac CT, 58.5% had a reduced peak V̇O₂ (< 80% predicted) with circulatory limitation to exercise. Despite normal pulmonary function tests, ventilatory impairment was present in 88% of patients, with dysfunctional breathing in 63%; hypocapnia (PetCO₂ < 35) in 61%; and/or increased V̇E/V̇CO₂ slope in 41%, which can be interpreted as a manifestation of autonomic dysfunction [44]. Similarly, Ladlow et al. studied 205 patients and identified dysautonomia in 25%, defined in this study as a resting HR > 75 bpm, HR increase with exercise < 89 bpm, and HR recovery < 25 bpm within 1 min after exercise. Patients with this condition also showed decreased workload, lower peak V̇O₂, and a steeper V̇E/V̇CO₂ slope [45].

Another possible explanation for reduced peak V̇O₂ could be a diminished oxygen extraction (muscular deconditioning) secondary to muscle damage in patients suffering from cytokine storm, prolonged hospital stay, direct viral damage, etc. [46]. It is associated with an earlier anaerobic threshold, accelerated heart rate response, lower workload achieved, lower peak O₂-pulse, slow V̇O₂ on-kinetics during unloaded cycling, and wider breathing reserve.

Additionally, chronotropic incompetence is one of the most frequently observed abnormalities among patients with reduced peak V̇O₂ after COVID-19 being present in up to 60%. Consequently, chronotropic incompetence is associated as a potential cause or contributor to reduced exercise capacity observed in PASC patients [41].

While approaching a patient with persistent symptoms after COVID-19, focus should be placed on determining whether his/her chief complaint derives from (1) a potentially fatal condition, (2) a condition that will carry adverse cardiovascular effects in the future, or (3) sequelae related to dysautonomia or deconditioning that can cause important limitations on the patient’s quality of life.

As previously stated, results from clinical trials using echocardiography have shown scarce pathological findings on patients with PASC, as opposed to acute COVID-19. Nonetheless, it continues to be a useful tool, readily accessible to most cardiology departments and can be performed during initial clinical evaluation. On the other hand, CMR, cardiac CT, or cardiopulmonary exercise testing are not as easy to access; therefore, we suggest getting transthoracic echocardiography as the first imaging modality accompanying physical examination. It is mandatory to emphasize that a normal echocardiogram does not exclude cardiovascular conditions, and moving to a more advanced image technique is highly advisable when feasible [47, 48].

A pathological finding on transthoracic echocardiography or the persistence of cardiovascular symptoms dictates the necessity of pursuing further study of the patient through more advanced imaging techniques. As has been exposed in this paper, CMR is postulated as the ideal imaging modality to approach PASC; it can diagnose pericarditis and/or acute myocarditis allowing a timely administration of anti-inflammatory drugs and therefore decreasing the risk of fibrosis and posterior cardiac remodeling. According to ESC guidelines, CMR can also help to diagnose significant ischemia through perfusion studies, guiding coronary revascularization, and optimizing anti-ischemic therapy [49].

In cases where images do not show any damage or the findings are not severe enough to explain the patient’s symptoms, we suggest performing an additional test to identify underlying mechanisms behind them. As shown before, CPET seems to be an excellent tool to be used in combination with CMR, due to its ability to give a more physiological approach to the evaluation of PASC through a ventilatory, circulatory, and metabolic analysis of patients under exercise. Accordingly, relevant limitations such as deconditioning, chronotropic incompetence, or ventilatory inefficiency can be identified. In some cases, additional exams like tilt-testing and ambulatory ECG-monitor can be employed for the detection of POTS, inappropriate sinus tachycardia, etc. (Fig. 2) [50, 51].