Extracorporeal membrane oxygenation for very high-risk transcatheter aortic valve implantation

Abstract

Introduction: Transcatheter aortic valve implantation (TAVI) can cause profound haemodynamic perturbation in the peri-operative period. Veno-arterial extracorporeal membrane oxygenation (ECMO) can be used to provide cardiorespiratory support during this time, either prophylactically or emergently. Methods: 100 TAVI procedures were performed between 2009 and 2013 in our institution. ECMO was used in 11 patients, including 8 prophylactic and 3 rescue cases. Rescue ECMO was required for ventricular fibrillation after valvuloplasty, and aortic annulus rupture. The criteria for prophylactic ECMO included heart failure requiring stabilisation pre-TAVI, haemodynamic instability with balloon aortic valvuloplasty performed to improve heart function pre-TAVI, moderate or severe left and/or right ventricular failure, or borderline haemodynamics at procedure. Differences in preoperative characteristics and postoperative outcomes between ECMO and non-ECMO TAVI patients were compared, and significant results were further assessed controlling for EuroSCORE. Results: Compared to TAVI patients who did not require ECMO, ECMO patients had significantly higher mean EuroSCORE (51 vs. 30%, p<.05). All-cause mortality occurred in nil prophylactic ECMO patients, one rescue ECMO patient, and two non-ECMO patients. The difference in mortality between ECMO and non-ECMO patients was not significantly different (9 vs. 2%; p>.05). Other non-significant outcomes (p<.05) included peri-procedural and spontaneous myocardial infarction (0 vs. 2% and 0 vs. 0%, respectively); disabling and non-disabling stroke (0 vs. 0% and 0 vs. 2%, respectively); transient ischaemic attack (0 vs. 0%); life-threatening, major and minor bleeding (12 vs. 3%, 0 vs. 12% and 38 vs. 13%, respectively); and major and minor vascular complications (12 vs. 5% and 0 vs. 13%, respectively). ECMO patients were more likely to develop acute renal failure than non-ECMO patients (36 vs. 8%, p<.05), which was most likely due to haemodynamic collapse and end-organ dysfunction in patients that required emergent ECMO rescue. Discussion: ECMO, in both prophylactic and rescue settings, is effective at providing cardiorespiratory support in very-high risk TAVI patients, and may have contributed to the low mortality in this series. Instituting prophylactic ECMO in an elective, controlled setting may help avoid consequences of intraoperative complications such as systemic hypoperfusion and acute renal failure, compared to emergent ECMO with rapid setup and cannulation. Limitations of this study include small patient sample size and observational nature. Future studies performed in a prospective, randomised fashion are required to evaluate the use of prophylactic ECMO, and to determine the ideal selection criteria.