Pediatr Cardiol 15:27-29, 1994

Pediatric Cardiology

Case Reports Inability to V i e w the Heart Through the Subxiphoid Echocardiographic Window: A Harbinger of Disaster N.L. Allgood, J.R. Brownlee, and G.A. Green Department of Pediatrics, USAF Medical Center Keesler, Keesler Air Force Base, Mississippi, USA

SUMMARY. Two case reports of critically ill neonates, one with pneumomediastinum, the other with pneumopericardium, illustrate how the inability to view the heart through the subxiphoid echocardiographic window occurs only in pneumopericardium. This echocardiographic nonfinding should alert the echocardiographer and neonatal team of impending pericardial tamponade.

KEY WORDS: Echocardiography - - Pneumomediastinum - - Pneumopericardium

Neonates with severe hyaline membrane disease, sepsis, or meconium aspiration syndrome are likely to develop intrathoracic air leaks secondary to positive pressure ventilation [2, 5]. The radiographic differentiation between pneumomediastinum and pneumopericardium can be particularly difficult in some cases. The following two cases illustrate how echocardiography, often used to evaluate these infants for cardiac defects, may provide clues to the location of the intrathoracic air. Furthermore, the second case underlines the need for increased suspicion of pneumopericardium and impending disaster when the heart cannot be visualized through the subxiphoid window.

(Fig. 1). This was felt to be consistent with pneumomediastinum. In light of the infant's progressively worsening cardiorespiratory status, an echocardiogram was performed at approximately 12 h of age to evaluate for possible confounding congenital heart disease. An echocardiogram performed after appearance of the pneumomediastinum showed clear visualization of the heart, possible only from the subxiphoid view. Parasternal and apical views were not obtainable because of air artifact. A very small pericardial effusion was present with no evidence of deformation of the cardiac cavities suggestive of cardiac tamponade, and no structural abnormalities were detected otherwise. Follow-up xrays at 24 h of age showed worsening pulmonary interstitial emphysema, but no enlargement of the pneumomediastinum and no change in the cardiac silhouette. The patient died a few hours later from his severe respiratory distress syndrome.

Case Reports

Case 2

Case I

A 32-week gestation male infant with Group B Streptococcal sepsis required intubation at birth and subsequent ventilatory support with very high pressure. On the second day of life, he was noted to have an intermittent murmur, palmar pulses, and a widening of his pulse pressure. An echocardiogram demonstrated bidirectional flow through at patent ductus arteriosus. Standard parasternal, apical, and subxiphoid views all gave clear unimpeded cardiac imaging. The patient was then treated with one dose of indomethacin. The clinical signs of a patent ductus arteriosus resolved, but the infant's cardiorespiratory status continued to deteriorate. An echocardiogram was therefore attempted to confirm ductal closure. At that time, the heart could not be visualized from any echocardiographic window including the subxiphoid view. Although a chest x-ray performed a few hours prior to the echocardiogram showed no evidence of air leaks, a chest x-ray obtained immediately following the echocardiogram demonstrated a large pneumopericardium (Fig. 2). At that time, the patient had a stable blood pressure and heart rate with a normal pulse pressure. However, the pulse pressure be-

A 27-week gestation male infant born prematurely due to a placental abruption presented with severe respiratory distress syndrome requiring ventilatory support with very high pressures. He rapidly developed pulmonary interstitial emphysema and bilateral pneumothoraces. An x-ray following bilateral chest tube placement showed not only resolution of the pneumothoraces, but also the development of a thin radiolucency along the left cardiac border with extension into the superior mediastinum The views expressed in this article are those of the authors and do not reflect the official policy or position of the Air Force, the Department of Defense, or the United States Government. Address offprint requests to: USAF Medical Center Keesler, Department of Pediatrics, c/o John R. Brownlee, MD, Keesler AFB, MS 39534-5300, USA.

28

Pediatric Cardiology Vol. 15, No. 1, 1994

Fig. 2. Pneumopericardium in case 2: At the time of this radiograph, the patient had clinical evidence of pericardial tamponade. Note that the cardiac interface with the diaphragm is interrupted.

Fig. 1. Pneumomediastinum in case 1: Radiolucency extends to the diaphragmatic-pericardial reflection, but does not interrupt it.

gan to narrow within the next few minutes and an emergency pericardiocentesis was performed which yielded 15 ml of air. The pneumopericardium immediately returned and therefore a pericardial tube was surgically placed. An echocardiogram subsequently showed good visualization of the heart utilizing all the standard echocardiography windows. No further clinical or radiographic evidence of pericardial air was detected. The infant died approximately 12 h later from cardiac dysrhythmia believed to be secondary to cardiomyopathy from his sepsis.

Discussion

Pneumomediastinum and pneumopericardium are complications of many of the same processes, including positive pressure ventilation, trauma, and asthma exacerbations. Considering the differing clinical significance and therapeutic management, it is important to know which of these two types of air leaks is present. A pneumomediastinum requires close observation for signs of progression to a

pneumothorax, but requires no intervention [7]. A pneumopericardium in children, especially neonates, however, is well known for producing cardiac tamponade [2, 3]. Intervention is therefore urgently needed in children and in some adults with pneumopericardium, making correct differentiation of the two entities crucial [5]. Radiographic evaluation for a pneumopericardium versus a pneumomediastinum is well described and relies largely upon two observations. Air within the pericardial sac moves with repositioning of the patient, but is limited by reflection of the pericardium on the great vessels. It often dissects superiorly above the level of the pericardial reflection, especially in a seated patient, although not always seen in the supine patient. In the critical care setting, a supine AP chest x-ray is the most common study performed for assessing intrathoracic air leads. Repositioning may not be possible, especially in neonates. The usual guidelines above may not clearly differentiate between pneumomediastinum and pneumopericardium. Chest x-ray findings of the halo sign or of a falling cardiothoracic ratio in tandem with deterioration of hemodynamic status is suggestive of a pneumopericardium [4]. However, since both of these findings are consistent with a large volume of air in the pericardial sac, by that point, there are usually obvious coexistent clinical findings of an evolving cardiac tamponade, such as a decrease in pulse pressure, cyanosis, and decreasing blood pressure. Additionally, since the same etiologies exist for both a pneumopericardium and a pneumomediastinum, both could easily be present within the same patient. The appli-

Allgood et al.: Visualization Through the Subxiphoid Window

cation of echocardiography in the diagnosis of a pneumopericardium and a pneumomediastinum is very limited because of the inability to obtain quality images through an air interface. An "air gap" sign has been described by echocardiography with both a pneumomediastinum and a pneumopericardium [5]. These investigators observed a cyclical appearance with either systole or held inspiration of a dense band of echos anteriorly and total loss of echos posteriorly when either a pneumomediastinum or a pneumopericardium was present. No echocardigraphic guidelines have been described, however, for the differentiation of these two entities. In these two cases, the echocardiographic difference of visualization of the heart only by the subxiphoid view with a pneumomediastinum compared with nonvisualization of the heart at all echocardiographic windows in the case of a pneumopericardium was clearly noted. This difference can best be understood by reviewing the pericardial reflections and attachments within the mediastinum. The pericardium is a two-layered structure consisting of an outer fibrous and an inner serous layer. Both attach superiorly to the great vessels approximately 2-3 cm from the origin of these vessels. Anteriorly, it attaches to the manubrium of the sternum by the superior pericardial ligament. Inferiorly, the pericardium is attached directly to the central tendon and muscular part of the left side of the dome of the diaphragm [10]. A large potential space is present anteriorly and bilaterally for mediastinal air. Some observers have noted that a potential space exists behind the pericardium which extends inferiorly to the posterior reflection of the pericardium with the diaphragm. Varying amounts of air in this location can occur and can produce the appearance of continuity of the left and right hemidiaphragms [1, 91. However, even in the presence of a collection of mediastinal air surrounding the heart in both the anterior and lateral aspects, the contact between the pericardium and the diaphragm should remain. The subxiphoid view would therefore allow for echocardiographic visualization of the heart in these types of pneumomediastina. This view may, however, be obscured by such confounders as free air in the abdomen or a large gastric bubble. In the rarer case of retrocardiac pneumomediastinum, the posterior aspect of the heart may be difficult to visualize because of the air dissecting between the diaphragm and the posterioinferior portion of the pericardial sac. The anterior portion of the pericardium should remain adhered to the diaphragm and thus

29

provide an echocardiographic window (although this is untested by these authors). Occasionally, small amounts of air are found within the pericardium (e.g., the immediate postoperative cardiac patient). When this occurs, the echocardiographic view may only be partially obscured. This was previously observed by the investigators who described the "air gap" sign [8]. Larger collections of air within the pericardium, can completely obstruct visualization of the heart on echocardiogram, as case 2 illustrates. In our experience, using 128-element phase-array transducers in the Accuson 128XP and HP 1000 Sonos systems, the subxiphoid view is obtainable to some degree in all pediatric patients, from premature infant to obese teenager. Although the complete clinical and radiographic findings must be taken into consideration when there is suspicion of intrathoracic air leaks, the inability to visualize the heart from any echocardiographic window should alert the examiner to the probability that a potentially life-threatening pneumopericardium is developing. This is especially true in the case of small children and infants in the critical care setting when imaging from the subxiphoid view is usually of excellent quality. References 1. Buckner CB, Harmon HB, PaUin JS (1988) The radiology of abnormal intrathoracic air. Curr Probl Diagn Radiol 17:4371 2. Burt TB, Lester PD (1982) Neonatal pneumopericardium. Radiology 142:8184 3. Emery RW, Foker J, Thompson TR (1984) Neonatal pneumopericardium: a surgical emergency. Ann Thorac Surg 37:128-132 4. Fanaroff AA, Martin RJ (1989) Neonatal-perinatal medicine: diseases o f the fetus and infant. CV Mosby, St Louis, pp 611-612 5. Hodson NA, Truog WE (1989) Critical care o f the newborn, 2nd ed. WB Saunders, Philadelphia, p 64 6. Mirvis SE, Indeck M, Schorr RM, Diaconis JN (1986) Posttraumatic tension pneumopericardium: the "small heart" sign. Radiology 158:663-669 7. Pernoll ML, Benda GI, Babson SG, Simpson K (1986) Diagnosis and management o f the fetus and neonate at risk: a guide for team care. CV Mosby, St Louis, p 265 8. Reid CL, Chandraratna AN, Kawanishi D, et al. (1983) Echocardiographic detection of pneumomediastinum and pneumopericardium: the "air gap" sign. J A m Coll Cardiol 1:916-921 9. Rosenfeld DL, Cordell CE, Jadeja N (1990) Retrocardiac pneumomediastinum: radiologic finding and clinical implications. Pediatrics 85:92-97 10. Williams PL, Warwick R, Dyson M, Bannister LH (1989) Gray's anatomy. Churchhill Livingston, New York, pp 694695, 1267-1272