Clinical Prognosis of Pericardial Effusion in COVID 19

Purpose In our study, we investigated the relationship between pneumonia severity and pericardial effusion, predisposing factors and the effect of pericardial effusion on clinical prognosis and mortality in COVID-19 patients. Methods A total of 3794 patients who were diagnosed with COVID- 19 by polymerase chain reaction (PCR), were hospitalized between March 21 and November 30, 2020 were included in the study. For each of the 3794 patients, the initial chest CT images, pericardial efusion (PE), pleural efusion and pneumonia severity were evaluated.


Introduction
The coronavirus disease 2019 , caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), was rst reported in Wuhan, Hubei Province of China, and then spread rapidly all over the world, and the World Health Organization (WHO) declared the outbreak a pandemic. While some cases may be asymptomatic in COVID-19, they can manifest from mild clinical symptoms to severe clinical involvement such as pneumonia and acute respiratory distress syndrome (ARDS). Some of the patients may apply to health institutions with atypical complaints related to gastrointestinal, genitourinary, cardiovascular and other systems. Additional comorbidities and demographic characteristics of patients Page 3/17 have an important place in the clinical course. Patients with cardiovascular disease or increased cardiovascular risk factors are more susceptible to the development of major clinical complications [1].
Cardiac damage is a clinical nding secondary to respiratory failure, hypoxemia, direct myocardial infection with virus, microvascular thrombosis and systemic in ammation and is a common clinical nding and is associated with poor prognosis [2]. Another cardiac nding in this process is pericardial effusion (PE), which is the most common clinical presentation of pericardial diseases. PE can be a complication of lung parenchyma infections, pleural infections, and some diseases, and it is often seen as viral acute pericarditis [3]. The effect of the virus in the pericardium occurs by a direct cytotoxic and/or immune-mediated mechanism [4]. In some studies in which a limited number of patients were examined, the incidence of PE in COVID-19 patients was found to be approximately 5% [5]. In another study, patients with severe/critical COVID-19 were found to have a higher incidence of PE than patients with mild disease [6]. As a matter of fact, extensive studies on the coexistence of COVID-19 and PE are limited. The main purpose of our study is to investigate the relationship between pneumonia severity and PE in COVID-19 patients, the predisposing factors and the effect of PE on clinical prognosis and mortality.

Patients
A total of 3794 patients who were diagnosed with COVID-19 by polymerase chain reaction (PCR), were hospitalized, and started treatment between March and November 2020 were included in the study. This study was performed in accordance with the Helsinki Declaration and with the approval of the local ethics committee. Patients' clinical and demographic characteristics, computarized tomography (CT) images, hematological and biochemical parameters were recorded. Patients with lung malignancy, a history of lobectomy, tuberculosis, or atelectasis, those under treatment for a recent diagnosis of pleural effusion, PE, and non-COVID-19 pneumonia were excluded.
De nitions PCR test: A combined swab sample was taken in accordance with the speci ed procedures in all patients admitted to the emergency room clinics [7].The patients were diagnosed with COVID-19 based on their positive PCR test results. Hypertension (HT) was de ned as a systolic blood pressure > 140 mmHg and/or a diastolic blood pressure > 90 mmHg or the use of an antihypertensive drug [8]. Diabetes mellitus (DM) was de ned according to the current American Diabetes Association guidelines [9]. A history of coronary artery disease (CAD) was de ned as patients who had invasive or noninvasive imaging studies showing evidence of coronary artery disease. Heart falure was de ned according to the current European Society of Cardiology guidelines [10] Renal failure (CRF) was de ned as creatinine clearance below 60 ml/min.The Cockroft-Gault equation was used to calculate the eGFR. The clinical comorbidities of the patients were obtained from the anamnesis recorded in the hospital automation.
CT is accepted as a complementary imaging modality to echocardiography in the evaluation of PE. Size and extent of PE can be better evaluated using CT or magnetic resonance imaging (MRI) than transthoracic echocardiography (TTE); the smallest amount of pericardial uid detectable by CT is approximately 10 ml [11]. The presence of > 4 mm of uid between both pericardial layers on CT is considered abnormal. At the beginning of the COVID-19 pandemic, patients were not administered routine TTE for the etiology of dyspnea, and a signi cant proportion of inpatients did not have TTE. Therefore, retrospective evaluation of PE was made with the ndings in CT. TTE reports of patients with massive effusion on CT and clinically suspected cardiac tamponade were obtained. Recommended classi cation for PE in current guidelines; the size of pericardial effusion on two dimensional echocardiography is qualitatively assessed by the end diastolic distance of the echo free space between the epicardium and parietal pericardium: small (10 mm), moderate (10-20 mm), large (20 mm). In our study, the classi cation of PE size in CT was performed as in the classi cation model according to TTE [3].
The severity of pulmonary involvement on CT -Assessment and patients groups. Chest CT severity score (Chest CT-SS) In some studies, it has been observed that the clinical classi cation (mild, widespread, severe or critical illness) formed as a result of the visual (semiquantitative) evaluation of patients with COVID-19 pneumonia according to CT ndings is compatible with the prognosis [6,12,13]. This method is an adaptation of a method previously used to describe CT ndings correlated with clinical and laboratory parameters in post-SARS patients, and the percentage of involvement of each lung lobe (5 lobes) was calculated semi-quantitatively (visually) [14]. According to the classi cation made according to the rate of involvement of the lung lobes, no involvement 0 points (no involvement), 1%-25% involvement 1 point (minimal involvement), 26%-50% involvement 2 points (mild involvement), 51%-75% % involvement was evaluated as 3 points (moderate involvement), and 76%-100% involvement was evaluated as 4 points (severe involvement). The total lung involvement severity score (CT-SS) (between 0-20) was obtained by summing the scores of all these lobes [12,13,15]. In addition, in another clinical study, parenchymal opaci cation in lung segments (20 segments) was evaluated to include 0%, less than 50% or equal to and/or more than 50% of each region, and 0, 1, 2 points were given, respectively, and created CT-SS. In this study, it has been shown that semiquantitative CT evaluation correlates with the disease severity, clinical follow-up and prognosis at admission, and its clinical use is bene cial [16,17]. In Our study, two radiologists were blinded to the clinical data evaluated the CT ndings in consensus. For each of the 3794 patients, the initial chest CT images were evaluated for the following characteristics: ground glass opacity (GGO), consolidation, interlobular septal thickening, bronchial wall thickening, subpleural line, lymph node enlargement, pleural effusion, and pericardial effusion in accordance with the standard morphologic descriptors, based on the previous similar studies [18,19]. The CT-SS was de ned by summing up individual scores from 5 lung lobes; scores of 0, 1, 2, 3 and 4 were respectively assigned for each region if parenchymal opaci cation involved 0%,1%-25%, 25%-50%, ≥ 50%-75% or 75%-100% of each region. In our study, CT-SS was de ned as the sum of the scores obtained from the involvement of 5 lung lobes, and CT-SS was determined as 0-20. Those without AC involvement were classi ed as group 1, those with AC involvement below 50% were classi ed as group 2 (minimal and mild involvement), and those with AC involvement ≥ 50% were classi ed as group 3 (moderate and severe involvement). The interobserver discrepancy was observed in the evaluation of CT of 87 patients. The nal decision regarding AC involvement of these patients was made according to the CT of those who had other CT and according to the clinical ndings of those who did not.

Chest CT Scan
All CT images were reviewed at a window width and level of 1000 to 2000 HU and − 700 to − 500 HU, respectively, for lung parenchyma. Chest CT imaging was performed using a 64-detector CT scanner (AQUILION; TOSHIBA). All patients were examined in the supine position. CT images were then acquired during a single inspiratory breath-hold. The scanning range was from the apex of the lung to the costophrenic angle. CT scan parameters were as follows: x-ray tube parameters 120 kVp, 110-270 mAs, FoV 400 mm; section thickness 5 mm.

Statistical Analysis
All statistics were analyzed via SPSS 22.0 software (SPSS Inc., Chicago, IL, USA). Categorical variables are presented as percentages, whereas continuous variables are presented as mean ± standard deviation or median (interquartile range). Baseline characteristics were classi ed according to prede ned subgroup and evaluated via appropriate statistical tests including independent samples t-test for continuous variables with normal distribution, Mann-Whitney U test for continuous variables with non-normal distribution, appropriate chi-square test for categorical variables and ANOVA for parametric variables with three independent groups. The Kruskal Wallis-H test was used when examining three groups in the analysis of variables with non-normal distribution. The regression analysis was performed on the statistically signi cant parameters obtained from the univariate analysis, and independent predictors of in-hospital mortality were investigated. Tomogra k parametrelerin mortalite ile ilişkisini araştırmak için bu parametreler regresyon analizine alındı. A p value ≤ 0.05 was considered signi cant.

Results
Patients were divided into three groups according to AC involvement. In the rst group, there were 560 patients who did not have AC involvement but were hospitalized considering other clinical ndings and laboratory characteristics, 2639 patients in the second group with AC involvement below 50%, and 595 patients in the third group with 50% or more AC involvement. The median age of group 1 was 47, group 2 was 63, and group 3 was 70 (p < 0.001). The rate of male patients was 45.4% in the 1st group, 47.9% in the 2nd group and 58.2% in the third group. (p < 0.001). The baseline cardiac and noncardiac comorbidities of the patients, laboratory data at the time of admission to the emergency department, pericardial and pleural involvement rates according to the severity of AC involvement, as well as the need for intensive care and mortality rates during follow-up are given in Table-1. Accordingly, as the severity of AC involvement of the patients increased, statistical signi cance was observed in mortality rates, the need for intensive care unit admission, the rate of deterioration in laboratory parameters, and the incidence of pleural and PE.
PE was present in 145 of the patients and PE was not observed in 3649. The median age of the patients with PE was 72, and the median age of the patients without effusion was 63 (p < 0.001). 51% of patients with pericardial effusion were male, and this rate was 49.1% in those without effusion. (p < 0.001). The clinical characteristics, demographic data, laboratory parameters, need for intensive care during follow-up and mortality rates of the groups separated according to the presence of PE are given in Table-2. In the group with PE, cardiac and non-cardiac comorbidities, laboratory ndings showing the severity of COVID-19 pneumonia at admission, rates of AC involvement, need for intensive care hospitalization and total mortality were found to be statistically signi cant. When these parameters obtained by computed tomography were evaluated, lung involvement, presence of PE and pleural effusion were found to be associated with mortality (p < 0.001 for all parameters). These parameters are evaluated by regression analysis; lung involvement, presence of PE, and presence of pleural effusion were found to be independent predictors of mortality ( p < 0.001, 0.019 and < 0.001, respectively). In addition, the relationship between AC involvement of the patients and the presence of PE and pleural effusion in the patients and their mortality is shown in Table-3.

Discussion
Although COVID-19 mainly affects the lungs, it can also affect the cardiovascular system and other organs and present with a wide clinical spectrum. Cardiovascular in uence such as acute cardiac injury, stroke, and pulmonary embolism are common. PE, occurs as direct involvement of COVID-19 or as a complication of the disease. The pericardium normally contains a small amount of uid (15-50 ml) [20]. The known pathophysiology of PE is often impaired lymphatic or venous drainage of the heart. The most common causes of PE are heart or kidney failure, followed by infection, neoplasia, and myocardial infarction. While patients with PE may sometimes be asymptomatic, sometimes they may present with general condition disorder and hemodynamic disorder [21,22]. In a study, the incidence of PE in the CT of patients with COVID-19 was reported to be approximately 5% [5]. However, the effect of PE on the incidence and clinical prognosis of PE according to the severity of the disease or the stage of AC involvement has not yet been demonstrated in large series [23].
In our study, AC involvement of 3794 patients was divided into three groups and its relationship with PE was investigated retrospectively. In patients without AC involvement (group 1), the group with AC involvement below 50% (group 2), and the group with AC involvement above 50% (group 3), as AC involvement or the severity of the disease increased, male gender and advanced age become statistically signi cant. As a matter of fact, the importance of age, gender and comorbidity in the progression of the disease is now well known [24,25] In all three groups, cardiovascular diseases and other comorbidities are observed more frequently in patients with increased severity of AC involvement, consistent with the literature [26].
Among the laboratory parameters that have prognostic importance in COVID-19 disease, especially markers such as lymphocyte level and percentage, d-dimer, ferritin, C-reactive protein (CRP) and troponin change statistically signi cantly in patients with severe disease and AC involvement. Presence of PE according to the degree of AC involvement, respectively; 0.7% in group 1, 2.3% in group 2, and 13.3% in group 3. In addition, when the intensive care needs of these patients in their follow-up are examined; It was observed as 5.4% in group 1, 11.3% in group 2 and 47.6% in group 3. The total mortality rates of these patients during the hospitalization were 1.6% in the 1st group, 8.5% in the 2nd group, and 37.5% in the 3rd group. Similarly, in studies including patients followed up with COVID-19 pneumonia, a close relationship was observed between the severity of AC involvement (with high CT-SS) and adverse clinical prognosis and mortality [27]. When we consider all patients according to the presence of PE, statistical signi cance is observed in patients with PE in age, gender, comorbidities such as HT, CAD, CHF, CRF, DM, and cardiac and in ammatory biomarkers in the serum. The high frequency of PE among these patients, especially in those with CHF and CRF, can also be considered to be associated with pericardial effusion secondary to CHF, independent of COVID-19 cytotoxicity [28]. In some series, a negative correlation was observed between PE and AC involvement and the clinical features, laboratory parameters, morbidity and mortality of the patients . [6] When we evaluated at the characteristics of patients with PE, it is seen that AC involvement is 50% or more in more than half (54%) of patients with effusion. Again, 52.6% of these patients are accompanied by pleural effusion. At the time of admission, 52.4% of the patients with PE developed the need for intensive care, and in-hospital death occurred in 44.8% of all PE patients. The total need for intensive care is 14.6% in patients without PE, and the mortality rate in all patients without PE is 10.7%. Studies have shown a higher incidence of PE in COVID-19 patients with severe and critical illnesses than in non-critical patients [6,16]. In our study, while the in-hospital mortality rate was 56.9% in group 3 patients with PE, it was found as 34.4% in group 3 patients without PE.
As seen in our study, the severity of AC involvement and the presence of conditions such as PE and pleural effusion give us important information on the progression of the disease. In the studies performed, PE frequently accompanies the deterioration that occurs in the laboratory and in the clinic. It has been found to be an important prognostic marker in morbidity and mortality that may develop due to COVID-19 [16]. Although the rate of severe effusion was low in patients with PE, the need for intensive care was seen at the rate of 50% during their follow-up, and the degree of effusion increased in approximately one third of the patients and progressed into a serious effusion.
CT is a useful method in diagnosing the disease, determining the degree of lung involvement and detecting additional pathologies (pulmonary embolism, AC cancer, cardiomegaly, etc.). Especially in our country, easy access, fast method and reproducibility are the most important advantages of CT. Findings that can be evaluated objectively with CT can be used to predict disease progression and mortality. From these ndings; parameters such as the degree of lung involvement, presence of PE and pleural effusion are associated with the severe course of the disease [6,16]. As it is known, the prevalence of AC involvement is the main nding that determines mortality and prognosis in patients with COVID-19. In our study, we found that the presence of PE, as well as the severity of AC involvement and other speci c accompanying ndings, were closely associated with mortality and prognosis. In the regression analysis, we determine that PE may be an independent predictor of mortality in addition to AC involvement and pleural effusion. Conclusion COVID-19 disease has been an important health problem that has affected millions of people around the world, and the epidemic still continues all over the world. As the severity of AC involvement increased and the clinical severity of the disease increased, PE occurred in patients or the degree of PE increased. For this reason, we think that PE should not be ignored when evaluating AC tomography ndings in COVID-19 patients, and it may be an important nding that should be emphasized in making appropriate treatment planning and predicting the course of the disease.

Limitations
First of all, our study was designed retrospectively and the data were obtained from les or electronic records. The diagnosis of PE was made by CT, and TTE data of the majority of these patients were not available. Since some of the patients, except those in need of intensive care, did not have a control CT, PE could not be followed up.  Tables   Table-1