Extracorporeal Life Support in Pediatric and Neonatal Critical Care: A Review

Therapeutic drug monitoring of voriconazole in a child with invasive aspergillosis requiring extracorporeal membrane oxygenation

We describe a patient with invasive pulmonary aspergillosis on extracorporeal membrane oxygenation therapy in which therapeutic drug monitoring and individualization of therapy by measuring voriconazole plasma concentrations were performed.

[ECMO: experience in paediatrics]

INTRODUCTION

ECMO (Extracorporeal Membrane Oxygenation) provides a vital support to patients with supposed reversible respiratory and/or cardiac failure, in whom conventional support techniques have been previously unsuccessful.

OBJECTIVES

To determinate the criteria used in our hospital to put paediatric patients on ECMO, compare their clinical course depending on their pathology (respiratory failure, congenital heart disease or sepsis) and identify the sequelae attributable to this technique.

MATERIAL AND METHOD

A retrospective review of clinical records of all patients on ECMO support in our centre, excluding those presenting typically in neonatal period.

RESULTS

ECMO was used on 16 patients from June 2001 to January 2007, of which 50% were males. The median age was 7 months (from 21 days to 11 years). The reason for starting ECMO was respiratory failure in 11 cases (oxygenation index >40 and/or alveolar-arterial oxygen gradient >605), congenital heart disease in 2 and sepsis in 3 (due to shock unresponsive to adequate resuscitation). The median time to starting ECMO from PICU admission was 3.58 days (from 12h to 9 days). Venovenous cannulation was used initially in 8 patients, but 5 of them needed venoarterial ECMO later. The technique was used for a mean of 8 days (from 1 to 28 days). The main complication was the isolation of bacteria in different cultures (8 patients). The overall survival was 50% (6 patients with respiratory failure and both patients submitted to cardiac surgery). Extracorporeal support was withdrawn in 7 children because their clinical situation was irreversible. Another patient died seven days after successful decannulation. We have not found any serious sequel among survivors that could be attributable to this technique.

CONCLUSIONS

Survival among children supported with ECMO in our hospital is similar to that recorded by the ELSO in 2004, although the prognosis depends on the initial pathology. There are different criteria for starting this technique depending on the underlying diseases: respiratory index of poor prognosis in patients with respiratory failure, haemodynamic instability in those with sepsis or cardiac failure after cardiovascular surgery. We have not found any serious sequel among the survivors which could be attributable to this technique.

Two-year survival, mental, and motor outcomes after cardiac extracorporeal life support at less than five years of age

OBJECTIVE

Comprehensive outcome assessment of children receiving cardiac extracorporeal life support.

METHODS

From 2000 to 2004, 39 consecutive children (aged 1 day to 4.4 years) had cardiac extracorporeal life support. Neurodevelopmental follow-up of all survivors was performed more than 6 months after life support (aged 53 +/- 12 months). Developmental delay was defined as a score of less than 70 on the Bayley Scales of Infant Development II or Wechsler Preschool and Primary Scale of Intelligence. Predictor variables for mortality (at 2 years' follow-up) and delay were examined by univariate and multivariate analyses.

RESULTS

Indications for extracorporeal life support were progressive low cardiac output in 14 (36%), failed weaning from cardiopulmonary bypass in 13 (33%), cardiac arrest in 9 (23%), and hypoxia in 3 (8%). Cardiac anatomy was single ventricle in 16 (41%), biventricular in 21 (54%), and myocarditis in 2 (5%). Survival was 18 (46%) at hospital discharge and 16 (41%) at 2 years. In survivors, mental score was 73 +/- 16 (normal 100 +/- 15), and 8 (50%) had mental delay. Initiating extracorporeal life support during cardiopulmonary resuscitation and duration of this resuscitation were not associated with death or mental delay. On multivariable Cox regression, lactate on admission to the pediatric intensive care unit (hazard rate 1.13; 95% confidence intervals 1.08-1.27) and single ventricle anatomy (hazard rate 3.93; 95% confidence intervals 1.62-9.49) were associated with death at 2 years. Stepwise multiple regression found time for lactate to normalize on extracorporeal life support, highest inotrope score during 120 hours of life support, and chromosomal abnormality explained 76.7% of the variance in mental score.

CONCLUSION

Cardiac extracorporeal life support had a 41% 2-year survival. Potentially modifiable variables (time for lactate to normalize and highest inotrope score early during extracorporeal life support) explained 69% of mental score variance.

Intravascular volume administration: a contributing risk factor for intracranial hemorrhage during extracorporeal membrane oxygenation?

OBJECTIVE

The objective of this study was to determine the relationship between the frequency and total volume of intravascular volume administration and the development of intracranial hemorrhage during venoarterial extracorporeal membrane oxygenation.

METHODS

In a retrospective, matched, case-control study, 24 newborns who developed an intracranial hemorrhage during venoarterial extracorporeal membrane oxygenation treatment were compared with 40 control subjects. Both groups were analyzed for gestational age, gender, race, Apgar scores at 1 and 5 minutes, birth weight, cardiopulmonary resuscitation before venoarterial extracorporeal membrane oxygenation, age at the start of treatment, duration of treatment, worst arterial blood gas sample preceding treatment, activated clotting time values, need for platelet transfusions, mean blood pressure, and the use of inotropics and steroids before the treatment. For both groups, total number and volume of intravascular infusions of normal saline, pasteurized plasma protein solution, erythrocytes, and platelets during the first 24 hours of treatment were determined. Variables were analyzed in their relationship to intracranial hemorrhage by using univariate and multivariate conditional logistic regression.

RESULTS

The only statistically significant difference in patient characteristics between the case patients and control subjects was arterial blood gas values. Newborns who developed intracranial hemorrhage during the treatment received both a statistically significantly higher number and a statistically significantly higher total volume of intravascular volume administrations compared with control patients. After adjustment for pH, Paco(2), and Pao(2) in the multivariate analysis, we found a significant relation between the development of intracranial hemorrhage and >8 infusions or >300 mL of volume infusion in the first 8 hours and >10 infusions in the first 24 hours of treatment.

CONCLUSIONS

The number and total volume of intravascular volume administration in the first 8 and 24 hours of venoarterial extracorporeal membrane oxygenation treatment are statistically significantly related to the development of intracranial hemorrhage.

Potential drug sequestration during extracorporeal membrane oxygenation: results from an ex vivo experiment

OBJECTIVE

Using an ex vivo simulation model we set out to estimate the amount of drug lost due to sequestration within the extracorporeal circuit over time.

DESIGN

Simulated closed-loop extracorporeal membrane oxygenation (ECMO) circuits were prepared using a 1.5-m2 silicone membrane oxygenator. Group A consisted of heparin, dopamine, ampicillin, vancomycin, phenobarbital and fentanyl. Group B consisted of epinephrine, cefazolin, hydrocortisone, fosphenytoin and morphine. Drugs were tested in crystalloid and blood-primed circuits. After administration of a one-time dose of drugs in the priming fluid, baseline drug concentrations were obtained (P0). A simultaneous specimen was stored for stability testing at 24 h (P4). Serial post-membrane drug concentrations were then obtained at 30 min (P1), 3 h (P2) and 24 h (P3) from circuit fluid.

MEASUREMENTS AND RESULTS

One hundred and one samples were analyzed. At the end of 24 h in crystalloid-primed circuits, 71.8% of ampicillin, 96.7% of epinephrine, 17.6% of fosphenytoin, 33.3% of heparin, 17.5% of morphine and 87% of fentanyl was lost. At the end of 24 h in blood-primed extracorporeal circuits, 15.4% of ampicillin, 21% of cefazolin, 71% of voriconazole, 31.4% of fosphenytoin, 53.3% of heparin and 100% of fentanyl was lost. There was a significant decrease in overall drug concentrations from 30 min to 24 h for both crystalloid-primed circuits (p = 0.023) and blood-primed circuits (p = 0.04).

CONCLUSIONS

Our ex vivo study demonstrates serial losses of several drugs commonly used during ECMO therapy. Therapeutic concentrations of fentanyl, voriconazole, antimicrobials and heparin cannot be guaranteed in patients on ECMO.

Venoarterial extracorporeal membrane oxygenation (VA-ECMO) in pediatric cardiac support

BACKGROUND

Resuscitation extracorporeal membrane oxygenation (R-ECMO) was introduced at our institution in July 2002. We reviewed the use of venoarterial (VA)-ECMO for cardiac diagnoses at our institution.

METHODS

Retrospective analysis of patients on VA-ECMO for cardiac failure was performed. Survival was defined as discharge from hospital.

RESULTS

Twenty-seven patients were supported with VA-ECMO (median age, 27 days; range, 1 to 640 days; median weight, 3.8 kg; range, 1.8 to 11.3 kg). Diagnoses were cardiomyopathy-myocarditis (CMM) in 8 (30%), systemic-to-pulmonary artery shunt-dependent single ventricle (SV) in 12 (44%), postcardiotomy for biventricular repair (BiV) in 6 (22%), and arrhythmia in 1 (4%). Sixteen of 27 patients survived (59%). Seven of 8 CMM patients survived (88%); 6 (75%) bridged to cardiac recovery, 1 to transplant (13%), and 1 death (13%). Seven of 12 SV patients survived (58%). The SV ECMO indications: post-Norwood ventricular dysfunction (n = 3, 2 deaths), postoperative cardiac failure (n = 6, 2 deaths), respiratory failure (n = 1, 1 death), and acute shunt occlusion (n = 2, 0 deaths). One of 6 BiV patients survived (17%). The BiV ECMO indications: failure to wean from CPB (n = 3, 3 deaths), postoperative cardiac failure (n = 2, 2 deaths), and pulmonary hypertension (n = 1, 0 deaths). Fifteen patients (56%) underwent cardiopulmonary resuscitation during ECMO cannulation. Eleven of 15 R-ECMO patients (73%) survived versus 5 of 12 non-R-ECMO patients (42%, p = 0.13). Median duration of R-ECMO: 66 hours (range, 18 to 179) versus 145 hours (range, 43 to 986, p = 0.01) for non-R-ECMO.

CONCLUSIONS

Resuscitation extracorporeal membrane oxygenation is an appropriate application in pediatric patients with cardiac disease. Single ventricle patients experiencing cardiopulmonary collapse and CMM patients have favorable outcomes. Failure to wean from CPB and postoperative ventricular failure are higher risk indications.