Table 1. Extracorporeal membrane oxygenation as treatment of graft failure after heart transplantation: review of the literature.
| Study | Design | Quality | Patients | ECMO survivors (discharged) | Short-term outcomes (<30 days or in-hospital) | Long-term outcomes |
|---|---|---|---|---|---|---|
| Tissot et al. (1) [2009] | Retrospective, pediatric | 3 | 310 HTx; 28 ECMO | 15/28 (54%) | Duration of support: survivors 2.8±0.6 vs. non-survivors 4.7±2.6 | Among ECMO survivors: ❖ Three years survival: 100%; ❖ Long-term survival (8.1±3.8 years): 87%, deaths related to transplant CAD and chronic graft failure; ❖ Stroke: 27%; ❖ Normal graft function (SF 36±7%, EF 65±10%, CI 4.5±1.3) |
| Causes of death on ECMO support: HF (69%), sepsis (30%), bleeding (23%) (overlap) | ||||||
| Longer ischemic time in ECMO patients (276±89 vs. 242±70 min) | ||||||
| Chou et al. (9) [2010] | Retrospective, adult | 1 | 366 HTx; 40 ECMO | 31/40 (78%) | None of the patients receiving ECMO support for >4 days survived | Late survival: 52% (survival details not reported) |
| D’Alessandro et al. (5) [2010] | Retrospective, adult | 3 | 394 HTx; 54 ECMO | 27/54 (50%) | Duration of support: 7±3 days Complications during support: infection (65%), bleeding (26%), stroke (20%), mediastinitis (19%) |
Among ECMO survivors: ❖ Three years survival: 94%; ❖ ECMO survivors have the same one-year and three-year survival of patients without early graft failure |
| Causes of death during support: HF (30%), sepsis (26%), bleeding (18%), stroke (18%), other (7%) | ||||||
| Marasco et al. (10) [2010] | Retrospective, adult | 2 | 239 HTx; 39 ECMO | 29/39 (74%) | Duration of support: 6.8±2.6 days | Among ECMO survivors: similar survival of patients without early graft failure (survival details not reported) |
| Median blood products used: packed cells 5,890 mL, fresh frozen plasma 2,250 mL, platelets 2,130 mL, cryoprecipitate 300 mL | ||||||
| Complications during support: bleeding (38%), sternal wound infection (10%), gastrointestinal hemorrhage (13%), renal failure (67%), pneumonia (46%), sepsis (49%), multi-organ failure (13%), limb complications (13%), stroke (8%). No differences in complications between central and peripheral cannulation | ||||||
| D’Alessandro et al. (11) [2011] | Retrospective, adult | 3 | 402 HTx; 91 ECMO | 42/91 (46%) | 33 patients died while on ECMO support: causes of death were low output cardiac syndrome (n=2), stroke (n=7), septic shock (n=10), postoperative surgical bleeding (n=7), refractory vasoplegia (n=6), and gastrointestinal bleeding (n=1) | Among ECMO survivors: similar conditional 1-year survival rates as non-EGF patients (93% at three years and 91% at five years without EGF versus 93% at three years and 84% at five years with EGF) |
| Both patients bridged to total artificial heart died (stroke and septic shock) | ||||||
| 21 patients died after weaning; causes of death were low-output cardiac syndrome (n=1), stroke (n=2), septic shock (n=11), hypoxic arrest during weaning after a tracheotomy (n=2), acute rejection (n=1), gastrointestinal bleeding (n=1), and iatrogenic hemothorax (n=1) | ||||||
| Two patients had refractory graft failure after removal of the ECMO. Both patients died after ECMO reimplantation due to surgical bleeding (n=1) and septic shock (n=1) | ||||||
| Mihaljevic et al. (12) [2012] | Retrospective, adult | 1 | 1,417 HTx; 43 ECMO | Not available | ECMO not differentiated from other TMSC in the presentation of results. Survival of patients requiring TMCS at one day, one week, one month: 94%, 66%, 43%, respectively | Survival of patients requiring TMCS at one year, five years, and ten years: 40%, 37%, 29%, respectively |
| Complications during support: stroke (4%), device infection (4%), reoperation for bleeding (7%). Cumulative number of events increased linearly with duration of TMCS (rate of 0.34 events/patient per week) | ||||||
| Chen et al. (8) [2014] | Retrospective, adult | 1 | 447 HTx; 26 ECMO | 12/26 (46%) | Not reported | Among ECMO survivors: similar one-year survival rates as non-EGF patients |
| Lehmann et al. (13) [2014] | Retrospective, adult | 2 | 298 HTx; 28 ECMO | 15/28 (54%) | All peripheral cannulations | Among ECMO survivors: ❖ one-year survival rate: 63%±14%; ❖ Ten-year survival rate: 63%±14% |
| Duration of support: 4.2±2.9 days | ||||||
| Concomitant IABP use: 46% | ||||||
| Complications during support: bleeding (25%), gastrointestinal bleeding (28%), stroke (7%), deep sternal wound infection (4%) | ||||||
| Ventilation time 28.6±24.6 days | ||||||
| Blood products use, units: red blood cells 56±29, fresh frozen plasma 13±11, platelet 43±26 | ||||||
| Causes of death during support: multiorgan failure (n=9), sepsis (n=9), lung failure (n=2), stroke (n=2) | ||||||
| Kaushal et al. (14) [2014] | Retrospective, pediatric | 2 | 92 HTx; 16 ECMO | 9/16 (56%) | Timing of cannulation: 50% in the operating room because of the inability to wean from cardiopulmonary bypass, 50% in intensive care unit (emergency) | Among ECMO survivors: ❖ Five-year survival rate 40%; ❖ Graft function normal, and similar to non-ECMO patients (SF 42±3 vs. 40±2) |
| Duration of support: 5.4 days in survivors vs. 8.0 days in non-survivors | ||||||
| Re-initiation of ECMO support in three patients | ||||||
| Cause of death during support: cardiac failure with multi-organ failure | ||||||
| Su et al. (15) [2015] | Retrospective, pediatric | 2 | 203 HTx; 29 ECMO | 18/29 (62%) | Duration of support: 7.6±8.1 days | Not reported |
| Causes of death during support: MOF (n=4), HF (n=2), stroke (n=1), sepsis (n=2) | ||||||
| Survival was not associated with gender, age, indication for HTx, past surgical history, ischemic time, CPB time, intra-operative or postoperative ECMO cannulation, time between HTx and ECMO, or duration of support | ||||||
| Loforte et al. (16) [2015] | Retrospective, adult | 3 | 119 ECMO; 26 HTx | 19/26 (73%) | Cannulation: peripheral (38%), central (62%) | One-year survival: 57.6% |
| Duration of support, median: 8 days | ||||||
| Duration of support >6 days: 54% | ||||||
| IABP on ECMO: 100% | ||||||
| Intubation time, median: 10 days | ||||||
| Hospital stay, median: 21 days Complications during support: dialysis (30%), reoperation for bleeding (46%), pneumonia (11%), liver failure (27%), multi-organ failure (27%), sepsis (4%) | ||||||
| Median blood products used, units: red blood cells 16.1, platelets 15.8, fresh frozen plasma 5.2 | ||||||
| Re-transplantation: 15% | ||||||
| Loforte et al. (7) [2016] | Retrospective, adult | 3 | 412 HTx; 29 ECMO | 13/29 (45%) | Cannulation: peripheral (20.7%), central (79.3%) | Among ECMO survivors: similar five-year conditional survival rate as HTx patients who had not suffered from EGF (88% without EGF vs. 84% with EGF treated with TMCS) |
| Duration of support: overall (7.9±6.4 days), peripheral (6.4±4.3 days), central (11.2±7.9 days) | ||||||
| Pre-HTx ECMO support: 44.8% | ||||||
| Mechanical ventilation time: 12.8±16.8 days | ||||||
| Intensive care unit stay: 28.2±20.4 days | ||||||
| Complications during support: leg ischemia (7%), reoperation for bleeding (38%), dialysis (55.2%), multi-organ failure (31%), stroke (10%), sepsis (24%) | ||||||
| Causes of death during support: multi-organ failure (47%), sepsis (36%), stroke (16%) | ||||||
| Takeda et al. (17) [2017] | Retrospective, adult | 3 | 597 HTx; 27 ECMO | 22/27 (81%) | Duration of support: 5.2±3.9 | The three-year post-transplant survival was 41% in the VAD group and 66% in the ECMO group |
| Cardiopulmonary bypass time: 216±65 min | ||||||
| Aortic cross clamp time: 193±52 min | ||||||
| Blood products use during support: red blood cells 1,359±1,415 mL, fresh frozen plasma 1,916±1,728 mL, platelets 646±455 mL | ||||||
| Peripheral cannulation in 85%, no patient required left ventricular vent during ECMO support | ||||||
| Compared to ECMO, VAD implantation is associated with: ❖ Longer cardiopulmonary bypass time (323±86 vs. 216±65 minutes); ❖ Longer support time (14±17 vs. 5±4 days); ❖ Higher incidence of major bleeding requiring reoperation (77% vs. 30%); ❖ Higher incidence of dialysis (53% vs. 11%); ❖ Higher in-hospital mortality (41% vs. 19%); ❖ Lower weaning from support (59% vs. 89%) | ||||||
| Phan et al. (3) [2017] | Systematic review | 2 | 11,555 HTx; 535 ECMO1 | 57% | Duration of support: 5.2 days | Not reported |
| Complications during support: infection (34%), reoperation (34%), renal failure (29%), stroke (12%), device-related complications (7%) | ||||||
| Causes of death during support: sepsis (40%), stroke (34%), cardiac (26%) (weighted results) | ||||||
| Tran et al. (2) [2018] | Retrospective, adult | 1 | 224 ECMO; 62 HTx | 30/62 (48%) | Not reported | Not reported |
1, result of ECMO patients are generally presented as percentages and the precise number of ECMO patients is reported differently in the various sections of the text. Quality of evidence ranges from 1 to 3, with 1 indicating poor quality (missing information, methods not reported) and 3 indicating excellent quality for the purpose of this article. CI, cardiac index; ECMO, extracorporeal membrane oxygenation; EF, ejection fraction; HF, heart failure; HTx, heart transplantation; RV, right ventricle, SF, shortening fraction.