J Thorac Cardiovasc Surg 1999;117:541-542
© 1999 Mosby, Inc.
SURGERY FOR CONGENITAL HEART DISEASE |
Commentary
Tom R. Karl, MS, MD,
Victoria, Australia
 |
Introduction
|
|---|
Although there is a large experience worldwide with the VAD in adults, and a proliferation of effective support systems, the pediatric experience remains relatively limited. This is due to size and technical considerations, but also to a perception that patients with complex congenital heart disease are unsuitable for univentricular support. Duncan and associates have provided evidence to the contrary, reporting their large pediatric experience with VAD and ECMO for cardiac and cardiopulmonary support. The clinical questions addressed are important for pediatric surgeons. The time frame and the similarities in patient populations and outcome invite commentary and comparison with experience at the Royal Children's Hospital.
Since 1989, we have been using both VAD and ECMO,
1-4 always employing centrifugal pumps. The centrifugal-pump VAD circuit is straightforward in concept and design, can be rapidly assembled and primed, and requires little technical attention. It is suitable for many cardiac arrest situations. In this format, the centrifugal pump is responsive to changes in the peripheral circulation, allowing fine tuning of inotropic drugs and vasodilators before weaning. The centrifugal pump is probably safer than a roller pump for long-term support outside the operating theater, since it generates constant pressure rather than flow.
The majority of our postoperative support has been with VAD rather than ECMO systems. In practice, at least for intraoperative placement, adequacy of univentricular support can be assessed by converting cardiopulmonary bypass to partial left heart assist, and ceasing gas exchange across the oxygenator. The bypass circuit can then be converted to centrifugal-pump VAD or ECMO, as appropriate, on the basis of hemodynamics, oxygenation and carbon dioxide clearance, and acid base status.
We have used the VAD in 53 patients, median age 3.5 months (range 2 days–19 years) and median weight 4 kg (range 1.9-70 kg). Our results 3-5 were remarkably similar to those of Duncan and colleagues. The time on VAD was similar for survivors and nonsurvivors (P = .69), with a median of 75 hours. Thirty-eight of 53 children were weaned from VAD (.72, CL = .57-.83) and 24 of 53 were eventually discharged (.46, CL = .31-.61). Kaplan-Meier survival probability at 1 year was .44 (CL = .31-.58). The post-weaning deaths reflected continued cardiac problems rather than specific VAD-related complications. Age, weight, support time, cardiopulmonary bypass time, ischemic time, univentricular anatomy, and timing of placement (intraoperative versus postoperative) were not strongly associated with survival probability. By comparison, of 40 patients supported with ECMO, 19 were weaned, 3 bridged to transplantation, and 19 eventually discharged from the hospital. Thus, although the weaning probability was better with VAD (.71 vs .48, P = .014), as in Boston the discharge probability was similar (.46 vs .48, P = 1.0). In interpreting these results, one must consider that most of the patients treated with a VAD could have been supported with ECMO, but the reverse would not generally apply.
In the Boston experience, younger patients were more likely to be supported with ECMO than with VAD, because of the complexity of the underlying heart disease and technical considerations. In our own practice, 38 of 53 patients supported with a VAD weighed less than 6 kg. 6 The probability of weaning was similar to that for the older patients (P = .07), but the 1-year survival probability was poorer, primarily related to irreversible cardiac disease. The smallest patient in our series was a 19-day-old, 1.9-kg baby with Taussig-Bing anomaly and arch obstruction, who was placed on VAD support in the postoperative period during a prolonged (1 hour) cardiac arrest and survived with no neurologic sequelae. As would be the case in Boston, we have been obliged to redefine the upper limit for reasonable resuscitation time in children in the era of circulatory support.
Neither VAD nor ECMO in this format is eminently suitable for long-term support. Our longest successful supports have been 120 hours (ECMO) and 144 hours (VAD), although nonsurvivors were well supported metabolically for 384 hours (ECMO) and 428 hours (VAD). Our longest ECMO support for noncardiac indications has been 6 weeks. The usefulness of centrifugal-pump VAD and ECMO as a bridge to transplantation will depend heavily on the immediate availability of suitable donor hearts, a major problem in many parts of the world. A major limitation is our inability to wean most patients from ventilatory support during centrifugal-pump VAD and ECMO, mobilize them, and make them independent of the intensive care unit.
There are two important questions to be addressed in analyzing this experience. First, do ECMO and VAD as reported herein adequately support the patient when cardiac function is compromised? The answer is clearly yes, although not in every patient, and not indefinitely. Second, does this type of support allow cardiac improvement to occur if improvement would not otherwise have occurred in a given patient? This is a much more complicated question. To ask it another way: What is the effect of patient selection on outcome? As in most areas of cardiac surgery, it is profound. As a group, our VAD patients with anomalous origin of the left coronary artery from the pulmonary artery and those with transposition of the great arteries (especially patients with intact ventricular septum operated on after 3 weeks of age) had a significantly better survival probability than others (.91, CL = .59-1.0, P = .04). 7 The worst outcome was in patients with technically imperfect repairs, coagulopathy, and those with complex cardiac anomalies that have an expected poor long-term outcome even if VAD and ECMO are not required. Perhaps this type of analysis of indications begs the question. In practice, if a reliable support device were available, almost any child accepted for a cardiac operation would be considered a candidate for support should the need arise.
Centrifugal-pump VAD is not expensive technology. In our own unit, an additional nurse is required when ECMO is used, but not for VAD, during which the child's primary nurse also monitors the VAD circuit (with a perfusionist and cardiac surgeon on call). Consequently, VAD adds only approximately $250 per day to normal intensive care unit and hospital expenses, whereas ECMO adds 4 to 5 times that amount.
The availability of paracorporeal pulsatile devices for support of infants and children may render the centrifugal pump system obsolete. Certainly there are several very elegant systems in use in Europe at this time that are proving to be suitable for long-term support, even in very small infants. However, for short-term assistance, the costs involved may place these devices out of the reach of many units, especially those not actively involved in cardiac transplantation.
The results with short-term mechanical support in Boston, Melbourne, and elsewhere strongly support continued application of this strategy for selected children with severe cardiac failure. The simplicity and efficacy of centrifugal-pump VAD should place it in the armamentarium of all units prepared to undertake cardiac operations in children. In the current era, decrement in surgical mortality is very difficult to come by, and Duncan and associates have demonstrated a way of achieving this end.
|
References
|
|---|
-
Karl TR, Horton SB, Mee RBB. Left heart assist for ischaemic postoperative ventricular dysfunction in an infant with anomalous left coronary artery. J Card Surg 1989;4:352-4. [Medline]
-
Karl TR, Horton SB, Sano S, Mee RBB. Centrifugal pump left heart assist in pediatric cardiac surgery: indications, technique, and results. J Thorac Cardiovasc Surg 1991;102:624-30. [Abstract]
-
Karl TR, Pennington GD. Extracorporeal circulatory support in infants and children. Sem Thorac Cardiovasc Surg 1994;6:154-60. [Medline]
-
Cochrane AD, Horton A, Butt W, Skillington P, Karl TR, Mee RBB. Neonatal and paediatric extracorporeal membrane oxygenation. Aust Assoc J Card Thorac Surg 1992;1:17-22.
-
Karl TR. Circulatory support in children. In: Hetzer R, Hennig E. Loebe M, editors. Mechanical circulatory support. Berlin: Springer; 1997. p. 7-20.
-
Thuys CA, Mullaly RJ, Horton SB, O'Connor EB, Cochrane AD, Brizard CP, et al. Centrifugal ventricular assist in children under 6 kg. Eur J Cardiothorac Surg 1998;13:130-4. [Abstract/Free Full Text]
-
Cochrane AD, Coleman DM, Davis AD, Brizard CPR, Wolfe R, Karl TR. Excellent long-term functional outcome after an operation for anomalous left coronary artery from the pulmonary artery. J Thorac Cardiovasc Surg 1998: In press.
Top
Introduction
References