Cardiac tamponade caused by serous pericardial effusion in patients on extracorporeal membrane oxygenation

Journal of Pediatric Surgery SEPTEMBER 1999

VOL 34, NO 9

Cardiac Tamponade Caused by Serous Pericardial Effusion in Patients on Extracorporeal Membrane Oxygenation By

M.S.

Kurian,

E.R.

Reynolds,

R.A.

Detroit,

Michigan

Humes,

and

M.D.

Klein

Purpose:Extracorporeal membrane oxygenation (ECMO) has been successful in the treatment of critically ill children; however, its use has been accompanied by a broad range of complications. The authors describe the presentation, clinical course, treatment, and outcome of 4 patients on ECMO in whom pericardial tamponade developed caused by a serous effusion.

noted, and echocardiography results confirmed a pericardial effusion. The diagnosis was recognized earlier in the course of the subsequent 3 patients. All 4 patients were treated with aspiration of serous fluid from the pericardium with an over-the-needle plastic catheter that was left in place. More than 1 aspiration was required in all cases. All 4 patients survived.

Methods:A retrospective review of patients placed on ECMO at our institution from 1993 to 1997 was performed. The case histories of 4 patients in whom pericardial tamponade developed caused by a serous effusion were reviewed in detail.

Conclusions: The authors have identified a group of ECMO patients with pericardial tamponade caused by serous effusion with good response to treatment. A high index of suspicion and early echocardiography is warranted to confirm the diagnosis in a patient with hypotension on ECMO.

Results: The first patient presented with hypotension while on venovenous (VV) ECMO. The hypotension improved with fluid resuscitation. The patient was converted from (W) to venoarterial (VA) ECMO when hypotension recurred. After a third episode of hypotension, a narrow pulse pressure was

J Pediatr Surg 34:1311-1374. Saunders Company.

T

HE CLINICAL MANIFESTATION and pathophysiology of hemorrhagic pericardial tamponade while on extracorporeal membrane oxygenation (ECMO) have been well described in the literature. Treatment in this population of patients includes percutaneous drainage, per&u-dial window, adjustment of heparin therapy, and platelet transfusion.1,2 Patients with pericar-dial tamponade on ECMO caused by hemorrhagic pericardial effusion have an overall mortality rate of 43%.3

We reviewed 4 of our ECMO patients with pericardial tamponade secondary to serous pericardial effusion and describe their clinical presentation and management. Serous effusions have not been described previously in the ECMO population. Unlike patients with hemorrhagic effusions, our patients with serous cardiac tamponade required pericardiocentesis alone as effective management while on ECMO.

Journal

of Pediatric

Surgery,

Vol

INDEX WORDS: Cardiac extracorporeal membrane

MATERIALS

Copyright

tamponade, oxygenation.

AND

o 1999 by W.B.

pericardial

effusion,

METHODS

We reviewed the records of the 206 patients placed on ECMO between January 1993 and October 1997 at the Children’s Hospital of Michigan, and those who had serous pericardial tamponade were selected. During this same period, some patients placed on ECMO after cardiac surgery deve1oped.mediastina.l tamponade related to bleeding, but pericardial effusions developed in only 2, one before cannulation, and one between 2 ECMO runs. These 2 patients are not included. No patients had purulent pericardial effusions. Pericardiocentesis was performed from a subxiphoid approach directing an over-the-needle catheter toward the left shoulder. When an

From the Departments of Surgery and Pediatrics (Cardiology), Wayne State University School of Medicine and the Children’s Hospital of Michigan, Detroit Medical Centel: Detroit, MI. Address reprint requests to Michael D. Klein, MD, Department of Surgery, Children k Hospital of Michigan, 3901 Beaubien Blvd, Detroit, MI 48201. Copyright 6 1999 by WB. Saunders Company 0022-3468/99/3409-0001$03.00/O

1312

KURIAN

eter for loculated serosanguinous effusions. Subsequently, serial echocardiograms were performed that showed slow reaccumulation of the effusion without significant ventricular compromise. The patient was decannulated on ECMO day 14, and a pericardial window was performed. The patient was extubated 12 days later and discharged home on room air 25 days after ECMO. Patient 2 is a 39-week-gestation girl with a birth weight of 4,070 g. She was the product of an uncomplicated pregnancy and had respiratory distress secondary to PPHN during the first hours of life. The initial echocardiogram documented normal anatomy without pericardial effusion. The patient was placed on VA ECMO at 41 hours of age. After 53 hours on ECMO, the patient had a narrow pulse pressure, loss of pulse oximetry, tachycardia, decrease in venous return, and an increase in Paoz. Volume resuscitation, including FFP, was initiated while the echocardiogram was obtained to confirm the clinical diagnosis of cardiac tamponade. Pericardiocentesis was performed with aspiration of 25 rnL of serous fluid, and an over-the-needle catheter was left in place. Follow-up echocardiograms showed no recurrence of the effusion and normal ventricular function. The patient was decannulated after 6% days on ECMO without further complications, and the pericardial catheter was removed the next day. The patient was extubated after 5 days and discharged home on room air 13 days after ECMO. Patient 3 is a 13-month-old child in whom respiratory

indwelling catheter was necessary, the needle was withdrawn and the catheter left in place as when obtaining venous access. When echocardiography was immediately available, an anterior transthoracic approach was used with echocardiographic guidance.

CASE HISTORIES

The clinical presentation and course of the 4 patients are summarized in Table 1. Patient 1 is a 39-weekgestation boy with a birth weight of 3,430 g. He had respiratory distress at birth with the initial diagnoses of persistent pulmonary hypertension of the newborn (PPHN) and meconium aspiration syndrome (MAS). Echocardiogram showed no major cardiac anomalies and no effusion. He was placed on venovenous (VV) ECMO several hours after birth. After 30 hours on ECMO, the patient had an abrupt drop in blood pressure, which was treated initially with volume expansion, including fresh frozen plasma (FFP). The patient was then converted to venoarterial (VA) ECMO. Hypotension persisted; however, unresponsive to volume resuscitation. Also noted were tachycardia, loss of pulse oximetry, narrowed pulse pressure, decrease in venous return to the circuit, and increasing PaoZ. An echocardiogram showed a large global pericardial effusion with tamponade. Pericardiocentesis was performed using an over-the-needle catheter with the removal of 90 mL of serous fluid, and the catheter was left in place. Within the next week, the patient underwent 3 more aspirations and one exchange under echocardiogram guidance of the indwelling cathTable 1. Patients Patient No.

1

With

Serous

Pericardial

Effusion

Newborn MAS

ECMO day 2 Lowered blood

VV ECMO

Lowered Lowered

Raised Newborn

pressure

VA ECMO

4

Raised heart rate Raised Paol ECMO day 12

with

Newborn PPHN Bilateral hydronephrosis VA ECMO

discontinuation

Pericardial

window

Survive

drain

after decannulation

heart Paol CVP

expansion

Pericardiocentesis Percutaneous pericardial

Volume

Lowered blood pressure Lowered pulse pressure Lost pulse oximetry signal Raised Raised Raised

Pericardiocentesis Percutaneous pericardial

Volume pressure

Lowered venous return Lost pulse oximetry signal

13 mo Near drowning VA ECMO Bleeding treated of heparin

signal

Results

Volume expansion Conversion to VA ECMO

Pao,

ECMO day 3 Lowered pulse

PPHN

3

on ECMO lnterventlons

pulse pressure venous return

Lost pulse oximetry Raised heart rate 2

and Tamponade

Presentation

Age and Diagnoses

ET AL

Survive drain

expansion

Pericardiocentesis Percutaneous pericardial

Survive drain

rate

ECMO hour 24 Lowered blood pressure Lowered pulse pressure Lost pulse oximetry Raised heart rate Raised Pao,

signal

Pericardiocentesis Percutaneous

Survive pericardial

drain

CARDIAC

TAMPONADE

IN PATIENTS

ON ECMO

1313

distress syndrome developed after a near drowning episode. The patient did not respond to conventional ventilatory support and was placed on VA ECMO 2 days after her initial injury. During her ECMO course, the patient had excessive bleeding (from the neck cannulation site and a previously placed chest tube) despite periodic discontinuation of heparin and required 2 circuit changes for clotting. On day 12 of ECMO, tachycardia and hypotension unresponsive to volume resuscitation, which included FFP developed. In addition, there was loss of pulse oximetry, narrowed pulse pressure, increased Pao,, and increased central venous pressure. An echocardiogram confirmed the diagnosis of cardiac tamponade (Fig 1). Initial pericardiocentesis was performed with removal of 25 mL of serous fluid. The patient had immediate improvement in her hemodynamics. However, the patient had a similar episode 2 days later, and repeat aspiration was performed under echocardiogram guidance. Only 1 mL of serosanguinous fluid was removed, and an over-the-needle catheter was left in place. The patient’s hemodynamics further improved with volume expansion. The patient was decannulated on day 22 of ECMO, and the pericardial catheter was removed subsequently. The patient was extubated 5 days after decannulation and discharged home on room air 21 days after ECMO. Patient 4 is a 35-week-gestation boy with a birth weight of 2,580 g. The pregnancy was complicated by oligohydramnios and a prenatal diagnosis of bilateral hydronephrosis. The patient had PPHN at birth requiring intubation and maximal ventilatory support. Postnatal ultrasound scan confirmed bilateral hydronephrosis. The patient was presumed to have pulmonary hypoplasia related to oligohydramnios and posterior urethral valves. On the sixth day of life, the patient was placed on VA ECMO. Initial echocardiogram showed no effusion. The

Fig 1. either

Echocardiograms

ventricle.

of patient

3. (Left)

An apical

4-chamber

view

patient had bleeding noted from his endotracheal tube and was placed on epsilon aminocaproic acid (Am&r, Immunex Carp, Seattle, WA), maintained at activated clotting times (ACTS) of 180 to 200 seconds and platelet counts greater than 120,000/mm3, and received FFP (10 mL/kg every 6 hours). After 24 hours of ECMO, the patient had a narrowed pulse pressure with a blood pressure of 66/59. The patient also had tachycardia, loss of pulse oximetry, and increased arterial Pao,. A confirmatory echocardiogram was obtained, and a pericardiocentesis removed 14 mL of serous fluid. An over-the-needle catheter was left in place but removed after 2 days. On the next day, the patient had recurrence of cardiac tamponade, and another pericardiocentesis was performed. A pericardial catheter was left in place with serosanguinous drainage. This patient had a long and complicated ECMO course but no further episodes of cardiac tamponade. Serial echocardiograms did not show reaccumulation of the effusion, The pericardial catheter was removed on ECMO day 43, and the patient was decannulated after 48 days of ECMO. The patient was extubated 20 days after decannulation and discharged home on oxygen 41 days after ECMO. DISCUSSION

All patients manifested the following clinical signs of cardiac tamponade on ECMO: tachycardia. narrow pulse pressure, loss of pulse oximetry, and increased Paoz. Two patients had decreased venous return to the circuit, and 3 patients had hypotension. Initial pericardiocentesis was therapeutic in all cases. All patients received FFP before pericardiocentesis. Three of the 4 patients were treated definitively with pericardial drainage alone. One patient required a pericardial window at the time of decannulation. All 4 patients

before

ECMO.

The cardiac

chamber

sizes

are normal.

(Right)

An apical

1314

KURIAN

were weaned successfully from ECMO and are shortterm survivors (11 months to 4% years). There are many case reports of cardiac tamponade occurring while on ECMO. All documented patients have had hemorrhagic effusions, and the management of these patients has required pericardiocentesis or pericardial window. Most cases are a result of previous intervention such as stemotomy, pericardiocentesis, and cannula injury to the vessels.1,3-8On review of the Extracorporeal Life Support Organization (ELSO) registry data, a 43% mortality rate occurs in this group of patients.3 Patients with myocardial stun (a more common complication of ECMO) also have a narrowed pulse pressure. Myocardial stun, however, occurs most commonly immediately after initiating ECMO, and is not associated with hypotension, increasing Pao,, and tachycardia.9 Clinical suspicion of cardiac tamponade must be high. Although all of our patients had an echocardiogram for confirmation, our diagnosis was made on clinical grounds. All of our patients received FFP before undergoing

ET AL

pericardiocentesis, and our patients had no bleeding problems related to the procedure. We have considered routine echocardiography in ECMO patients, because we could also monitor pulmonary artery hypertension in this manner. Weaning from ECMO, however, is still done based on blood gases, mixed venous oxygen saturation, and transcutaneous oxygen saturation. We doubt that routine echocardiography would be cost effective if done for surveillance for pericardial effusion. The clinical signs and symptoms are clear enough that echocardiography can be done when hypotension, decreased pulse pressure, and increased PaoZ are recognized. Once a serous pericardial tamponade is identified on ECMO, adequate resolution should be obtained with pericardiocentesis alone. The likelihood of recurrence is high, so a catheter should be left in place. Most importantly, if serous per&dial tamponade is recognized and treated, it can be expected to have little effect on outcome.

REFERENCES 1. Zwischenberger JB, Cilley RE, Hirsch1 RB, et al: Life-threatening intrathoracic complications during treatment with extracorporeal membrane oxygenation. .I Pediatr Surg 23:599-604,1988 2. Medary I, Steinherz LJ, Aronson DC, et al: Cardiac tamponade in the pediatric oncology population: Treatment by percutaneous catheter drainage. J Pediatr Surg 31:197-200,1996 3. Meehan JJ, Laney D, Georgeson KE, et al: Cardiac tamponade during extracorporeal life support. Abstract, ELSO Meeting, Detroit, MI, September 1997 4. Glauber M, Szefner J, Seti M, et al: Reduction of haemorrhagic complications during mechanically assisted circulation with the use of a multi-system anticoagulation protocol. Int J Artif Organs 18:649-655, 1995 5. Alexi-Meskishvili V, Weng Y, Uhlemann F, et ah Prolonged open

sternotomy after pediatric open heart operation: experience with 113 patients. Ann Thorac Surg 59:379-383, 1995 6. Hulyalkar AR, Watkins L, Reitz BA, et al: Transesophageal echocardiographic diagnosis of covert cardiac tamponade during extracorporeal membrane oxygenation in an adult. J Thorac Cardiovasc Surg 104:1756-1757,1992 7. Rhine WD, Hartman GE, Schochat SJ, et al: Hemopericardium from coronary artery laceration complicating extracorporeal membrane oxygenation. J Perinatol 17:189-192,1997 8. Evans MJ, McKeever PA, Pearson GA, et al: Pathological complications of non-survivors of newborn extracorporeal membrane oxygenation. Arch Dis Child 71:F88-92, 1994 9. Martin GR, Short BL, Abbott C, et al: Cardiac stun in infants undergoing extracorporeal membrane oxygenation. J Thorac Cardiovast Surg 101:607-611, 1991