Ann Thorac Surg 2004;78:1723-1727
© 2004 The Society of Thoracic Surgeons
Original article: cardiovascular
Venovenous Extracorporeal Membrane Oxygenation for Cyanotic Congenital Heart Disease
Michiaki Imamura, MD, PhDa,
Michael L. Schmitz, MDb,
Bryan Watkins, MDb,
Carl W. Chipman, RNa,
Sherry C. Faulkner, CCPa,
William P. Fiser, Jr, MDa,
Stephen H. Van Devanter, MDa,
Jonathan J. Drummond-Webb, MDa,*
a Department of Pediatric and Congenital Heart Surgery, Little Rock, AR, USA
b Pediatric Cardiovascular Anesthesiology, Arkansas Children's Hospital, Little Rock, Arkansas, USA
Accepted for publication May 4, 2004.
* Address reprint requests to Dr Drummond-Webb, Department of Pediatric and Congenital Heart Surgery, Arkansas Children's Hospital, 800 Marshall St, Slot 677, Little Rock, AR 72202, USA
Presented at the Fiftieth Annual Meeting of the Southern Thoracic Surgical Association, Bonita Springs, FL, Nov 13–15, 2003.
 |
Abstract
|
|---|
BACKGROUND: Severe, refractory hypoxemia complicating uncorrected cyanotic congenital heart disease is a potentially lethal condition, even when urgent surgical intervention is undertaken. When a viral pneumonia initiates hypoxemia, the likelihood of a satisfactory outcome is further reduced. We examined our policy of venovenous extracorporeal membrane oxygenation support through the hypoxic event and performing delayed surgery, if required, to separate from extracorporeal membrane oxygenation.
METHODS: A single institution, retrospective review of an Institutional Review Board approved database was undertaken. Over a 6-year period, 18 instances were identified for 17 patients who became acutely hypoxemic from either inadequate pulmonary blood flow (8 instances) or a viral pneumonia (10 instances) complicating their cyanotic heart disease. Demographics, duration of venovenous extracorporeal membrane oxygenation and outcomes are reported.
RESULTS: The length of venovenous extracorporeal membrane oxygenation ranged from 13.5 to 362.5 hours (mean 130 ± 121 hours). During 10 supports, operations were performed to facilitate weaning from support. In 7 patients, extracorporeal support was weaned during this surgery. Follow-up was obtained in all patients over a period ranging from 4 months to 7 years (mean 39.0 ± 23.0 months). There were two late deaths due to sepsis 1.4 and 2.5 months after extracorporeal support.
CONCLUSIONS: Venovenous extracorporeal membrane oxygenation allows time for the recovery of acute hypoxic insult and resolution of some viral pneumonia processes. Palliative surgical procedures may be safely undertaken during extracorporeal support. Viral pneumonia is a risk for prolonged support. Venovenous extracorporeal membrane oxygenation is useful in these high-risk patients.
|
Introduction
|
|---|
Venoarterial extracorporeal membrane oxygenation (VAECMO) therapy was introduced for the management of respiratory failure until 1982, when it was adapted for neonatal respiratory and cardiac failure [1]. VAECMO readily supports both pulmonary and cardiac function. Venovenous extracorporeal membrane oxygenation (VVECMO) has several advantages over VAECMO including a lower incidence of neurologic complications, lack of arterial compromise, the potential for a single venous cannula in smaller patients, preservation of pulsatile perfusion, and increased oxygen delivery to the myocardium and pulmonary vascular bed. In the absence of cyanotic congenital heart disease, VVECMO has allowed infants with compromised cardiac function to recover from severe respiratory compromise as a result of an acute viral pneumonia manifesting hypoxemia and respiratory acidosis [2]. Cyanotic congenital heart disease limits the tolerance for further hypoxemia induced by pulmonary failure or reduced pulmonary blood flow.
We have encountered a subgroup of patients with cyanotic congenital heart disease who present in extremis as a result of hypoxia induced by either the presence of a concurrent viral pneumonia or acute reduction in pulmonary blood flow. This is a very high-risk group of patients, for whom a conventional strategy to surgically improve pulmonary blood flow or attempt complete correction would be extremely hazardous [3]. Our preference for these patients is to provide support with VVECMO until sufficient recovery from the pneumonia, and or hypoxic insult occurs.
We report a consecutive series of 17 pediatric patients with cyanotic congenital heart disease who were supported in 18 instances with VVECMO at a single tertiary institution between July 1997 and May 2003.
|
Material and Methods
|
|---|
A retrospective review of an Institutional Review Board approved cardiac surgical database and an extracorporeal membrane oxygenation (ECMO) database at Arkansas Children's Hospital was undertaken.
Patient Profiles
From July 1997 to June 2002 ECMO support (both VAECMO and VVECMO) was utilized in 196 instances. We identified 17 patients with cyanotic congenital heart disease who were supported in 18 instances with VVECMO. We were able to identify 8 patients who presented with an acute viral pneumonia (group 1) and 9 who had acute severe hypoxemia as a result of inadequate pulmonary blood flow (group 2). Congenital heart disease diagnoses are listed in Table 1. One patient in group 1 received two periods of VVECMO support. Two patients in group 2 were supported with VVECMO within 24 hours of birth. Three patients in group 2 had undergone cardiac catheterization and balloon atrial septostomy before VVECMO support (Fig 1). Additionally, 11 of 17 patients (4 in group 1 and 7 in group 2) had previously undergone a total of 12 palliative cardiac surgical procedures (Table 2). Of the 17 patients, 11 were male and 6 were female. Patient ages were similar in both groups, ranging from birth to 4 years old (0.48 ± 0.93 years) and weighting from 2.6 to 15.2 kg (5.3 ± 2.9 kg). Demographic data for each group is listed in Table 3.
View larger version (15K):
[in this window]
[in a new window]
|
Fig 1. Flow diagram of 18 employments of venovenous extracorporeal membrane oxygenation (VVECMO) for 17 patients with cyanotic heart disease; the relationship of cardiac catheterization study. *represents a category with the patient having VVECMO on two occasions. (BAS = balloon atrial septostomy; Cath = catheterization.)
|
|
Technique
In 3 patient the right atrium was accessed by a transthoracic approach because either VVECMO support was initiated in the early postoperative period or because the internal jugular vein had been ligated previously. In all other instances, a double-lumen coaxial cannula (Origen Biomedical, Austin, TX; Kendall Healthcare Products, Mansfield, MA; or Jostra AG, Hirrlingen, Germany) was introduced through the right internal jugular vein to the middle of the right atrium. A computer-aided roller pump system was utilized for ECMO support (Stockert Sorin Computer-Aided Perfusion System, Saluggia, Italy; Cobe Cardiovascular, Arvada, CA; and Sorin Biomedical, Munich, Germany). Systemic anticoagulation was set to maintain the active coagulation time between 200 and 220 seconds (Medtronic ACT II; Medtronic Inc., Minneapolis, MN). Ventilation was maintained during the period of support with pressure regulated volume support, tidal volume of 10 mL/kg, rate of 20/min, end-expiratory pressure of 10 cm of water pressure, and 30% inspired oxygen (modified from Gattinoni and colleagues [4]). At the time of decannulation, patients cannulated through the internal jugular vein underwent internal jugular vein repair (7 patients) or ligation (7 patients).
|
Results
|
|---|
The duration of VVECMO support ranged from 50.5 to 362.5 hours (mean 205.3 ± 135.4 hours) in group 1, and from 13.5 to 164.3 hours (mean 66.3 ± 48.0 hours) in group 2. During VVECMO support, 6 patients underwent cardiac catheterization (Fig 1). Surgical interventions to increase pulmonary blood flow were performed in 10 patients (5 in each group) in order to facilitate weaning from VVECMO support (Fig 1). In 7 patients (3 in group 1 and 4 in group 2), VVECMO was weaned and discontinued near the completion of these surgical procedures. Two other patients were weaned from support 1 day (group 2) and 2 days (group 1) after the surgical procedure. One patient in group 1 underwent an arterial switch operation with closure of a ventricular septal defect from VVECMO. This infant was converted to cardiopulmonary bypass from VVECMO during surgery, and then to VAECMO at the conclusion of surgery. This patient had a total of 17 days of postoperative VAECMO support and was successfully weaned. No other patients required conversion to VA support from VVECMO. Patient outcomes are summarized in Figure 2.
View larger version (15K):
[in this window]
[in a new window]
|
Fig 2. Flow diagram of 18 employments of venovenous extracorporeal membrane oxygenation (VVECMO) for 17 patients with cyanotic heart disease. *represents a category with the patient having VVECMO on two occasions. (VAECMO = venoarterial extracorporeal membrane oxygenation.)
|
|
Complications related to VVECMO were limited to one instance of hyperkalemia and cardiac arrest while changing the oxygenator and circuit as a result of priming the circuit with packed red blood cells with a high potassium concentration. All patients successfully weaned from VVECMO.
Follow-up was obtained for all patients ranging from 4 months to 7 years (mean 39.0 ± 23.0 months). In comparing the patients who presented for ECMO on the basis of a viral pneumonic process versus the patients presenting as a result of inadequate pulmonary blood flow, the only variable that demonstrated statistical significance was the duration of ECMO support (mean 205 ± 135 vs 66 ± 47 hours; p < 0.05), which was longer in those patients with a pneumonic process.
There was 1 hospital death 1.4 months after weaning from ECMO, and 1 late death at 2.5 months. Both patients died of complications related to septicemia and subsequent multiple organ dysfunction.
|
Comment
|
|---|
Due to recent advancements in medical technology, an expanded role for ECMO support has evolved. VVECMO has become an important tool for the support of pediatric patients with severe, refractory cardiac and respiratory failure in whom the cardiac compromise arises from hypoxemia induced by inadequate pulmonary oxygenation. Infants and children with cyanotic congenital heart disease often fall into this category because they have very little reserve to tolerate worsening hypoxemia.
Venovenous ECMO has several documented advantages over VAECMO. Myocardial oxygenation is improved because the pulmonary venous return is fully oxygenated, the coronary artery blood flow is derived from the left ventricle output (or systemic ventricle) and this fully saturated blood is provided to the coronary arteries [5]. The pulmonary vascular bed is supplied with more highly oxygenated blood as well, potentially reducing pulmonary vascular resistance [5]. In a VA circuit, there is some reduction in the saturation due to the mixing of coronary arterial blood by the relatively desaturated pulmonary venous blood excluded from the circuit mixing with the arterial cannula blood. There may be a reduction of emboli to the systemic circulation as much of the blood from the ECMO circuitry is reintroduced to the right heart and filtered through the pulmonary circulation [5]. Often only one great vessel needs to be cannulated, sparing compromise of an artery such as the carotid [5, 6]. Pulsatile flow is preserved to the pulmonary and systemic circulations [5]. Lower flow rates are required to adequately oxygenate the patient, reducing the total exposure of blood to foreign surfaces in the circuitry, and theoretically allowing improved oxygenator longevity [5]. The incidence of neurologic complications is lower overall for VVECMO versus VAECMO [7].
Disadvantages of VVECMO include the inability to directly support cardiac output and the requirement for ongoing intensive respiratory/mechanical ventilation manipulation. Ventilation cannot be suspended and as such, may become a difficult issue for the transport of patients while on VVECMO. VVECMO is more labor intensive, and does require more frequent interventions to optimize respiratory and cardiac performance.
Some patients, however, still require surgical interventions in order to wean from VVECMO, but these are likely patients that would have required emergent surgical intervention to enhance pulmonary blood flow under adverse conditions of respiratory failure with severe hypoxemia and myocardial dysfunction. The presence of a concurrent viral pneumonia, especially respiratory syncytial virus, is unfortunately associated with a worse outcome in any patient needing cardiopulmonary bypass. Hospital admissions related to respiratory syncytial virus-associated illness are two- to threefold higher for infants less than 6 months of age with congenital heart disease than for low-risk infants [8] and infants and preschool age children manifest postoperative complications of greater number and severity when undergoing cardiac surgery within 6 months of a respiratory syncytial virus infection [3]. VVECMO allows for the recovery of both the myocardial as well as the pulmonary function. Under these circumstances we use VVECMO as a resuscitative tool, allowing time for the myocardium and the pulmonary function to recover.
Surprisingly, none of our patients required direct conversion from VVEMCO to VAECMO. All weaned from VVECMO with the exception of one patient who was converted to cardiopulmonary bypass for the arterial switch and ventriculoseptal defect (VSD) closure. Postoperatively, VAECMO was used to support the patient for pulmonary hypertension refractory to nitric oxide. He subsequently underwent perforation of his VSD patch and was weaned from support.
Results obtained from the database of the Extracorporeal Life Support Organization (ELSO) [9] did not yield outcomes as promising as those we report for similar cardiac diagnoses. Our preliminary experience suggests that VVECMO is an important tool for the support of patients with cyanotic congenital heart disease with respiratory compromise due to inadequate pulmonary blood flow, or pulmonary compromise from an acute viral pneumonia. A period of "cooling off" allows recovery of all organ systems. VVECMO does not preclude concurrent surgical intervention. Certainly this approach should be considered in this select group of patients and using this strategy, we were able to achieve hospital discharge for 95% of our patients.
|
DISCUSSION
|
|---|
DR GEORGE R. DAICOFF (St. Petersburg, FL): I agree with the concept of providing additional support to keep these little desperate babies alive during palliative operations. We ran through a patch at one time when our anesthesia would have difficulty maintaining oxygenation and blood pressure especially when clamps were applied during simple Blalock-Taussig shunts. Many times the shunts were hurriedly completed and sometimes while resuscitation was being applied, which commonly resulted in a less than ideal shunt. It would seem unorthodox to use cardiopulmonary bypass for closed heart procedures. Besides the obvious advantage of sustaining life of these desperate infants, the use of this support allowed for a technically superior anastomosis. Not only were there fewer inadequate shunts, but also the subsequent palliation was better. I would congratulate the authors for this concept and would wonder if a small amount of hypothermia might be of additional protective benefit for these patients during the operation and for a short time afterwards?
DR IMAMURA: Thank you for your question. It does depend on the state of the patient prior to extracorporeal membrane oxygenation (ECMO) support. In those patients with severe cardiogenic or pulmonary shock, we will use mild hypothermia with the institution of ECMO. I agree with your comment that hypothermia is a valuable consideration in select patients.
DR ROSS M. UNGERLEIDER (Portland, OR): Just a couple of quick questions. It's a nice study. I am curious about three things.
First of all, in your advantages of VVECMO you mentioned a reduced incidence of embolic complications, but it seems to me that in patients with cyanotic heart disease, your incidence would be the same since there is usually a single ventricle type physiology.
The second question I have relates to the technique in which you applied VVECMO to the larger patients. I saw you had at least one up through an age range of 4, and venovenous ECMO (VVEMCO) in those patients probably required a different kind of cannulation since the cannula that you showed wouldn't be adequate for a larger child.
And thirdly, I am just curious about the way you applied VVECMO, and I think you had at least 1 patient who had a previous bidirectional Glenn, and if you could tell us a little bit about cannulation to put a patient on VVECMO who has had a previous Glenn. Thank you.
DR IMAMURA: Thank you very much for your comments. About the first question of embolic event, I totally agree with your comment. My comment of advantages of venovenous ECMO was in general, so this advantage is mainly for the biventricular patient.
The second question of the two cannulas. Fortunately in our series the maximum weight was 15 kg. In the case of adult patients, I agree, that kind of patient may need two cannulas. However, our patients are pretty small in size, so we only used the one cannula and we could obtain good oxygenation.
DR UNGERLEIDER: And the patients with a bidirectional Glenn shunt?
DR IMAMURA: That patient had a relatively early postoperative period, so we cannulated the right atrial appendage directly.
DR ANDREW C. FIORE (St. Louis, MO): What was your incidence of stroke or bleed in the patients, cerebral bleed?
DR IMAMURA: Cerebral bleeding was fortunately not in this series.
|
Acknowledgments
|
|---|
We gratefully acknowledge the contributions of the ECMO support teams at our hospital including cardiologists, neonatologists, operating room nurses, and ECMO specialists. Work was supported by funding from the Arkansas Log-A-Load Endowed Chair of Pediatric Cardiovascular Surgery at Arkansas Children's Hospital, University of Arkansas for Medical Sciences, Department of Surgery, Little Rock, Arkansas.
|
References
|
|---|
- Bartlett RH, Roloff DW, Custer JR, Younger JG, Hirschl RB. Extracorporeal life support: the University of Michigan experience. J Am Med Assn. 2000;283:904–908[Abstract/Free Full Text]
- Dickson ME, Hirthler MA, Simoni J, Bradley CA, Goldthorn JF. Stunned myocardium during extracorporeal membrane oxygenation. Am J Surg. 1990;160:644–646[Medline]
- Khongphatthanayothin A, Wong PC, Samara Y, et al. Impact of respiratory syncytial virus infection on surgery for congenital heart disease: postoperative course and outcome. Crit Care Med. 1999;27:1974–1981[Medline]
- Gattinoni L, Agostoni A, Pesenti A, et al. Treatment of acute respiratory failure with low-frequency positive-pressure ventilation and extracorporeal removal of CO2. Lancet. 1980;2:292–294[Medline]
- Zwischenberger JB, Steinhorn RH, Bartlett RH. Principles and practice of venovenous extracorporeal membrane oxygenation. Pettignano R, Clark RH, Cornish JD. ECMO: extracorporeal cardiopulmonary support in critical care. Second edition. Ann Arbor, MI: Extracorporeal Life Support Organization; 2000. p. 113–131
- Durandy Y, Chevalier JY, Lecompte Y. Single-cannula venovenous bypass for respiratory membrane lung support. J Thorac Cardiovasc Surg. 1990;99:404–409[Abstract]
- Zwischenberger JB, Nguyen TT, Upp JRJ, et al. Complications of neonatal extracorporeal membrane oxygenation: collective experience from the Extracorporeal Life Support Organization. J Thorac Cardiovasc Surg. 1994;107:838–849[Abstract/Free Full Text]
- Boyce TG, Mellen BG, Mitchel EFJ, Wright PF, Griffin MR. Rates of hospitalization for respiratory syncytial virus infection among children in Medicaid. J Pediatr. 2000;137:865–870[Medline]
- Extracorporeal Life Support Organization. ECLS Registry Report. 2004;2004:1–28
Top
Abstract
Introduction
