Extracorporeal Membrane Oxygenation Support in Postcardiotomy Elderly Patients: The Mayo Clinic Experience Pankaj Saxena, FRACS, PhD, James Neal, CCP, Lyle D. Joyce, MD, PhD, Kevin L. Greason, MD, Hartzell V. Schaff, MD, Pramod Guru, MD, William Y. Shi, MBBS, Harold Burkhart, MD, Zhuo Li, William C. Oliver, MD, Roxann B. Pike, MD, Dawit T. Haile, MD, and Gregory J. Schears, MD Division of Cardiovascular Surgery, Perfusion Services, and Departments of Anesthesiology and Biostatistics, Mayo Clinic, Rochester, Minnesota; and University of Melbourne, Melbourne, Australia
Background. We conducted a retrospective study to assess whether providing extracorporeal membrane oxygenation (ECMO) support to elderly patients (aged 70 years or more) who failed separation from cardiopulmonary bypass after cardiac surgery was a viable option. Methods. From 2003 to 2013, 45 patients aged 70 years or more underwent 47 runs of ECMO postoperatively. Results. There were 31 men (68.9%). The mean age was 76.8 years. Five patients were in cardiogenic shock preoperatively. Forty-four patients required venoarterial ECMO support for cardiogenic shock. Mean duration of support was 103.8 ± 74.3 hours. Twenty-one patients (46.6%) died while on ECMO support. Twenty-four patients (53.3%) were weaned off ECMO initially, and 11 patients were discharged from hospital. Inhospital mortality was 75.6%. Postoperative complications included
acute kidney injury in 30 patients (44.4%), pneumonia in 12 (26.7%), and sepsis in 11 (24.4%). There were 30 deaths (88.2%) attributable to cardiac causes. Preoperative atrial fibrillation, chronic kidney injury, lactic acidosis on ECMO support, and persistent coagulopathy were associated with higher mortality. Conclusions. Postcardiotomy ECMO support in elderly patients is associated with high postoperative morbidity and mortality. Nevertheless, it often provides the last line of therapy for these critically ill patients and may provide positive outcomes in selected subgroups.
E
from January 2003 to January 2013 (Institutional Review Board reference number 13-001149).
xtracorporeal membrane oxygenation (ECMO) support can be used effectively for postcardiotomy cardiopulmonary failure. Cardiogenic shock after cardiac surgery in elderly patients has been associated with high postoperative mortality [1]. The rate of successful ECMO removal in adult cardiac surgical series varies from 30% to 60% and is associated with inhospital mortality of 60% to 80% [2–5]. Literature on the use of ECMO in elderly cardiac surgical patients is limited. We reviewed our 10 years of experience with patients aged 70 years or more who had ECMO support after cardiac surgery to assess the futility or viability of this approach.
(Ann Thorac Surg 2015;-:-–-) Ó 2015 by The Society of Thoracic Surgeons
Cardiac Surgery Cardiac surgery was performed through either a median sternotomy or a thoracotomy. The chest was left open at the end of the procedure if there was significant bleeding due to coagulopathy or myocardial edema. The decision to proceed to ECMO was based on failure to respond to the standard hemodynamic support. Intraaortic balloon pump was placed either preoperatively or intraoperatively while attempting to wean the patient off CPB. Reexploration of the chest was undertaken during the ECMO support as clinically indicated.
Material and Methods The study was conducted at the Mayo Clinic, Rochester, Minnesota, and included patients who were aged 70 years or older and required postcardiotomy ECMO support Accepted for publication Nov 26, 2014. Address correspondence to Dr Schears, Department of Anesthesiology and Critical Care Medicine, Mayo Clinic, 200 First St SW, Rochester, MN 55905; e-mail: [email protected].
Ó 2015 by The Society of Thoracic Surgeons Published by Elsevier
ECMO Circuits A standard ECMO circuit was used. The circuits were primed with Plasmalyte A (Abbott, Chicago, IL) first, and then the clear prime was displaced in a transfer bag by packed red blood cells to maintain a hemoglobin level greater than 9 g/dL to 10 g/dL, 500 USP heparin, and 50 mL sodium bicarbonate. Usually 1 to 2 units of packed red blood cells were used to displace the clear prime. 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2014.11.075
2
SAXENA ET AL POSTCARDIOTOMY ECMO SUPPORT IN THE ELDERLY
Ann Thorac Surg 2015;-:-–-
ECMO Management
Table 1. Baseline Clinical Characteristics of Study Population
Heparin was usually given in the dose of 5,000 USP if the patient was not on CPB. The ECMO circuit flow was maintained at a cardiac index of 2.0 to 2.4 L $ min1 $ m2 as allowed by the cannula size and the patient’s volume status, with an aim to keep mixed venous saturation greater than 60%. Heparin infusion was started in the postoperative period when the chest tube output was less than 1 to 2 cc/kg per hour for the first few hours with normalized coagulation values. Heparin infusion was typically started within the first 12 hours after ECMO initiation. Anticoagulation was measured using either activated clotting time (POC iStat; Abbott), aiming for a range of 140 s to 170 s, or activated partial thromboplastin time, maintained in the range of 50 s to 80 s. Weaning off ECMO support was based on the clinical and investigative evidence of recovery of cardiac and pulmonary function. Anticoagulation therapy was adjusted to maintain activated clotting time greater than 200 s or activated partial thromboplastin time of 80 s or more at the time of weaning. The oxygenator’s blender gas flow and fraction of inspired oxygen (FiO2) was adjusted to maintain PaO2 greater than 150 mm Hg and PCO2 between 35 mm Hg and 45 mm Hg.
Clinical Characteristics
Statistical Methods Categoric variables are reported as frequencies and percentages, and continuous variables are reported as mean SD or median (range) as appropriate. Categoric variables were compared between patients who survived ECMO and those who did not using Fisher’s exact test, and continuous variables were compared using twosample t test or Wilcoxon rank sum test where appropriate. The Kaplan-Meier method was used to evaluate survival. All statistical tests were two sided, with the alpha level set at 0.05 for statistical significance.
Results Baseline Clinical Characteristics We identified 45 patients who underwent 47 runs of ECMO support after cardiac surgery (Table 1). Forty-one patients underwent primary cardiac surgery through median sternotomy, 3 had thoracotomies, and 1 patient had transfemoral transcatheter aortic valve replacement. Hypothermic circulatory arrest was utilized in 6 patients (mean circulatory arrest time 24.2 minutes; range, 11 to 44). Mean left ventricular ejection fraction was 48.5%. Twelve patients had a preoperative left ventricular ejection fraction of 30% or less.
Indications for ECMO Institution The ECMO was instituted when there was failure to wean patients from CPB despite maximal inotropic or intraaortic balloon pump support or both, cardiogenic shock with systolic blood pressure 85 mm Hg or less, cardiac index less than 1.5 L $ min1 $ m2, cardiac arrest in the operating room or in the ICU, right ventricular failure with or without concomitant left ventricular failure based
Value
Number of patients Age, years Sex Male Female Body mass index, kg/m2 (n ¼ 44) Chronic obstructive pulmonary disease Former smokers Current smokers Diabetes mellitus Hypertension Acute kidney injury Chronic kidney injury Peripheral vascular disease Preoperative cerebrovascular accident Recent myocardial infarction Previous myocardial infarction Preoperative cardiogenic shock Preoperative CPR Preoperative congestive heart failure Preoperative intraaortic balloon pump Preoperative LVAD Elective operation Urgent operation Emergent operation Redo cardiac surgery
45 76.8 4.6 31 (68.9) 14 (31.1) 29.5 5.1 10 (22.2) 8 (17.8) 1 (2.2) 16 (35.6) 31 (68.9) 5 (11.1) 12 (26.7) 7 (15.6) 5 (11.1) 5 (11.1) 6 (13.3) 5 (11.1) 5 (11.1) 24 (53.3) 6 (13.3) 1 (2.2) 35 (77.8) 6 (13.3) 4 (8.9) 25 (55.6)
Values are mean SD or n (%), unless otherwise indicated. CPR ¼ cardiopulmonary resuscitation; device.
LVAD ¼ left ventricular assist
on the information obtained from Swan-Ganz catheter, and intraoperative transesophageal echo (TEE). Fortyfour patients (97.8%) received ECMO to provide mechanical circulatory support. One patient had ECMO support to decompress the bleeding right ventricle during redo cardiac surgery.
Intraoperative Complications During Cardiac Surgery The primary cardiac surgical procedure was complicated intraoperatively in 26 patients, and possibly contributed to the need for ECMO support (Table 2).
ECMO Data The ECMO support was instituted using peripheral cannulation in 15 patients (33%) and central cannulation in 30 (66%) patients. Venoarterial ECMO was utilized in 44 patients (97.8%), and a right ventricular assist device with oxygenator was used in 1 patient. Mean duration of ECMO support for the whole group was 103.8 74.3 hours (range, 1.0 to 378.5). Mean duration of ECMO support for the nonsurvivors was 106.8 81.8 hours; the survivors were supported for 93.9 41.3 hours (p > 0.99). ECMO was begun after a lag period due to delayed appearance of cardiogenic shock in 19 patients (42.2%);
Ann Thorac Surg 2015;-:-–-
SAXENA ET AL POSTCARDIOTOMY ECMO SUPPORT IN THE ELDERLY
3
Table 2. Intraoperative Events During Primary Cardiac Surgery and Outcome Pt. Age (years)
Primary Cardiac Surgical Procedure
71 85 71 83 75 75
Aortic root replacement CABG Redo AVR CABG AVR AVR, aortic root enlargement, CABG
76 70 72 77 76 73 71 82 79 79 84 81
Pericardiectomy Robotic MV repair CABG AVR, MVR, CABG MVR AVR, refashioning vein graft Maze, LA thrombectomy MVR, TV repair, CABG Repair of descending thoracic aorta Aortic root replacement, MV repair TAVI TA on bypass TVR and PVR, intraop stent placed in LM
84
Ascending aortic replacement, AVR
75 80
AVR, refashioning vein graft AVR
77
AVR, MVR, CABG
82
AVR, TV repair
79 79
TAVI TF AVR, valve retrieval from LV after failed TAVI, aortoplasty, CABG 1 AVR, ascending aorta replacement
74
Intraoperative Event
Outcome
RCA graft kink Revision of grafts Cardiac injury during resternotomy Poor coronary targets Vein graft injury during redo sternotomy Cardiac injury to LIMA and RV during redo sternotomy, aortic prosthesis reimplanted Failure to graft coronary arteries RCA thrombus Regrafting of coronary arteries Reimplantation of aortic valve, heavily calcific annulus Cardiac injury during redo sternotomy Coronary graft vasospasm Injury to RCA Bleeding from friable RV Cardiac failure, coagulopathy Myocardial protection issues VF during aortic root injection during TAVI Failure to graft coronaries and had intraoperative stent placement in LM Friable aorta bleeding from aortotomy requiring replacement of aorta Protamine reaction Heavily calcified aortic annulus and calcific atheroma of aorta Heavily calcified aorto-mitral calcification, mitral valve damage during debridement Bleeding requiring repair on multiple runs of CPB and cross clamping Intraoperative embolization to RCA during TAVI Valve embolization, failed TAVI, LIMA injury during sternal reentry Intraoperative anterior wall LV rupture
Survived Survived Died Died Died Died Died Survived Died Died Died Died Died Died Died Died Died Died Died Died Survived Died Died Died Survived Died
AVR ¼ aortic valve replacement; CABG ¼ coronary artery bypass graft surgery; CPB ¼ cardiopulmonary bypass; intraop ¼ intraoperative; LA ¼ left atrium; LIMA ¼ left internal mammary artery; LM ¼ left main coronary artery; LV ¼ left ventricle; MVR ¼ mitral valve replacement; Pt. ¼ patient; RCA ¼ right coronary artery; RV ¼ right ventricle; TAVI ¼ transcatheter aortic valve implantation; TF ¼ transfemoral approach; TV ¼ tricuspid valve; TVR ¼ tricuspid valve replacement; VF ¼ ventricular fibrillation.
the survival was poor if ECMO was placed after a delayed period after cardiac surgery (p ¼ 0.03).
Postoperative Complications Numerous postoperative complications were encountered in the study group, as shown in Table 3.
factors that determined the survival in the study group (Tables 4–6).
Causes of Death The cause of death in most patients was multifactorial. Cardiac causes accounted for 30 deaths (88.2%). Multiorgan failure occurred in 13 patients (38.2%) and septicemia in 6 (17.6%).
Postoperative Survival Twenty-one patients (46.7%) died during ECMO support, including patients for whom support was withdrawn owing to futility. Twenty four patients (53.3%) survived initial weaning of ECMO support; however, 13 of these patients subsequently died as inpatients (inhospital mortality 75.6%). Eleven patients (24.4%) were discharged from hospital. Using univariate analysis, we found clinical, biochemical, and hematologic
Late Follow-Up Eleven patients were discharged from hospital. Mean follow-up was 3.2 years (median 2 years; Fig 1). Of the 11 patients who were discharged from hospital alive, all survived at least 1 year. There were 3 late deaths, all due to end-stage heart failure. The remaining 8 patients were alive at the time of follow-up. Among those discharged from hospital, survival was 69% at 3 years and 51% at 5
4
SAXENA ET AL POSTCARDIOTOMY ECMO SUPPORT IN THE ELDERLY
Table 3. Postoperative Complications Postoperative Complication
No. of Patients
Acute kidney injury Hepatic failure Atrial fibrillation Septicemia Pneumonia Adult respiratory distress syndrome Respiratory failure Delirium Cerebrovascular accident Intracranial hemorrhage Limb ischemia Compartment syndrome Coronary graft thrombosis VT/VF/cardiac arrest Bleeding requiring reexploration Gastrointestinal bleeding Gastric ischemia and necrosis Ischemic bowel Peritonitis Pseudomembranous enterocolitis Acute cholecystitis Intraperitoneal hemorrhage Sternal wound infection Mediastinitis
20 2 11 11 12 1 3 3 2 2 6 2 1 8 7 6 1 3 1 2 1 1 4 1
(44.4) (4.4) (24.4) (24.4) (26.7) (2.2) (5.5) (5.5) (4.4) (4.4) (13.3) (4.4) (2.2) (17.8) (15.6) (13.3) (2.2) (5.5) (2.2) (4.4) (2.2) (2.2) (8.9) (2.2)
Values are n (%). VF ¼ ventricular fibrillation;
VT ¼ ventricular tachycardia.
Ann Thorac Surg 2015;-:-–-
years. At the time of last follow-up, 3 patients were in NYHA class I, 4 in class II, 3 in class III, and 1 in class IV.
Comparison With Younger Population Requiring ECMO Support We compared the study population with a younger group of patients (n ¼ 48) who required ECMO support after cardiac surgery during the same period as the patients aged 70 years and more. When comparing the groups, there was a statistically significant difference in survival to discharge, with the younger cohort having better survival (p ¼ 0.05). Also significant was that more deaths were attributed to multiorgan failure in the group aged 50 to 69 years than in the group aged 70 years and more (p < 0.001). When comparing the survival differences after discharge, the 1-year survival was not statistically different, with 100% of the 70 and older group alive and 81% of the 50 to 69 group alive (p ¼ 0.75; Fig 2). Other areas of comparison, including discharge location, NYHA class at discharge and follow-up, and cause of death from cardiac or neurologic causes were not statistically different between groups (Table 7).
Comment Approximately 1% of adult cardiac surgery patients have postcardiotomy cardiogenic shock [6, 7]. It has been suggested that ECMO may be less suitable for older patients [8–10]. We did a comparison of outcome between older patients (the study cohort) and younger patients
Table 4. Clinical Characteristics of Patients Who Survived to Discharge Compared With Patients Who Died in Hospital After Extracorporeal Membrane Oxygenation Support Died in Hospital
Survived to Discharge
Variable
n
%
n
%
p Value
Age group 70–80 years Age group 80 years Atrial fibrillation Preoperative IABP Preoperative CHF Preoperative cardiogenic shock Preoperative CKI COPD Diabetes mellitus Hypertension Previous stroke Redo cardiac surgery Coronary artery disease Aortic stenosis Tricuspid regurgitation Valve repair/replacement Valve þ CABG Primary cardiac surgical procedure, other
28 8 13 7 21 6 12 10 14 27 8 21 23 15 10 12 9 10
77.778 22.222 36.111 19.444 58.333 16.667 33.333 27.778 38.889 75 22.222 58.333 63.889 41.667 27.778 33.333 25 27.778
7 4 0 0 4 1 0 1 3 6 1 5 6 6 1 4 4 2
63.636 36.364 0 0 36.364 9.091 0 9.091 27.273 54.545 9.091 45.455 54.545 54.545 9.091 36.364 36.364 18.182
0.4352
a
0.0215a 0.1751 0.2012 1 0.0438a 0.416 0.7218 0.2628 0.6631 0.505 0.7258 0.505 0.416 1 0.467 0.7027
Statistically significant (p < 0.05).
CABG ¼ coronary artery bypass graft surgery; CHF ¼ congestive heart failure; pulmonary disease; IABP ¼ intraaortic balloon pump.
CKI ¼ chronic kidney injury;
COPD ¼ chronic obstructive
Ann Thorac Surg 2015;-:-–-
5
SAXENA ET AL POSTCARDIOTOMY ECMO SUPPORT IN THE ELDERLY
Table 5. Comparison of Survivors and Nonsurvivors in Terms of Hematologic and Biochemical Variables Perioperatively Died Variable
n
Baseline BUN Baseline Cr Baseline platelet count Mean postop ALT Mean postop AST Mean postop total bilirubin Mean postop BUN Mean postop C3 Mean postop Cr Mean postop CRP Mean postop lactate Mean postop platelet count Max postop ALT Max postop AST Max postop total bilirubin Max postop C3 Max postop Cr Max postop CRP Max postop lactate Max postop platelet count
22 29 29 34 34 34 34 34 34 34 34 34 34 34 34 34 34 34 34 34
a
Survived to Discharge
Mean SD
n
6 9 9 11 11 11 11 11 11 11 11 11 11 11 11 11 11 11 11 11
32 1.44 233.24 191.94 663.13 4.91 39.99 29.35 1.91 59.12 4.4 72.79 616.53 2,126.56 12.14 63.91 2.95 169.6 10.33 119.06
13.57 0.57 103.24 380.51 1507.32 6.3 12.55 22.98 0.79 48.92 4.62 40.49 891.67 3,991.99 13.52 28.86 1.08 101.99 6.71 96.18
Mean SD
p Value
0.0096a 0.1675 0.7573 0.0461a 0.0042a 0.0522 0.4359
Background. We conducted a retrospective study to assess whether providing extracorporeal membrane oxygenation (ECMO) support to elderly patients (aged 70 years or more) who failed separation from cardiopulmonary bypass after cardiac surgery was a viable option. Methods. From 2003 to 2013, 45 patients aged 70 years or more underwent 47 runs of ECMO postoperatively. Results. There were 31 men (68.9%). The mean age was 76.8 years. Five patients were in cardiogenic shock preoperatively. Forty-four patients required venoarterial ECMO support for cardiogenic shock. Mean duration of support was 103.8 ± 74.3 hours. Twenty-one patients (46.6%) died while on ECMO support. Twenty-four patients (53.3%) were weaned off ECMO initially, and 11 patients were discharged from hospital. Inhospital mortality was 75.6%. Postoperative complications included
acute kidney injury in 30 patients (44.4%), pneumonia in 12 (26.7%), and sepsis in 11 (24.4%). There were 30 deaths (88.2%) attributable to cardiac causes. Preoperative atrial fibrillation, chronic kidney injury, lactic acidosis on ECMO support, and persistent coagulopathy were associated with higher mortality. Conclusions. Postcardiotomy ECMO support in elderly patients is associated with high postoperative morbidity and mortality. Nevertheless, it often provides the last line of therapy for these critically ill patients and may provide positive outcomes in selected subgroups.
E
from January 2003 to January 2013 (Institutional Review Board reference number 13-001149).
xtracorporeal membrane oxygenation (ECMO) support can be used effectively for postcardiotomy cardiopulmonary failure. Cardiogenic shock after cardiac surgery in elderly patients has been associated with high postoperative mortality [1]. The rate of successful ECMO removal in adult cardiac surgical series varies from 30% to 60% and is associated with inhospital mortality of 60% to 80% [2–5]. Literature on the use of ECMO in elderly cardiac surgical patients is limited. We reviewed our 10 years of experience with patients aged 70 years or more who had ECMO support after cardiac surgery to assess the futility or viability of this approach.
(Ann Thorac Surg 2015;-:-–-) Ó 2015 by The Society of Thoracic Surgeons
Cardiac Surgery Cardiac surgery was performed through either a median sternotomy or a thoracotomy. The chest was left open at the end of the procedure if there was significant bleeding due to coagulopathy or myocardial edema. The decision to proceed to ECMO was based on failure to respond to the standard hemodynamic support. Intraaortic balloon pump was placed either preoperatively or intraoperatively while attempting to wean the patient off CPB. Reexploration of the chest was undertaken during the ECMO support as clinically indicated.
Material and Methods The study was conducted at the Mayo Clinic, Rochester, Minnesota, and included patients who were aged 70 years or older and required postcardiotomy ECMO support Accepted for publication Nov 26, 2014. Address correspondence to Dr Schears, Department of Anesthesiology and Critical Care Medicine, Mayo Clinic, 200 First St SW, Rochester, MN 55905; e-mail: [email protected].
Ó 2015 by The Society of Thoracic Surgeons Published by Elsevier
ECMO Circuits A standard ECMO circuit was used. The circuits were primed with Plasmalyte A (Abbott, Chicago, IL) first, and then the clear prime was displaced in a transfer bag by packed red blood cells to maintain a hemoglobin level greater than 9 g/dL to 10 g/dL, 500 USP heparin, and 50 mL sodium bicarbonate. Usually 1 to 2 units of packed red blood cells were used to displace the clear prime. 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2014.11.075
2
SAXENA ET AL POSTCARDIOTOMY ECMO SUPPORT IN THE ELDERLY
Ann Thorac Surg 2015;-:-–-
ECMO Management
Table 1. Baseline Clinical Characteristics of Study Population
Heparin was usually given in the dose of 5,000 USP if the patient was not on CPB. The ECMO circuit flow was maintained at a cardiac index of 2.0 to 2.4 L $ min1 $ m2 as allowed by the cannula size and the patient’s volume status, with an aim to keep mixed venous saturation greater than 60%. Heparin infusion was started in the postoperative period when the chest tube output was less than 1 to 2 cc/kg per hour for the first few hours with normalized coagulation values. Heparin infusion was typically started within the first 12 hours after ECMO initiation. Anticoagulation was measured using either activated clotting time (POC iStat; Abbott), aiming for a range of 140 s to 170 s, or activated partial thromboplastin time, maintained in the range of 50 s to 80 s. Weaning off ECMO support was based on the clinical and investigative evidence of recovery of cardiac and pulmonary function. Anticoagulation therapy was adjusted to maintain activated clotting time greater than 200 s or activated partial thromboplastin time of 80 s or more at the time of weaning. The oxygenator’s blender gas flow and fraction of inspired oxygen (FiO2) was adjusted to maintain PaO2 greater than 150 mm Hg and PCO2 between 35 mm Hg and 45 mm Hg.
Clinical Characteristics
Statistical Methods Categoric variables are reported as frequencies and percentages, and continuous variables are reported as mean SD or median (range) as appropriate. Categoric variables were compared between patients who survived ECMO and those who did not using Fisher’s exact test, and continuous variables were compared using twosample t test or Wilcoxon rank sum test where appropriate. The Kaplan-Meier method was used to evaluate survival. All statistical tests were two sided, with the alpha level set at 0.05 for statistical significance.
Results Baseline Clinical Characteristics We identified 45 patients who underwent 47 runs of ECMO support after cardiac surgery (Table 1). Forty-one patients underwent primary cardiac surgery through median sternotomy, 3 had thoracotomies, and 1 patient had transfemoral transcatheter aortic valve replacement. Hypothermic circulatory arrest was utilized in 6 patients (mean circulatory arrest time 24.2 minutes; range, 11 to 44). Mean left ventricular ejection fraction was 48.5%. Twelve patients had a preoperative left ventricular ejection fraction of 30% or less.
Indications for ECMO Institution The ECMO was instituted when there was failure to wean patients from CPB despite maximal inotropic or intraaortic balloon pump support or both, cardiogenic shock with systolic blood pressure 85 mm Hg or less, cardiac index less than 1.5 L $ min1 $ m2, cardiac arrest in the operating room or in the ICU, right ventricular failure with or without concomitant left ventricular failure based
Value
Number of patients Age, years Sex Male Female Body mass index, kg/m2 (n ¼ 44) Chronic obstructive pulmonary disease Former smokers Current smokers Diabetes mellitus Hypertension Acute kidney injury Chronic kidney injury Peripheral vascular disease Preoperative cerebrovascular accident Recent myocardial infarction Previous myocardial infarction Preoperative cardiogenic shock Preoperative CPR Preoperative congestive heart failure Preoperative intraaortic balloon pump Preoperative LVAD Elective operation Urgent operation Emergent operation Redo cardiac surgery
45 76.8 4.6 31 (68.9) 14 (31.1) 29.5 5.1 10 (22.2) 8 (17.8) 1 (2.2) 16 (35.6) 31 (68.9) 5 (11.1) 12 (26.7) 7 (15.6) 5 (11.1) 5 (11.1) 6 (13.3) 5 (11.1) 5 (11.1) 24 (53.3) 6 (13.3) 1 (2.2) 35 (77.8) 6 (13.3) 4 (8.9) 25 (55.6)
Values are mean SD or n (%), unless otherwise indicated. CPR ¼ cardiopulmonary resuscitation; device.
LVAD ¼ left ventricular assist
on the information obtained from Swan-Ganz catheter, and intraoperative transesophageal echo (TEE). Fortyfour patients (97.8%) received ECMO to provide mechanical circulatory support. One patient had ECMO support to decompress the bleeding right ventricle during redo cardiac surgery.
Intraoperative Complications During Cardiac Surgery The primary cardiac surgical procedure was complicated intraoperatively in 26 patients, and possibly contributed to the need for ECMO support (Table 2).
ECMO Data The ECMO support was instituted using peripheral cannulation in 15 patients (33%) and central cannulation in 30 (66%) patients. Venoarterial ECMO was utilized in 44 patients (97.8%), and a right ventricular assist device with oxygenator was used in 1 patient. Mean duration of ECMO support for the whole group was 103.8 74.3 hours (range, 1.0 to 378.5). Mean duration of ECMO support for the nonsurvivors was 106.8 81.8 hours; the survivors were supported for 93.9 41.3 hours (p > 0.99). ECMO was begun after a lag period due to delayed appearance of cardiogenic shock in 19 patients (42.2%);
Ann Thorac Surg 2015;-:-–-
SAXENA ET AL POSTCARDIOTOMY ECMO SUPPORT IN THE ELDERLY
3
Table 2. Intraoperative Events During Primary Cardiac Surgery and Outcome Pt. Age (years)
Primary Cardiac Surgical Procedure
71 85 71 83 75 75
Aortic root replacement CABG Redo AVR CABG AVR AVR, aortic root enlargement, CABG
76 70 72 77 76 73 71 82 79 79 84 81
Pericardiectomy Robotic MV repair CABG AVR, MVR, CABG MVR AVR, refashioning vein graft Maze, LA thrombectomy MVR, TV repair, CABG Repair of descending thoracic aorta Aortic root replacement, MV repair TAVI TA on bypass TVR and PVR, intraop stent placed in LM
84
Ascending aortic replacement, AVR
75 80
AVR, refashioning vein graft AVR
77
AVR, MVR, CABG
82
AVR, TV repair
79 79
TAVI TF AVR, valve retrieval from LV after failed TAVI, aortoplasty, CABG 1 AVR, ascending aorta replacement
74
Intraoperative Event
Outcome
RCA graft kink Revision of grafts Cardiac injury during resternotomy Poor coronary targets Vein graft injury during redo sternotomy Cardiac injury to LIMA and RV during redo sternotomy, aortic prosthesis reimplanted Failure to graft coronary arteries RCA thrombus Regrafting of coronary arteries Reimplantation of aortic valve, heavily calcific annulus Cardiac injury during redo sternotomy Coronary graft vasospasm Injury to RCA Bleeding from friable RV Cardiac failure, coagulopathy Myocardial protection issues VF during aortic root injection during TAVI Failure to graft coronaries and had intraoperative stent placement in LM Friable aorta bleeding from aortotomy requiring replacement of aorta Protamine reaction Heavily calcified aortic annulus and calcific atheroma of aorta Heavily calcified aorto-mitral calcification, mitral valve damage during debridement Bleeding requiring repair on multiple runs of CPB and cross clamping Intraoperative embolization to RCA during TAVI Valve embolization, failed TAVI, LIMA injury during sternal reentry Intraoperative anterior wall LV rupture
Survived Survived Died Died Died Died Died Survived Died Died Died Died Died Died Died Died Died Died Died Died Survived Died Died Died Survived Died
AVR ¼ aortic valve replacement; CABG ¼ coronary artery bypass graft surgery; CPB ¼ cardiopulmonary bypass; intraop ¼ intraoperative; LA ¼ left atrium; LIMA ¼ left internal mammary artery; LM ¼ left main coronary artery; LV ¼ left ventricle; MVR ¼ mitral valve replacement; Pt. ¼ patient; RCA ¼ right coronary artery; RV ¼ right ventricle; TAVI ¼ transcatheter aortic valve implantation; TF ¼ transfemoral approach; TV ¼ tricuspid valve; TVR ¼ tricuspid valve replacement; VF ¼ ventricular fibrillation.
the survival was poor if ECMO was placed after a delayed period after cardiac surgery (p ¼ 0.03).
Postoperative Complications Numerous postoperative complications were encountered in the study group, as shown in Table 3.
factors that determined the survival in the study group (Tables 4–6).
Causes of Death The cause of death in most patients was multifactorial. Cardiac causes accounted for 30 deaths (88.2%). Multiorgan failure occurred in 13 patients (38.2%) and septicemia in 6 (17.6%).
Postoperative Survival Twenty-one patients (46.7%) died during ECMO support, including patients for whom support was withdrawn owing to futility. Twenty four patients (53.3%) survived initial weaning of ECMO support; however, 13 of these patients subsequently died as inpatients (inhospital mortality 75.6%). Eleven patients (24.4%) were discharged from hospital. Using univariate analysis, we found clinical, biochemical, and hematologic
Late Follow-Up Eleven patients were discharged from hospital. Mean follow-up was 3.2 years (median 2 years; Fig 1). Of the 11 patients who were discharged from hospital alive, all survived at least 1 year. There were 3 late deaths, all due to end-stage heart failure. The remaining 8 patients were alive at the time of follow-up. Among those discharged from hospital, survival was 69% at 3 years and 51% at 5
4
SAXENA ET AL POSTCARDIOTOMY ECMO SUPPORT IN THE ELDERLY
Table 3. Postoperative Complications Postoperative Complication
No. of Patients
Acute kidney injury Hepatic failure Atrial fibrillation Septicemia Pneumonia Adult respiratory distress syndrome Respiratory failure Delirium Cerebrovascular accident Intracranial hemorrhage Limb ischemia Compartment syndrome Coronary graft thrombosis VT/VF/cardiac arrest Bleeding requiring reexploration Gastrointestinal bleeding Gastric ischemia and necrosis Ischemic bowel Peritonitis Pseudomembranous enterocolitis Acute cholecystitis Intraperitoneal hemorrhage Sternal wound infection Mediastinitis
20 2 11 11 12 1 3 3 2 2 6 2 1 8 7 6 1 3 1 2 1 1 4 1
(44.4) (4.4) (24.4) (24.4) (26.7) (2.2) (5.5) (5.5) (4.4) (4.4) (13.3) (4.4) (2.2) (17.8) (15.6) (13.3) (2.2) (5.5) (2.2) (4.4) (2.2) (2.2) (8.9) (2.2)
Values are n (%). VF ¼ ventricular fibrillation;
VT ¼ ventricular tachycardia.
Ann Thorac Surg 2015;-:-–-
years. At the time of last follow-up, 3 patients were in NYHA class I, 4 in class II, 3 in class III, and 1 in class IV.
Comparison With Younger Population Requiring ECMO Support We compared the study population with a younger group of patients (n ¼ 48) who required ECMO support after cardiac surgery during the same period as the patients aged 70 years and more. When comparing the groups, there was a statistically significant difference in survival to discharge, with the younger cohort having better survival (p ¼ 0.05). Also significant was that more deaths were attributed to multiorgan failure in the group aged 50 to 69 years than in the group aged 70 years and more (p < 0.001). When comparing the survival differences after discharge, the 1-year survival was not statistically different, with 100% of the 70 and older group alive and 81% of the 50 to 69 group alive (p ¼ 0.75; Fig 2). Other areas of comparison, including discharge location, NYHA class at discharge and follow-up, and cause of death from cardiac or neurologic causes were not statistically different between groups (Table 7).
Comment Approximately 1% of adult cardiac surgery patients have postcardiotomy cardiogenic shock [6, 7]. It has been suggested that ECMO may be less suitable for older patients [8–10]. We did a comparison of outcome between older patients (the study cohort) and younger patients
Table 4. Clinical Characteristics of Patients Who Survived to Discharge Compared With Patients Who Died in Hospital After Extracorporeal Membrane Oxygenation Support Died in Hospital
Survived to Discharge
Variable
n
%
n
%
p Value
Age group 70–80 years Age group 80 years Atrial fibrillation Preoperative IABP Preoperative CHF Preoperative cardiogenic shock Preoperative CKI COPD Diabetes mellitus Hypertension Previous stroke Redo cardiac surgery Coronary artery disease Aortic stenosis Tricuspid regurgitation Valve repair/replacement Valve þ CABG Primary cardiac surgical procedure, other
28 8 13 7 21 6 12 10 14 27 8 21 23 15 10 12 9 10
77.778 22.222 36.111 19.444 58.333 16.667 33.333 27.778 38.889 75 22.222 58.333 63.889 41.667 27.778 33.333 25 27.778
7 4 0 0 4 1 0 1 3 6 1 5 6 6 1 4 4 2
63.636 36.364 0 0 36.364 9.091 0 9.091 27.273 54.545 9.091 45.455 54.545 54.545 9.091 36.364 36.364 18.182
0.4352
a
0.0215a 0.1751 0.2012 1 0.0438a 0.416 0.7218 0.2628 0.6631 0.505 0.7258 0.505 0.416 1 0.467 0.7027
Statistically significant (p < 0.05).
CABG ¼ coronary artery bypass graft surgery; CHF ¼ congestive heart failure; pulmonary disease; IABP ¼ intraaortic balloon pump.
CKI ¼ chronic kidney injury;
COPD ¼ chronic obstructive
Ann Thorac Surg 2015;-:-–-
5
SAXENA ET AL POSTCARDIOTOMY ECMO SUPPORT IN THE ELDERLY
Table 5. Comparison of Survivors and Nonsurvivors in Terms of Hematologic and Biochemical Variables Perioperatively Died Variable
n
Baseline BUN Baseline Cr Baseline platelet count Mean postop ALT Mean postop AST Mean postop total bilirubin Mean postop BUN Mean postop C3 Mean postop Cr Mean postop CRP Mean postop lactate Mean postop platelet count Max postop ALT Max postop AST Max postop total bilirubin Max postop C3 Max postop Cr Max postop CRP Max postop lactate Max postop platelet count
22 29 29 34 34 34 34 34 34 34 34 34 34 34 34 34 34 34 34 34
a
Survived to Discharge
Mean SD
n
6 9 9 11 11 11 11 11 11 11 11 11 11 11 11 11 11 11 11 11
32 1.44 233.24 191.94 663.13 4.91 39.99 29.35 1.91 59.12 4.4 72.79 616.53 2,126.56 12.14 63.91 2.95 169.6 10.33 119.06
13.57 0.57 103.24 380.51 1507.32 6.3 12.55 22.98 0.79 48.92 4.62 40.49 891.67 3,991.99 13.52 28.86 1.08 101.99 6.71 96.18
Mean SD
p Value
0.0096a 0.1675 0.7573 0.0461a 0.0042a 0.0522 0.4359
