Delayed hemopericardium after penetrating chest trauma: thoracoscopic pericardial window as a therapeutic option.

Abstract: A 41-year-old male developed a hemothorax after sustaining a stab wound in the right chest. The patient was managed conservatively with thoracostomy tube drainage for 3 days and was subsequently discharged home. Two weeks later the patient returned to the hospital with pleuritic chest pain and shortness of breath. Imaging studies revealed a right-sided pleural effusion and an enlarged cardiac silhouette, which was consistent with pericardial effusion as per ultrasonography. Thoracoscopic exploration revealed an enlarged heart, that following pericardiotomy drained 400 mL of frank blood. Subsequently, cardiac contractility improved, and no further bleeding was evident.

Key Words: hemopericardium, penetrating chest trauma, thoracoscopic pleuropericardial window

**********

The majority of patients suffering penetrating wounds to the heart do not survive long enough to receive any medical assistance. However, among those who reach the hospital, most cardiac injuries are discovered at admission and treated accordingly, whether initially decompressed with a subxiphoid pericardial window, or approached with an open thoracotomy. (1,2)

Infrequently, a penetrating injury to the heart may be missed on initial assessment, the patient returning to the hospital a few weeks later with different degrees of hemopericardium. Delayed hemopericardium after penetrating chest injury has been described in the literature, with the therapeutic approach invariably involving pericardiocentesis or open thoracotomy. (3)

Thoracoscopic pleuropericardial window has been popularized as a way to drain different types of pericardial effusion, with the advantage of better exposure than the traditional subxiphoid pericardial window, but without the morbidity associated with an open thoracotomy. This technique has been increasingly utilized during the last ten years in numerous situations where pericardial drainage is required, including conditions such as inflammatory processes, neoplastic exudates, or postsurgical effusions. (4-7) As of this writing, the use of the thoracoscopic approach has not been reported in the management of delayed hemopericardium.

We report the use of a thoracoscopic pericardial window as a means of draining a delayed hemopericardium after penetrating chest injury. The findings, technique, and outcome are described for this approach.

Case Report

A 41-year-old male was seen in the emergency department after a stab wound to the right chest. At admission the patient was in stable condition, with a chest radiograph positive for hemopneumothorax, and without evidence of cardiac enlargement. A thoracostomy tube was placed in the right hemithorax, and 3 days later the patient was discharged after the chest tube was removed and adequate lung expansion verified.

Two weeks later, the patient returned to the emergency department complaining of increasing right-sided pleuritic chest pain and shortness of breath. Initial assessment revealed bilateral pleural effusions on chest radiograph, predominantly in the right side, as well as an enlarged cardiac silhouette (Fig. 1). A thoracostomy tube was placed in the right chest again and connected to wall suction, draining 300 mL of serosanguineous fluid upon insertion.

Further imaging studies included a 2-D echocardiogram, which was positive for pericardial effusion. A computed tomographic scan of the chest showed bilateral pleural effusions and fluid around the pericardium with an attenuation of 30 Hounsfield units, number suggesting the presence of blood or purulent fluid (Fig. 2). The patient was taken to the operating room for thoracoscopic exploration, with the presumptive diagnosis of bilateral loculated hematomas and associated hemopericardium.

It is worth mentioning that during the first admission, pericardial ultrasound was not performed on the patient, since at that point it was not yet readily available in the emergency department.

The operation was performed under general anesthesia with double-lumen orotracheal intubation. The patient was placed in the right lateral position and draped in the standard fashion as for a formal thoracotomy. After deflation of the left lung, a thoracoscope was introduced one finger breadth below the tip of the scapula, next to the posterior axillary line, and between the sixth and seventh intercostal spaces. Full assessment of the left hemithorax was performed, and 200 mL of blood was drained. During inspection, the heart was revealed to be enlarged, suggesting a retained hemopericardium after penetrating injury to the heart. After identifying the phrenic nerve, with the use of dissectors and electrocautery, a 4 cm longitudinal incision was made in the pericardial sac. Subsequent to opening the pericardium, 400 mL of frank blood was drained from the pericardial cavity, with immediate evidence of improved cardiac contraction. The camera was advanced and introduced inside the sac, visualizing sparse clots and no active bleeding evident at that time. After complete inspection of the left hemithorax, anterior and posterior chest tubes were left in place for continuous drainage.

The patient was then placed in the left lateral position to approach the right hemithorax. Access was gained following the same landmarks used for the left chest, and with selective deflation of the left lung. Full inspection of the right hemithorax revealed sparse adhesions, and 400 mL of retained blood was removed. The adhesions were taken down, the chest cavity irrigated, and a chest tube left in place.

The patient tolerated the procedure and was extubated on the first postoperative day. With drainage progressively decreasing, the thoracostomy tubes were removed four days later. Chest films revealed no reaccumulation of pleural or pericardial effusions. The patient was finally discharged with no major complaints, and 8 months after surgical intervention remains asymptomatic.

Discussion

Few cases of delayed hemopericardium following either penetrating or blunt chest trauma have been described in the literature. Delayed hemopericardium secondary to blunt trauma has been reported primarily in the pediatric population. Taylor (8) recently reported a case of delayed hemopericardium in a 10-month-old infant, which was missed in the initial admission. Ultrasonography was not performed in this instance. (3,9-11)

In 1991, Aaland (3) reported a case of delayed pericardial tamponade following penetrating chest injury, and a literature review yielded seven similar cases reported since 1950. No other comparable cases have been reported, except for one patient developing a delayed pericardial tamponade following a pacemaker insertion (12) and a second one induced by an acupuncture needle. (13)

Invariably, all of the cases which involved penetrating injury have been treated with open thoracotomy, except for one reported in 1953, which was managed with repeated pericardiocentesis. (14) We present a similar case of delayed hemopericardium after penetrating injury that was approached thoracoscopically, with bilateral evacuation of hemothoraces and the creation of a pleuropericardial window.

The practice of thoracoscopic pericardial window was initially described by Vogel (15) in 1990. This method was applied in four patients with malignant pericardial effusion, permanently eliminating the cardiac tamponade. Since then, this technique has been increasingly used in other causes of pericardial effusion, such as inflammatory or neoplastic disease, (4,6) pericardial tamponade following coronary artery bypass graft or heart transplantation, (7,16) and in the acute setting for the assessment of cardiac injury after penetrating trauma. (5)

Traditionally, pericardial effusions have been managed with two different approaches, subxiphoid pericardial window, and open thoracotomy. Subxiphoid pericardial window has the advantage of being a simple procedure, allowing quick drainage of pericardial tamponade. However, due to limited exposure, this technique may result in inadequate pericardial resection, and therefore be less suitable for cases susceptible to relapse, which has been reported at between 3 and 18% depending on the etiology. (4) On the other hand, an open thoracotomy prevents these shortcomings; however, it is also more frequently associated with pulmonary complications and a prolonged postoperative recovery.

[FIGURE 1 OMITTED]

Videothoracoscopy provides a thorough exposure of the thoracic cavity, allowing the simultaneous management of pericardial, pleural, and pulmonary disorders, avoiding the increased morbidity related to an open thoracotomy. In addition, this technique allows a more extensive pericardial resection, tissue sampling, and the ability to fashion a pleuropericardial window if needed.

Mack (17) reported no complications in his review of 45 cases of pericardial effusion managed with this technique. Nataf (4) and Liu (6) reported similar results in their series (22 and 28 patients, respectively) with no perioperative or postoperative complications. In all these studies, the thoracoscopic approach was used to assess pericardial effusions of different types, although hemopericardium following trauma was not among them.

Thoracoscopy, even when it offers a thorough inspection of the heart surface following pericardiotomy, also has limitations. Inspection of the posterior surface of the heart is not easily accomplished. Similarly, detail in the assessment of an injury to the coronary arteries may also be limited with this approach. However, in the setting of a delayed hemopericardium, without signs of an ischemic insult to the heart and without active exsanguination following pericardiotomy, such issues would not have much relevance for the thoracoscopic approach.

[FIGURE 2 OMITTED]

Another therapeutic approach described recently is the laparoscopic transdiaphragmatic pericardial window. (18) This technique has been used in cases of thoracoabdominal injuries, when the upper abdomen and lower chest are at risk for both intraabdominal and cardiac injury. Proponents of this approach claim that it can be rapidly performed in the stable patient with no need of specialized airway management. In the case here discussed, thoracoscopy offers the advantage of a complete assessment of the thoracic cavity for associated injuries, and the opportunity to execute treatment accordingly.

Thoracoscopy has also been used in the acute setting of penetrating chest trauma. Morales (5) reported on 108 patients in whom a thoracoscopic pericardial window was used as a means of diagnosing cardiac injury in cases of penetrating chest trauma in proximity to the heart when no signs of cardiac trauma were evident, and the patient was in stable condition.

In Morales' study, thoracoscopy identified hemopericardium in 30% of the pericardial windows performed, all of them being followed by immediate open thoracotomy. There was no mortality or morbidity associated with this procedure. This study group included only thoracic injuries in the acute setting. In the current report, pericardial ultrasound was not used as part of the diagnostic workup since it was not available on an emergency basis in the facility where the procedure was performed.

Pericardial ultrasound has, for the most part, replaced pericardiocentesis and pericardial window in the emergency setting in level I trauma centers. It has been shown to have very high sensitivity and specificity for the diagnosis of fluid accumulation around the pericardium. Thourani, (19) in his report of a 22-year experience of penetrating cardiac trauma, suggested that surgeon-performed ultrasound of the pericardium should improve survival in normotensive or mildly normotensive patients, even though the overall survival for such injuries did not change during the study interval.

Rozycki (20) reported a sensitivity of between 96.9 and 100% for pericardial ultrasound in the setting of penetrating thoracic trauma. In this study, 261 patients with penetrating truncal injuries, from five different level I trauma centers, underwent pericardial ultrasound examination as part of the initial diagnostic workup. Even though false negative results were not found in this study, other authors have cautioned that false-negative readings may occur. In the presence of large hemothorax, subcutaneous emphysema, lacerations over the proposed windows, and multiple chest tubes, it has been reported that the presence of hemopericardium may be obscured, which could in turn lead to indeterminate or false negative results. (21)

Even though there is no doubt that ultrasound is a very useful adjunct in the setting of acute penetrating thoracic trauma, the majority of large scale studies assessing the efficacy of pericardial ultrasound in the emergency setting come from level I trauma centers with readily available units, and personnel properly trained in the use of such resources. In the setting of less well-structured systems without the appropriate training, it is hard to believe that sensitivities in the range of 100% would be found, as it was reported by Rozycki. It is also noteworthy to mention that all these studies lack follow-up of the patient after discharge, since patients with negative studies were admitted for observation and finally discharged without any prospective follow-up. As a result of such deficiencies, these studies could possibly have failed to identify a very small group of patients with false-negative results who may have developed a subsequent event secondary to a delayed hemopericardium.

Unfortunately, pericardial ultrasound was not initially performed on our patient, since it was not readily available on an emergency basis. Even though this situation is hardly ever the case in high-volume trauma centers, this is still a relatively frequent issue in the smaller community hospitals that often deal with penetrating thoracic injuries. In the case presented, as well as in previous reports of delayed hemopericardium in the literature, the hemopericardium was most likely present at the original injury. Such an injury may be missed in the initial assessment, given the more insidious and asymptomatic nature of the presentation. The purpose of this report is to stress the fact that, should this unusual situation occur, it is feasible to be dealt with through a thoracoscopic approach.

The incidence of missed injuries to the heart in the setting of penetrating thoracic injuries would be very difficult to estimate accurately. As noted previously, ultrasonography is a very useful adjunct in acute trauma; however, it is operator-dependent, and adequate training is essential for its efficacy. Furthermore, the lack of adequate follow-up of patients after discharge from trauma centers makes such outcomes even more troublesome to estimate.

Another interesting point in this case is the presence of bilateral hemothoraces. Even though only 200 mL of blood was drained from the left hemithorax, the origin of this injury was a penetrating wound to the right chest. It could have been caused by delayed rupture of the pericardium into the left chest, or a tangential injury to the heart that violated the pleural boundaries of the left hemithorax.

In cases of delayed hemopericardium, the initial heart injury is more likely to have sealed by the time it is discovered, as was the case in our patient, who presented to the hospital two weeks after the original insult. For this reason, it would be safe to assess the pericardial sac thoracoscopically, and to proceed with drainage of the retained blood, since a primary cardiac repair would not be necessary at this time. We think close monitoring of the drainage output and overall condition of the patient is a reasonable option in a case of delayed hemopericardium, reserving open thoracotomy and primary cardiac injury repair in the event of a failure of this initial, conservative management.

Conclusion

In conclusion, this is one of the few cases reported in the literature of delayed hemopericardium following a penetrating injury to the chest. The video-assisted approach reported here represents a reasonable alternative to open techniques for the assessment of this condition. The fashioning of a thoracoscopic pleuropericardial window is a safe therapeutic decision in the setting of a delayed hemopericardium following penetrating chest injury.

If you can count your money, you don't have a billion dollars.
--J. Paul Getty

Accepted December 8, 2003.

Please see Krishnan Raghavendran's editorial on page 921 of this issue.

References

1. Demetriades D, Van der Veen BW. Penetrating injuries of the heart: experience over two years in South Africa. J Trauma 1983;23:1034-1041.

2. Sugg WL, Rea WJ, Ecker RR, et al. Penetrating wounds of the heart: an analysis of 459 cases. J Thorac Cardiovasc Surg 1968;56:531-545.

3. Aaland MO, Sherman RT. Delayed pericardial tamponade in penetrating chest trauma: case report. J Trauma 1991;31:1563-1565.

4. Nataf P, Cacoub P, Regan M, et al. Video-thoracoscopic pericardial window in the diagnosis and treatment of pericardial effusions. Am J Cardiol 1998;82:124-126.

5. Morales CH, Salinas CM, Henao CA, et al. Thoracoscopic pericardial window and penetrating chest trauma. J Trauma 1997;42:273-275.

6. Liu HP, Chang CH, Lin PJ, et al. Thoracoscopic management of effusive pericardial diseases: indications and technique. Ann Thorac Surg 1994;58:1695-1697.

7. Hurley JP, Subarreddy K, McCarthy J, et al. Video-assisted thoracic surgery for delayed pericardial effusion post-CABG. Chest 1994;106:1617-1619.

8. Taylor MW, Garber JC, Boswell WC, et al. Delayed hemopericardium and associated pericardial mass after blunt chest trauma. Am Surg 2003;69:343-345.

9. Cil E, Senkaya I, Tarim O. Delayed hemopericardium due to trivial chest trauma. Cardiology in the young 1998;8:390-392.

10. Bowers P, Harris P, Truesdell S, et al. Delayed hemopericardium and cardiac tamponade after unrecognized chest trauma. Pediatr emerg care 1994;10:222-224.

11. Solomon D. Delayed cardiac tamponade after blunt chest trauma: case report. J Trauma 1991;31:1322-1324.

12. Gershon T, Kuruppu J, Olshaker J. Delayed cardiac tamponade after pacemaker insertion. J Emerg Med 2000;18:355-359.

13. Hasegawa J, Noguchi N. Yamasaki J, et al: Delayed cardiac tamponade and hemothorax induced by an acupuncture needle. Cardiology 1991;78:58-63.

14. Mason LB, Warshauer SE, Williams RW. Stab wound of the heart with delayed hemopericardium. J Thorac Surg 1954;29:524.

15. Vogel B, Mall W. Thoracoscopic pericardial fenestration. Diagnostic and therapeutic aspects. Pneumologie. 1990;44 Suppl 1:184-185.

16. Brandt M. Bruckner M. Hirt SW, et al. Thoracoscopic creation of a pericardial window for recurrent pericardial effusion after heart transplantation. Eur J Cardiothorac Surg 1996;10:220-222.

17. Mack MJ, Landreneau RJ, Hazelrigg SR, et al. Video-thoracoscopic management of benign and malignant pericardial effusions. Chest 1993;103(Suppl):390S-393S.

18. Porter JM. Diagnostic laparoscopy and laparoscopic transdiaphragmatic pericardial window in a patient with an epigastric stab wound: a case report. J Laparoendosc Surg 1996;6:51-54.

19. Thourani VH, Feliciano DV, Cooper WA, et al. Penetrating cardiac trauma at an urban center: a 22-year perspective. Am Surg 1999;65:811-818.

20. Rozycki GS, Feliciano DV, Ochsner MG, et al. The role of ultrasound in patients with possible penetrating cardiac wounds: a prospective multicenter study. J Trauma 1999;46:543-551.

21. Meyer DM, Jessen ME, Grayburn PA. Use of echocardiography to detect occult cardiac injury after penetrating thoracic trauma: a prospective study. J Trauma 1995;39:902-907.

RELATED ARTICLE: Key Points

* Thoracoscopic drainage of delayed hemopericardium following penetrating thoracic trauma is an acceptable therapeutic approach.

* Delayed hemopericardium, even though unusual, is possible in the setting of penetrating thoracic injuries.

* Hemopericardium can be missed, particularly when FAST scan is not used in the initial assessment.

Manuel Caceres, MD, Kennan Buechter, MD, Jaime A. Rodriguez, MD, and Donald Liu, MD, PHD

From the Department of Surgery, Louisiana State University Health Sciences Center, New Orleans, LA, and the Department of Surgery, University of Chicago Printzker School of Medicine, Chicago, IL.

Reprint requests to Manuel Caceres, Louisiana State University School of Medicine, Department of Surgery, 1542 Tulane Avenue, New Orleans, LA 70112. Email: [e-mail]