Echocardiographic Evaluation of Pericardial Effusion and Cardiac Tamponade

Pericardial effusion (PEff) is defined by an increase in the physiological amount of fluid within the pericardial space. It can appear following different medical conditions, mainly related to inflammation and cardiac surgery. Cardiac tamponade is a critical condition that occurs after sudden and/or excessive accumulation of fluid in the pericardial space that restricts appropriate filling of the cardiac chambers disturbing normal hemodynamics and ultimately causing hypotension and cardiac arrest. It is, therefore, a life-threatening condition that must be diagnosed as soon as possible for correct treatment and management. Echocardiographic evaluation of PEff is paramount for timely and appropriate diagnosis and management. A structured echocardiographic approach including two-dimensional, M-mode, and Doppler echocardiographic evaluation assessing (i) quantity and quality of pericardial fluid, (ii) collapse of cardiac chambers, (iii) respiratory variation of the ventricular diameters, (iv) inferior vena cava collapsibility, and (v) flow patterns in atrioventricular valves should give the bedside clinician the necessary information to appropriately manage PEff. Here, we review these key echocardiographic signs that will ensure an appropriate assessment of a patient with PEff and/or cardiac tamponade.

ease of use, availability, costeffectiveness and its comprehensive appraisal of the heart and its hemodynamics.

ANATOMY OF THe PeRiCARDiUM
The pericardium constricts the heart into position, limiting distension of the cardiac chambers and also protecting it from infections within the chest. The outermost layer of the pericar dium is formed by a fibrous envelope composed primarily of col lagen fibers with interspersed short elastic fibrils (8). The fibrous envelope is continuous to the adventia of the great vessels superi orly and attached to the central tendon of the diaphragm inferi orly. It is attached anteriorly to the sternum by sternopericardial ligaments. Posteriorly, it is anatomically related to the bronchi, esophagus, and the thoracic descending aorta. The inner layer of the pericardium is composed of visceral and parietal layers. The space between these two inner layers is the pericardial cavity and normally contains a small amount of fluid (<50 ml, which serves as lubricant between the visceral and the parietal pericardial layers) (9). The visceral layer covers the epicardium, where a variable amount of fat is found. Parietal layer has a vari able thickness (0.7-1.7 mm) as described from echocardiogram, CT, and CMR (10)(11)(12). For this reason, the use of transthoracic echocardiography may not be optimal, due to hyperechogenic signal received from epicardial fat.

Absence of the Pericardium
Absence of the pericardium is a rare condition due to failure in the embryogenesis of the pericardial sac. This anomaly can be classified based on extension (total or partial) and location (left, right, or bilateral) (13). Complete absence of left pericardium sac is most common (9); however, patients are usually asymptomatic and diagnosis often incidental even though symptoms such as chest pain, palpitations, or dyspnea have been described. Partial absence of the pericardium can lead to sudden death due to com pression, herniation, or cardiac strangulation (14). Chest radiog raphy, electrocardiogram, echocardiography, CT, and CMR have proven helpful diagnostic studies. Specifically, echocardiography is useful, since the image of the parasternal longaxis view show ing a posterior and leftward displacement of the apex without its normal round shape is different than the usual one (Figure 1). No intervention is needed if patients are asymptomatic, but surgery is indicated in symptomatic cases or in those at high risk of herniation (partial defects) (15).

PeRiCARDiAL eFFUSiON
Physiology Pericardial effusion is defined by an increase in the amount of fluid within the pericardial space. Etiology is very diverse. Tuberculo sis is the primary cause in developing countries, while viral infec tions and postsurgical complications are the most frequent causes in developed countries (16)(17)(18). Parietal pericardium compli ance follows a nonlinear relationship (Figure 2). Therefore, the effect on the cardiac chambers and resulting clinical symptoms from PEffs will not only depend on the etiology but also on the amount of fluid accumulation and the time since onset. There is an initial flat part of the curve, when the increase of pericardial fluid represents either no or minimum pressure increases within the pericardial space. However, once a threshold point is reached, small changes in fluid volume can cause large pressure variations in the intrapericardial space. Thus, an acute onset, small PEff can cause cardiac tamponade, while chronic and slow PEff may allow the elastic fibrils of the pericardium to stretch so that the volume-pressure curve is shifted to the right, causing minimal effects on the hemodynamics (19). echocardiography Echocardiography should be the initial imaging modality used should PEff be suspected (7,20). It is an innocuous technique, highly reliable, and readily available at the patient's bedside. Due to its versatility, echocardiography is the technique of choice to identify hemodynamic compromise in the setting of PEff.
Pericardial effusion appears as an echofree space between the two layers of the pericardium. Echocardiographic assess ment should be structured and focused to (i) differentiate between global or localized effusion; (ii) quantify the effusion; (iii) describe fluid appearance; and (iv) analyze hemodynamic compromise. Standard views along with twodimensional (2D) echocardiographic, MMode, and Doppler analysis should be routinely used in the assessment of PEff. Four standard views should be used for reproducibility [subcostal, fourchamber, and parasternal long and short axis (Figures 3 and 4)]; however, in emergent situations, a single view (usually the subcostal one) FiGURe 2 | volume-pressure relationship in cardiac tamponade physiology. Rapid onset effusion leads faster to cardiac tamponade, comparing to a slow increase of the pericardial fluid that will allow the pericardium to stretch reaching the critical point with a higher volume of effusion.   is enough to diagnose significant PEff. A combination of the aforementioned four standard views should be used to determine whether the effusion is global or localized. Quantitative assessment of the PEff can be helpful for diag nosis and followup. However, as previously mentioned, the volume of PEff does not necessarily correlate with the clinical symptoms. PEffs in adults are commonly classified according to the size including trivial (seen only in systole), mild (<10 mm), moderate (10-20 mm), and severe (>20 mm) (21). The most common and easy view to reproduce is the parasternal long axis (Figure 3), but it is still important to compare with other standard views (Figure 4), such as the parasternal short axis and the fourchamber view. A subcostal view is also very commonly used, but it is important to know that depending on the angle of the probe, it is easy to overestimate the size of the effusion. At end diastole, echofree space between both pericardial layers can be measured along the posterior wall of the left ventricle (LV). Mmode echocardiography in the parasternal long axis shows an echofree space throughout between the visceral and the parietal pericardium. This sign is only seen in systole in case of mild effusions and during the complete cycle when the effusion is at least moderate.
Since PEff can be of different etiologies (7), it is usually difficult to make an appropriate qualitative assessment of fluid charac teristics just by echocardiographic signal. A qualitative assessment can be made by defining whether it is global or loculated, transudative, exudative, or containing clotted blood or fibrin. Intrapericardial fibrin and pyogenic exudates can create tabications and loculate the effusion. Particular echo genic densities such as the ones seen in complex exudates or  Quantitative and qualitative assessment can be made fol lowing comments made earlier regarding PEff.
(ii) Collapse of cardiac chambers A common sign of cardiac tamponade with significant hemo dynamic compromise is collapse of the right atrium (RA) and the RV. This happens during their relaxation phase, when intrachamber pressures are lower than pericardial pressures (26). Atrial and ventricular collapses are observed at different points of the cardiac cycle. Atrial collapse is usu ally observed before ventricular collapse in the sequence of progressive cardiac tamponade. Right atrium collapse is commonly observed during systole. In early systole (near the peak of the R wave), intracavity pressure is lower and the atrial indentation of the thin free wall is seen. Moreover, duration of atrial collapse (collapse longer than onethird of the cardiac cycle) has been described as an almost 100% sensitive and specific sign of clinical cardiac tamponade (27) (Figure 7). While isolated RA collapse is frequently observed, collapse of the left atrium is, although described, rarely observed as a single chamber col lapse (28). It is usually seen in cardiac tamponade along with collapse of the RA (Figure 8). Collapse of both atria increases the sensitivity and specificity of cardiac tamponade.
Right ventricle collapse is commonly observed in diastole. During early diastole (at the end of the T wave), intracavity pressures are lower (Figure 8). Collapse of the free wall appears with moderate increases in pericardial pressures and will decrease stroke volume, initially without significant changes in the systemic blood pressure due to hemopericardium with the presence of fibrin strands and clots are easily recognized by echocardiography. On the other hand, transudates will usually show a more echofree signal giving a characteristic sinusoid image in Mmode echocardiography that could help differentiate complex exudates from transudates. Hemopericardium can be differentiated because of the sponta neous contrast signal resulting from the swirling of the blood within the pericardial space ( Figure 6). Nevertheless, a complete clinical assessment complemented by laboratory tests and imag ing studies should always be taken into consideration to find the underlying condition (22).

CARDiAC TAMPONADe Physiology
Cardiac tamponade is a lifethreatening condition caused by accumulation of fluid in the pericardial space that compresses the cardiac chambers and restricts them from normal filling (9,23,24). The rate of fluid accumulation will determine the clinical mani festations. As the pericardial space is filled with fluid, increased pressure within the pericardial space will overpower the relaxing pressures of the cardiac chambers, which varies during the respi ratory cycle. During spontaneous inspiration, regardless of pres ence of cardiac tamponade, systemic venous return is increased and filling pressures of the right ventricle (RV) are higher than the filling pressures of LV ( Figure 5) (25). If left ventricular diastolic filling pressures are equal or lower than intrapericardial pressures, the volume the ventricle can accept will be restricted, causing a decrease in the cardiac output. An exaggerated inspiratory drop in the systemic blood pressure (>10 mmHg) is known as pulsus paradoxus and is a typical sign of significant PEff (sensitivity of 80% and specificity of 40%) (26).
As cardiac tamponade becomes more severe, compensatory mechanisms will ultimately fail, preload will be insufficient to sustain cardiac filling, coronary perfusion will be in jeopardy leading to rapid bradycardia and abrupt blood pressure drop as the usual terminal event.
echocardiography Echocardiography should be obtained immediately if cardiac tamponade is suspected. The information obtained from this study should assess (i) quantity and quality of pericardial fluid; (ii) collapsibility of cardiac chambers; (iii) diastolic ventricular    compensatory mechanisms (29). Initially, collapse of the RV free wall will only be present during expiration, but as collapse progresses, detection is possible throughout the respiratory cycle. Duration of collapse of the RV free wall is again an indicator of severity (30). Collapse will last as long as pericardial pressures remain higher than RV filling pres sures. Thus, the longer the indentation on the free wall, the more severe the tamponade. Mmode through the affected wall (always along with appropriate EKG tracing) is useful to assess duration and timing of collapse. Experi mental studies demonstrate that RV diastolic indentation is more sensitive, specific, and has a better predictive value of cardiac tamponade than pulsus paradoxus (26). Collapse of the LV is unusual, due to its thicker wall and may be seen in patients with severe pulmonary hypertension (31) or loculated PEff, typically around the free posterior wall of the LV following cardiac surgery (32) (Figure 9). All these echocardiographic signs of cardiac tamponade may not be present in those cases in which right ventricular Echocardiography in Cardiac Tamponade Frontiers in Pediatrics | www.frontiersin.org April 2017 | Volume 5 | Article 79 diastolic pressures are elevated at baseline such as in pulmo nary hypertension, positive pressure ventilation, severe LV failure, or other cardiac congenital or inherited conditions that would increase RV diastolic pressures (33,34). On the other hand, collapse of the right chambers may occur earlier than expected in those conditions with reduced baseline intracavitary pressures, such as hypovolemia (35). Even with these limitations, absence of collapse of any cardiac chamber has a 90% negative predictive value (36).
(iii) Diastolic ventricular size variability with respiratory cycle When cardiac tamponade is present, Mmode in both parasternal long and short axes is a useful tool to assess exaggerated ventricular interdependence with the respira tory cycle. During inspiration, RV filling is increased, while LV size in diastole decreases (see Figure 10). The opposite scenario is present during expiration. This is a physiological phenomenon that can cause a variation of no more than 5% of the cardiac output in absence of tamponade (10).
(iv) Assessment of IVC An important sign of tamponade seen in 2D echocardiogra phy is dilatation of the IVC (>20 mm in an adult size heart) and hepatic veins (Figure 11). This is known as IVC plethora, and, although not very specific, it is a very sensitive sign of cardiac tamponade (92%). A decrease of the physiological collapsibility of the IVC during inspira tion is the commonly observed sign. Mmode through the IVC will demonstrate a less than a 50% reduction of caliber in most cases of significant PEff (Figure 11) (37). Echocardiography in Cardiac Tamponade Frontiers in Pediatrics | www.frontiersin.org April 2017 | Volume 5 | Article 79 (v) Septal "bounce" An inspiratory "bounce" of the interventricular septum toward the LV is a common, but not specific finding in cardiac tamponade (Figure 12). An Mmode through the parasternal long axis will show this abnormal movement of the interventricular septum. It may not be present in case of LV hypertrophy or increased LV filling pressures.

Doppler Echocardiography
Pulsedwave Doppler is key in the assessment of respiratory varia tions of the transvalvular flows in patients with cardiac tamponade (32). In spontaneously breathing healthy individuals, Doppler peak velocities in the tricuspid and pulmonary valves increase during inspiration, while they decrease in the mitral and aortic valves. As previously mentioned in this review, a physiological variation of the cardiac output of around 5% exists with normal respiratory cycle, which is reflected in normal peak Ewave variation of up to 10% (7). In cardiac tamponade, these variations in the Doppler veloci ties are larger.
1. Mitral valve: research studies demonstrate an inspiratory reduction in mitral peak Ewave velocity in cardiac tamponade of at least 25% (32,33,36). The value of the Ewave peak will vary based on the phase of the respiratory cycle at the time of early diastole (lower filling pressures) (Figure 13). 2. Tricuspid valve: peak Ewave tricuspid valve physiological variation is larger than the mitral valve fluctuations. In cardiac tamponade, the peak Ewave velocity will drop at least 40% in expiration compared to inspiration (changes seen mainly in first two heart beats of expiration) (Figure 13).

Right and left ventricular outflow tracts: physiological
variations in aorta and pulmonary trunk are typically less than 10%. In presence of hemodynamically significant PEff, peak velocities in both outflow tracts will show a larger difference with respiratory cycle. During inspiration a drop of >10% of the peak velocity will be seen in the aorta (Figure 14), while the opposite will happen in the right ventricular outflow tract, where an increase of at least 10% will be noticed (38). 4. Hepatic veins: normal flow pattern shows a biphasic wave fol lowed by a reversal wave corresponding with atrial contraction (see Figure 15). In cardiac tamponade, while venous return will still increase with inspiration, decreased or even reversal flow will be seen in diastole (before atrial contraction). High positive and negative predictive values have been reported for changes in Doppler signal in hepatic veins (82 and 88%, respectively) (36). Nevertheless, this has some limitations since hepatic vein flow is difficult to assess in up to onethird of patients. 5. Pulmonary veins: pulmonary vein flow shows a biphasic wave with reversal flow during atrial contraction. During inspira tion, D wave is smaller due to reduced left ventricular filling (increased LV filling pressures). In expiration, S wave is larger following decrease in LV filling pressures (increased flow toward the ventricle).

The verdict
Echocardiography is the technique of choice for the assessment of PEff. Nevertheless, as stated in this review, published data continue to reveal a large variability in specificity and sensitivity of the typical echocardiographic findings in patients with signifi cant PEff. Of all assessments, the absence of collapse of any of the cardiac chambers seems to be the echocardiographic sign with a higher and more consistent negative predictive value to rule out cardiac tamponade. Unfortunately, although very helpful, none of these echocardiographic findings is 100% diagnostic of this critical condition. Thus, final decision and management of PEff should always be made following a detailed physical exami nation and a proper clinical assessment.

ULTRASOUND-GUiDeD PeRiCARDiOCeNTeSiS
When clinical findings and echocardiography signs of cardiac tamponade are present, pericardiocentesis must be performed without delay. Ultrasound is useful before, during and after the procedure. Its use has shown to reduce the incidence of major complications, reported as 3%, comparing to the blind subxi phoid approach that has an associated incidence of complications of 5-25% (39)(40)(41). Assessment of the effusion must be done from standard views to find the largest area of effusion closest to the transducer avoiding vital structures. Several approaches have been used, and a longaxis inplaneguided insertion using a linear probe has been recently validated as an easy and safe method to perform the pericardiocentesis in small children (42). Access from the anterior chest wall, typically subcostal, is the most common point of incision. Parasternal left access (between fifth and sixth intercostal spaces) can also be used, but that carries a higher risk of pneumothorax. With either approach, Seldinger technique is typically used and placement of a pigtail drain is usually recommended. During the whole procedure, correct positioning of the wire and the catheter should be monitored with echocardiography (43).

AUTHOR CONTRiBUTiONS
APC: primary author, writing the bulk of the manuscript; image acquisition. LBG: image acquisition and image prepara tion, writing sections. SC: image acquisition and image edition, critical revision of the manuscript, and writing sections. JSdT: drafting of the manuscript, image acquisition, editing, critical revision of the manuscript, and responsible for the overall content.