Extracorporeal support for septic shock*

In this issue of Pediatric Critical Care Medicine, Dr. MacLaren and colleagues (1) describe a series of 45 children with profound septic shock refractory to conventional treatment, managed with venoarterial extracorporeal circulation: extracorporeal life support (ECLS) and extracorporeal membrane oxygenation (ECMO). Twenty-one children (47%) were discharged and survived with no serious disability. These are very impressive results, especially considering that 18 were resuscitated from cardiac arrest, all had three or more organs failing, and all were failing despite maximal inotropic and pressor support. Although all had profound respiratory failure, patients who presented with primary pneumonia and septic shock were excluded from this group, to define the characteristics and results of ECLS management of septic shock itself. The cause of septic shock was bacterial infection in all but one patient, with meningococcus and streptococcus predominating. As ECLS was being developed, profound sepsis was initially considered a contraindication (2). However sepsis with shock and respiratory failure in new born infants responded well during ECLS support. The overall survival for this indication is 75%, and sepsis is now a standard indication in neonates (3, 4) This group of investigators from the Royal Children’s Hospital in Melbourne, Australia, have been leaders in evaluation of extracorporeal support for sepsis in children (5, 6). This is a full review of their experience and by far the largest experience with ECLS management of septic shock in pediatrics in the literature. ECLS has been proven very effective in prospective randomized trials in neonatal respiratory failure (7–9) and in a contemporary matched pairs study in pediatric respiratory failure (10). As the authors point out, there will not be a prospective randomized trial of ECLS for septic shock in children for many reasons, but it is reasonable to ask what the survival might have been in this series without extracorporeal support. Eighteen children were in cardiac arrest, and we might assume that two could ultimately survive without disability. Twenty-seven had three or more organs failing and were deteriorating despite maximal support, with an average lactate of eight and negative buffer base deviation of 10 meq/L. We might generously estimate that a third of these children could have survived without severe disability. This gives us a total of 11 potential survivors of 45 (24%). We can conclude that ECLS is effective for children in septic shock, but why should it be? One reason is perfusion. With conventional treatment, systemic perfusion is maintained with high doses of pressor and inotropic drugs, each of which has its own deleterious side effects. Furthermore, profound respiratory failure requires positive pressure ventilation at high mean intrathoracic pressures, decreasing venous return and further increasing the need for fluid loading, inotropes, and pressors. All of this results in metabolic acidosis secondary to poor respiration at the cellular level (including the myocardial cells), leading to the vicious cycle of septic shock. By taking over the circulation with a mechanical pump and taking over gas exchange with a mechanical membrane lung, ECLS eliminates the iatrogenic problems of fluid overload, high ventilator pressure and Fio2, and vasoactive drugs. Equally important is simply buying time. The doctor has time to identify the culprit bacteria and institute culture-specific antibiotics. The patient has time to survive while this treatment is undertaken. Futility (in the form of brain death or apparently irreversible organ failure) becomes clear, and efforts are directed to truly salvageable patients. Other less obvious factors that might be involved include systemic anticoagulation (perhaps avoiding microvascular sludging and consumption coagulopathy), treatment of anemia (improving systemic oxygen delivery), treatment of renal failure and fluid overload (because continuous hemofiltration is a simple addition to the extracorporeal circuit), and perhaps removal of deleterious humoral mediators, some of which become dissolved in the large plastic surfaces of the extracorporeal circuit. Because this is a very experienced team, there were few complications of extracorporeal support. The need for oxygenator, circuit, and pump changes is common and not a serious problem (although one patient died during a circuit change, which is always a risk in patients who are totally dependent on the extracorporeal circuit). It is interesting that children whose venous drainage catheters were placed directly in the right atrium via thoracotomy had a better survival rate (73%) than children cannulated through the neck vessels (44%). The flow rate goals were set very high for this population, and these goals are much more easily achieved with direct thoracotomy and cannulation of the atrium and aorta, which might account for improved results. The authors state that they are now using this approach by choice, and we look forward to seeing the results of a series of patients managed in that fashion. The authors refer to guidelines for treatment of septic shock in children published by the Society of Critical Care Medicine (11). Those guidelines include a recommendation to use ECLS when all other measures fail (a surprisingly perceptive recommendation, perhaps prescient of this report). Dr. MacLaren and colleagues (1) present convincing evidence that extracorporeal support is a valuable management technique in septic shock that is refractory to conventional treatment. Robert H.