Does single ventricle physiology affect survival of children requiring extracorporeal membrane oxygenation support following cardiac surgery?

BACKGROUND

Improved survival with postoperative extracorporeal membrane oxygenation (ECMO) has expanded its application to children with single ventricle (SV) anomalies. We examine current-era outcomes of postoperative ECMO with special focus on patients with SV.

METHODS

Demographic, anatomic, surgical, and support details of 100 consecutive children requiring postoperative ECMO (2007-2012) were included into multivariable regression models to identify factors affecting survival.

RESULTS

Median age was 73 days (4 days-16.2 years), 31 patients had SV physiology. The ECMO indication was failure to wean cardiopulmonary bypass (34%) and postoperative low cardiac output (66%) including 37% having extracorporeal cardiopulmonary resuscitation (ECPR). Median ECMO duration was four days (1-21). The ECMO decannulation and survival to hospital discharge were 62% and 37%. In SV group, decannulation and survival rates were 55% and 32%. The SV-ECMO outcomes were best in ECPR subgroup (54%), following shunt (57%) or Norwood (46%) and worst following Glenn, Fontan, or total anomalous pulmonary venous connection repair (0% survival). On multivariable analysis, factors affecting odds of survival were performing angiogram (odds ratio [OR]: 15.28, confidence interval [CI]: 2.34-99.89, P = .004), prolonged ECMO duration (OR: 0.64, CI: 0.47-0.88 per day, P = .005), leaving cannulation snares (OR: 28.41, CI: 2.65-304.70, P = .006), higher HCO3 (OR: 1.19, CI: 1.04-1.36, P = .01), renal failure requiring hemodialysis (OR: 0.21, CI: 0.06-0.76, P = .02), bleeding requiring re-exploration (OR: 0.21, CI: 0.06-0.75, P = .02), ECPR in patients with SV (OR: 11.84, CI: 1.11-126.07, P = .04), delayed lactate normalization (OR: 0.95, CI: 0.90-0.99 per hour, P = .02), and elevated liver enzymes (OR: 0.97, CI: 0.95-1.00 per 10 unit/L, P = .05).

CONCLUSIONS

The ECMO is valuable in patients with SV however results depend on anatomy, procedure, and support indication. Persistent markers of poor perfusion, end-organ injury, and prolonged ECMO duration are associated with mortality. Those factors could be modified by early ECMO application before organ damage, meticulous homeostasis to ensure adequate perfusion, early diagnosis, and reoperation on residual lesions to expedite weaning.