- Email: [email protected]
An extracorporeal membrane oxygenation–based approach to cardiogenic shock in an older population
An Extracorporeal Membrane Oxygenation–Based Approach to Cardiogenic Shock in an Older Population Christopher Smith, FRACS, Rinaldo Bellomo, MD, Jai S. Raman, FRACS, George Matalanis, FRACS, Alexander Rosalion, FRACS, Jonathan Buckmaster, MD, Graeme Hart, MD, William Silvester, MD, Geoffrey A. Gutteridge, FANZCA, Ben Smith, CCP, Laurie Doolan, MD, and Brian F. Buxton, FRACS Departments of Cardiothoracic Surgery, Intensive Care, and Anaesthesia, Austin & Repatriation Medical Centre, Melbourne, and Warringal Private Hospital, Melbourne, Australia
Background. We investigated the efficacy of an integrated system of advanced supportive care based on extracorporeal membrane oxygenation (ECMO) in older patients with an estimated mortality of more than 90% to establish whether its use is justifiable. Methods. Treatment was provided by cardiac surgeons and critical care physicians and included the following key elements: (1) ECMO, (2) early application of continuous venovenous hemofiltration, (3) inhaled nitric oxide, (4) maintenance of perfusion pressure with norepinephrine, (5) maintenance of pulmonary blood flow by ventricular filling with intravenous colloids, (6) avoidance of early postoperative anticoagulation, (7) frequent use of transesophageal echocardiography, and (8) low tidal volume ventilation. Demographic features, intraoperative details, postoperative course, ECMO weaning rate, mor-
bidity, survival to hospital discharge, and the quality of life of survivors were recorded. Results. Seventeen consecutive patients (median age, 69 years) with refractory cardiogenic shock were studied. The median duration of ECMO was 86 hours (20 to 201 hours). Eleven patients (65%) were successfully weaned from ECMO. Seven patients (41%) survived to discharge. The major causes of morbidity were bleeding and leg ischemia. All patients who survived to discharge were alive and well at follow-up (median, 21 months) and reported a satisfactory quality of life. Conclusions. An ECMO-based approach can be used with acceptable results in the treatment of refractory cardiogenic shock, even in older patients. (Ann Thorac Surg 2001;71:1421–7) © 2001 by The Society of Thoracic Surgeons
V
benefits, and the potential of serious disability among survivors. We hypothesized that, even in these patients, ECMO could be used with acceptable outcomes using an integrated management strategy. Thus, we applied this strategy to a group of patients who, in our institution, had a predicted mortality of more than 90%.
enoarterial extracorporeal membrane oxygenation (ECMO) is a technique aimed at providing prolonged but temporary cardiac or respiratory support. The initial results of ECMO for adult respiratory distress syndrome were poor [1]. However, several subsequent studies have produced improved results in both respiratory failure [2] and cardiogenic shock [3– 6]. Although the majority of patients can be weaned from cardiopulmonary bypass with little or no inotropic support, up to 3% to 5% develop severe postcardiotomy myocardial dysfunction [6]. Most of these patients can be weaned with the aid of higher doses of inotropic agents and intraaortic balloon counterpulsation [6]. However, those who have persistent shock despite these measures face almost certain death. In this small group of patients, ECMO may provide sufficient hemodynamic support to allow recovery from reversible myocardial injury [7]. This use of ECMO, especially when applied to older patients, is open to criticism because of its cost, its unproved Accepted for publication Oct 17, 2000. Address reprint requests to Dr Bellomo, Department of Intensive Care, Austin & Repatriation Medical Centre, Studley Rd, Heidelberg, Victoria 3084, Australia; e-mail: [email protected].
© 2001 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
Material and Methods Study Group Between June 1996 and December 1998, ECMO was instituted in 17 consecutive patients with cardiogenic shock. The indications for establishing extracorporeal support in these patients were the following: 1. Failure to wean from cardiopulmonary bypass despite maximal inotropic and mechanical support. 2. Failure to maintain satisfactory systemic perfusion (systolic blood pressure ⬍ 80 mm Hg, cardiac inThis article has been selected for the open discussion forum on the STS Web site: http://www.sts.org/section/atsdiscussion/
0003-4975/01/$20.00 PII S0003-4975(00)02504-2
1422
SMITH ET AL ECMO IN OLDER PATIENTS
dex ⬍ 2.0 L 䡠 min⫺1 䡠 m⫺2, base excess ⬎ ⫺8 mEq/L, and anuria) despite the use of high-dose inotropic agents and intraaortic balloon counterpulsation, in the absence of a surgically correctable cause such as tamponade. In these patients, cardioplegia was administered in 800 mL of blood and contained 100 mL of tris(hydroxymethyl)aminomethane (Abbott, Australasia, NSW, Australia), 100 mg of lidocaine HCl, 20 mmol of KCl, and 10 mmol of MgCl2. Cardioplegia was delivered by the antegrade route and the coronary sinus cannula (approximately 500 mL at a temperature of 14°C to 18°C). Further cardioplegia solution was infused through the coronary sinus cannula at approximately 15-minute intervals. This consisted of 600 mL of blood with 5 mmol of KCl and 20 mmol of MgCl2 at a temperature of 14°C to 18°C. A “hot shot” was given through the aortic root immediately before aortic cross-clamp removal. This consisted of 600 mL of blood with 5 to 10 mmol of MgCl2 at 34°C to 36°C.
Extracorporeal Membrane Oxygenation The ECMO circuit consisted of a closed, Carmeda Bioactive Surface-coated circuit of polyvinyl chloride tubing (Medtronic, Minneapolis, MN) and a centrifugal pump (Biomedicus Biopump, Medtronic, Minneapolis, MN), which propelled blood through a hollow-fiber membrane oxygenator (Maxima P.R.F., Medtronic). Blood flows were monitored by a Doppler flow probe placed on the arterial side of the circuit. Cannulation was peripheral (femorofemoral) in 11 patients, using Biomedicus Carmeda Bioactive Surface-coated ECMO cannulas placed by direct cutdown into the femoral vessels; and central (aortoatrial) in 5 patients, using in situ cardiopulmonary bypass cannulas. No left atrial or ventricular drainage was used. In one further patient with central arterial cannulation, the venous cannula was moved to the femoral vein because of right atrial bleeding. Limb ischemia was a major problem early on in the study. Therefore, two methods of optimizing distal limb perfusion were instituted. One used a distal cannula passing from the side port of the arterial cannula to a point on the femoral artery distal to the cannulation site. The other was based on the cannulation of the femoral artery by means of a polyethylene terephthalate T graft.
Integrated Management Strategy Pump flows were chosen to supply adequate systemic circulatory support (2.0 to 2.5 L 䡠 min⫺1 䡠 m⫺2). Pulsatile pulmonary artery flow was maintained by means of right ventricular filling with intravenous colloids. The aim was to minimize the risk of intracardiac clot and reperfusion lung injury. A continuous cardiac output pulmonary artery catheter was used to assess pulmonary blood flow. Inotropic support was reduced or withdrawn to decrease myocardial oxygen demand and facilitate myocardial recovery. Norepinephrine was infused to maintain a mean systemic arterial pressure of more than 70 to
Ann Thorac Surg 2001;71:1421–7
75 mm Hg. To oxygenate any blood delivered to the lungs by the patient’s heart and to minimize atelectasis and lung injury, low tidal volume (5 to 6 mL/kg tidal breaths) mechanical ventilation with positive end-expiratory pressure at 8 to 10 cm H2O was applied. The activated clotting time was monitored, with the aim of ensuring an activated clotting time of approximately 140 to 150 seconds with full flows, and an activated clotting time of more than 200 seconds when flows were reduced less than 1.5 L/min during the weaning process. Early in our study, systemic heparinization was used in three patients; however, the remaining 14 patients were supported without any heparin except at low flows. Independent of the activated clotting time, no heparin was used if the patient’s chest drainage was more than 100 mL/h. If chest drainage was less than 100 mL/h for 6 to 8 hours, correction of the activated clotting time to the desired levels by means of heparin was considered. Intraaortic balloon counterpulsation was used in the majority of patients (14 of 17). Continuous venovenous hemofiltration was used in all 17 patients and was commenced early (within the first 8 hours) in the postoperative course. Continuous venovenous hemofiltration was used to regulate intravascular volume and overall fluid balance and to enable the rapid administration of blood and blood products without the induction of volume overload. To decrease right ventricular afterload, inhaled nitric oxide was used in 6 patients in whom there was a component of right ventricular dysfunction. Transesophageal echocardiography was used to guide the placement of the venous cannulas in patients with peripheral access. Serial transesophageal echocardiography was used during the period of extracorporeal support to enable the progressive assessment of myocardial recovery. It was also used to exclude intracardiac clot or clots within the cannulas, and to provide useful information regarding myocardial contractility and ventricular filling during the weaning process.
General Patient Care In all patients, enteral feeding was instituted in the first 24 hours and continued if tolerated. Muscle relaxants were only used for a short period (⬍ 8 hours) after admission to the cardiothoracic intensive care unit or to facilitate invasive procedures. Sedation was morphineand midazolam-based and used to achieve patient comfort while enabling daily neurologic assessment. Native urine output, if present, was maintained at more than 1 mL 䡠 kg⫺1 䡠 h⫺1 with a furosemide infusion. Packed red blood cells were transfused to maintain a hemoglobin concentration of approximately 100 g/L. Antibiotics were used if infection was clinically suspected. Hyperglycemia was controlled by continuous infusion of insulin. Atrial fibrillation with a fast ventricular response was controlled by the continuous intravenous infusion of amiodarone. The lower limbs were formally assessed for ischemia every 6 hours by physical examination, Doppler examination, and 6-hourly creatine kinase measurements.
Ann Thorac Surg 2001;71:1421–7
SMITH ET AL ECMO IN OLDER PATIENTS
Table 1. Clinical Pre-ECMO Features of Study Patients Age (y) 69 83 74 75 64 37 56 68 74 71 43 75 48 62 83 81 69
Sex
CPB Time (min)
X-Clamp Time (min)
IABP
Nature of Surgery
Preop NYHA
M M M M F M F M F M M M M M M F M
177 216 337 198 173 156 165 182 204 457 141 156 N/A 417 192 184 174
118 96 158 132 134 126 95 144 116 252 120 124 N/A 253 115 157 136
Y Y Y Y Y N Y Y Y Y N Y Y Y Y N Y
Elective Urgent Emergency Elective Elective Elective Urgent Elective Elective Elective Emergency Emergency N/A Elective Elective Urgent Urgent
III III IV IV III II III III III III I IV IV II III IV IV
CPB ⫽ cardiopulmonary bypass; ECMO ⫽ extracorporeal membrane oxygenation; IABP ⫽ intraaortic balloon pulsation; N/A ⫽ not applicable; preop NYHA ⫽ preoperative class according to the New York Heart Association; X-clamp ⫽ cross-clamp.
Data Collection Data collection was performed focussing on patient demographics, operative details, cannulation method, duration of support, blood product usage, use of hemofiltration, inotropic agents, counterpulsation, and nitric oxide. Data were obtained for morbidity and in-hospital outcome. All patients who survived to discharge were followed up after hospital discharge. All survivors completed a quality-of-life questionnaire. This questionnaire had been previously used at our institution to assess survivors of severe multiple organ failure [8], and was based on an activity index, mental function index, and a simplified version of the Nottingham Health Profile [9].
Results In the majority of patients, ECMO was required in the postcardiotomy setting either because of the inability to wean from cardiopulmonary bypass or refractory cardioTable 2. Operative Procedures Requiring Extracorporeal Membrane Oxygenation Procedure Coronary bypass Aortic valve/CABG Aortic root replacement Postinfarct VSD Pulmonary embolectomy CABG ⫽ coronary artery bypass grafting; defect.
No. of Patients 7 4 2 2 1 VSD ⫽ ventricular septal
1423
genic shock in the early postoperative period. This type of support was only required in 1.1% of all cardiac surgical operations during the 30-month period (mean age of overall patient population in question, 64.7 years). One patient also required ECMO for acute deterioration of dilated cardiomyopathy leading to cardiac arrest. Tables 1 and 2 summarize the clinical and operative features of the study patients. The median age of patients treated with ECMO was 69 years (range, 37 to 83 years). There were 8 patients older than 70 years of age, and 13 (76%) were male. The patients were supported for a median period of 86 hours (range, 20 to 201 hours). Only 4 patients were supported for less than 48 hours. Of these 4 patients, 2 were octogenarians who died early in the study because of severe limb ischemia and severe ongoing coagulopathic bleeding. The remaining 2 patients were weaned successfully after rapid recovery of right ventricular function. Eleven (65%) of the 17 ECMO-supported patients were successfully weaned. The median duration of support required for those patients who were successfully weaned was 106 hours. Overall, 7 patients (41%) survived to hospital discharge. Only 3 of 8 patients (38%) older than 70 years, however, were weaned from ECMO, and only 1 survived. Of the 9 patients younger than 70 years of age, 8 were successfully weaned (89%) and 6 (67%) survived. The median age of survivors was 62 years (range, 37 to 74 years) and of nonsurvivors, 74 years (range, 56 to 83 years). There were no significant differences in overall peak cardiac enzyme values between survivors and nonsurvivors. However, all patients with a peak creatine kinase of more than 3,000 IU or a peak postoperative cardiac troponin I of more than 500 ng/mL died. Mortality was caused by failure to wean because of intractable heart failure in 4 patients. Withdrawal of support occurred in 2 patients owing to severe limb ischemia and severe coagulopathic bleeding. Massive nonhemorrhagic stroke caused the death of 2 patients. Two patients who were successfully weaned died of overwhelming sepsis, one a week after weaning and the other 2 months after intensive care unit discharge. Survivors spent a median duration of 16 days (range, 4 to 34 days) postoperatively in the intensive care unit. The median postoperative hospital stay of survivors was 25 days (range, 10 to 40 days). Bleeding was a major problem in this group of patients, requiring the administration of platelets, fresh frozen plasma, and cryoprecipitate. Table 3 shows the median blood product requirement for the group. Six patients Table 3. Median Number of Blood Product Units Administered per Patient Product Platelets Fresh frozen plasma Cryoprecipitate Blood
Units 44 (14 –100) 30 (10 –70) 11 (0 –30) 26 (10 –56)
1424
SMITH ET AL ECMO IN OLDER PATIENTS
Ann Thorac Surg 2001;71:1421–7
Table 4. Postoperative Quality-of-Life Assessment (7 Patients) Employment Full-time Retired Unable to work Activities of daily living Shopping Washing self Cooking Gardening Use public transport Walk ⬎ 200 m Normal energy levels Normal sleep Personal relationships unaffected Satisfied with health Treatment worthwhile
2 4 1 6 6 6 6 6 6 6 6 6 7 7
(35%) required reoperation for excessive bleeding. In 3 patients, intraoperative aprotinin infusions were continued into the early postoperative period. Leg ischemia occurred in 4 patients (36%). One patient had limb ischemia despite the presence of a distal perfusion cannula. There was no limb ischemia in those with T-graft cannulation. All survivors were followed up at a median of 21 months after discharge with a quality-of-life questionnaire. Table 4 shows the results of the quality-of-life assessment. Six of 7 patients reported their ability to perform activities of daily living as satisfactory. Two patients had resumed full-time employment, and only 1 patient was unable to work because of illness, the remaining being retired. The majority reported normal energy levels and sleep patterns. All were satisfied with their general health, and considered their treatment to be worthwhile.
However, the mean age of the patients in these studies was more than a decade younger than in our cohort. In this investigation, we report a 41% overall survival in a group of patients with a mean age of 69 years. Importantly, this is also the first study to provide long-term follow-up of patients treated with ECMO for cardiogenic shock and to assess their quality of life. All hospital survivors remain alive at the time of follow-up, and the majority reports a satisfactory quality of life. Several surgical aspects must be considered when the decision is made to connect a patient to an ECMO circuit. For example, the configuration of arterial and venous cannulas is dependent on several factors: accessibility and state of femoral arteries and veins, condition of the aorta and right atrium, fragility of intrathoracic tissues, and extent of coagulopathy. Preexisting cannulas in the aorta and right atrium may be used to connect the patient to the ECMO circuit. This approach avoids cannulation in the groin with its risk of limb ischemia and infection but increases the chance of bleeding around central cannulas. In this setting, it may be difficult to close the sternum, because the underlying heart is quite edematous. When cardiopulmonary bypass cannulas or fresh aortic and right atrial cannulas are used, they should ideally be reinforced by extra pledgetted pursestring sutures and thoroughly secured. The cannulas and tubing may be brought out through the lower end of the wound or through separate stab incisions (Fig 1). Some allowance should also be made at the time of fixing the cannulas for recovery and resolution of myocardial edema to avoid
Comment Prolonged extracorporeal circulatory support has been previously described in the treatment of cardiogenic shock [3–7]. There are various mechanical assist devices available to provide circulatory support in this setting. These devices may be univentricular or biventricular; short-term or long-term, and ambulatory or nonambulatory. The choice of device depends on many clinical and logistic considerations. There are several potential advantages in using ECMO for short-term support, especially if cardiac transplantation is not available or appropriate. In particular, the combined biventricular and respiratory support provided by ECMO appears more desirable than that provided by ventricular assist devices in the presence of global myocardial dysfunction and markedly impaired gas exchange [10] as was the case in our patients. Recent studies suggest that in-hospital survival rates with the use of ECMO vary from 32% to 52% [3– 6].
Fig 1. Diagram showing the position and exit site of intrathoracic cannulas used for extracorporeal membrane oxygenation.
Ann Thorac Surg 2001;71:1421–7
SMITH ET AL ECMO IN OLDER PATIENTS
1425
Fig 4. Illustration of one of the techniques used to ensure maintenance of blood flow to the distal lower limb during femoral cannulation. The side cannula entering the femoral vessel distal to the cannulation site ensures flows greater than 150 mL/min. (ECMO ⫽ extracorporeal membrane oxygenation.)
Fig 2. Illustration of the preferred site for extrathoracic cannulation for extracorporeal membrane oxygenation.
traction on the cannulas. If femoral cannulation is chosen, the femoral vessels are isolated, and pursestring sutures are placed on each vessel with 5-0 Prolene
Fig 3. Diagram illustrating the appearance and position of the arterial cannula used for the return limb of the extracorporeal membrane oxygenation (ECMO) circuit.
(Ethicon, Somerville, NJ) sutures (Fig 2). It is preferable to cannulate the vein first. If the femoral artery is large, a small-bore cannula can be placed within it with little risk of distal limb ischemia (Fig 3). If the femoral artery is small or diseased, distal perfusion of the lower limb must be sought. Such perfusion may be achieved by inserting a small cannula (size 10F to 12F) distally and connecting it to the sidearm of the arterial inflow (Fig 4). Alternatively, a polyethylene terephthalate tube graft of 8-mm diameter can be anastomosed as a T graft to the femoral artery. The arterial cannula is then inserted into the graft and fastened securely (Fig 5). A distal draining cannula is placed in the femoral vein and connected to the venous
Fig 5. Illustration of the T-graft approach to maintaining both central arterial blood flow and distal lower limb perfusion. A polyethylene terephthalate graft is attached to the femoral artery, and the cannula is inserted into and tied to the graft. (ECMO ⫽ extracorporeal membrane oxygenation.)
1426
SMITH ET AL ECMO IN OLDER PATIENTS
Fig 6. Diagram illustrating the internal and external appearance of a double-lumen dialysis catheter used for continuous hemofiltration in all of our patients.
outflow cannula to reduce limb edema. Once the above precautions are taken, peripheral cannulation is our preferred means of access. The correct approach to sedation, pain relief, and muscle relaxation in these patients is unknown. We have used muscle relaxation with great caution because of the risks associated with it [11]. We have found that patient comfort and limited mobility can be achieved with a combination of morphine and midazolam given by continuous intravenous infusions at doses between 2 and 5 mg/h for each. Regular careful decrease or temporary cessation of these agents for daily neurologic assessment is desirable. Continuous venovenous hemofiltration through a double-lumen catheter (Fig 6) was used early in the postoperative course in all patients, because it allows the physician to control intravascular volume despite the need for massive and rapid transfusion of blood products. Such control prevents volume overload, cardiac edema, and pulmonary edema, and optimizes hemodynamics [12]. Excessive bleeding is common at the time the patient is placed on ECMO [6], is often life-threatening, and mandates the avoidance of anticoagulants and the rapid and massive administration of blood products. With this approach, we only experienced one death from bleeding and no episode of sudden clotting of the oxygenatiing membrane. Continuous venovenous hemofiltration can also proceed simultaneously without anticoagulation and remove plasma water at 500 to 700 mL/h to enable the safer transfusion of such products. Intracardiac clotting has been reported in 20% of patients during ECMO [6]. In our patients, despite avoiding systemic heparinization, intracardiac clotting was identified in only 1 patient. These findings suggest that with Carmeda Bioactive Surface-coated cannulas, clinicians can pursue a strategy that restricts the early administration of anticoagulants. Transesophageal echocardi-
Ann Thorac Surg 2001;71:1421–7
ography is essential before weaning to exclude the presence of intracardiac clot before native pump function is restored. Other issues of general patient care are also worth emphasizing. Early enteral nutrition appears desirable [13], as does the use of inhaled nitric oxide if right ventricular dysfunction is prominent [14]. Maintaining some lung perfusion at all times may prevent reperfusion pulmonary edema. Mechanical ventilation with low volume tidal breaths (6 mL/kg) and maintenance of alveolar patency by the use of positive end-expiratory pressure is recommended [15]. Cardiovascular support with ECMO may not ensure the maintenance of an adequate mean arterial blood pressure. Such hypotension may predispose the patient to cerebral and renal ischemia [16, 17]. Accordingly, a mean systemic blood pressure of more than 70 to 75 mm Hg was maintained with the continuous intravenous infusion of norepinephrine. All oliguric patients received continuous infusion of furosemide, titrated to maintain a urine output of more than 1 mL 䡠 kg⫺1 䡠 h⫺1 [18]. To prevent arrhythmias, we frequently administered intravenous magnesium [19] and amiodarone [20]. If such prophylaxis failed, the rate of ventricular response to atrial fibrillation was also controlled with the continuous infusion of intravenous amiodarone. Hyperglycemia was treated with insulin by continuous infusion titrated to maintain a serum glucose concentration between 7 and 10 mmol/L. Weaning patients from ECMO is often slow and difficult. We have used transesophageal echocardiography in all of our patients to study the effects of decreasing pump flows on myocardial chamber size and contractility. In all of our patients, we also used milrinone to increase myocardial contractility at the time of weaning [21]. We chose milrinone in preference to other inotropic agents because of it does not induce tachycardia, it has a minimal effect on myocardial oxygen consumption, it induces diastolic relaxation, and it increases coronary graft blood flow [22, 23]. We found milrinone safe and effective and support its use in preference to -adrenergic drugs. Milrinone also induces vasodilation and accumulates in the presence of renal dysfunction. The use of an ␣-adrenergic drug (norepinephrine) to maintain mean systemic blood pressure and dose adjustments may become necessary under such circumstances. In conclusion, this study reports the long-term outcome of the use of an integrated extracorporeal life support approach to the treatment of cardiogenic shock after cardiac operations in older patients. Previously, the mortality rate for this patient group at our institution was more than 90%. However, the use of ECMO alongside the integrated management strategy described above resulted in a 41% survival rate at approximately 2 years postoperatively, as well as a satisfactory quality of life. Accordingly, we believe that the use of an ECMO-based strategy is justified if intractable but potentially reversible cardiogenic shock is present or patients fail to wean from cardiopulmonary bypass.
Ann Thorac Surg 2001;71:1421–7
SMITH ET AL ECMO IN OLDER PATIENTS
References 1. Zapol WM, Snider MT, Hill JD, et al. Extracorporeal membrane oxygenation in severe acute respiratory failure. JAMA 1979;242:2193– 6. 2. Gattinoni L, Pesenti A, Mascheroni D, et al. Low-frequency positive pressure ventilation with extracorporeal CO2 removal in severe acute respiratory failure. JAMA 1986;256: 881– 6. 3. Reedy JE, Swartz MT, Raithel SC, et al. Mechanical cardiopulmonary support for refractory cardiogenic shock. Heart Lung 1990;19:514–23. 4. Magovern GJ, Magovern JA, Benckart DH, et al. Extracorporeal membrane oxygenation: preliminary results in patients with postcardiotomy cardiogenic shock. Ann Thorac Surg 1992;57:1462–71. 5. Muehrcke DD, McCarthy PM, Stewart RW, et al. Extracorporeal membrane oxygenation for postcardiotomy cardiogenic shock. Ann Thorac Surg 1996;61:684–91. 6. Golding LAR. Postcardiotomy mechanical support. Semin Thorac Cardiovasc Surg 1991;3:29–33. 7. Muehrcke DD, McCarthy PM, Stewart RW, et al. Complications of extracorporeal life support systems using heparinbound surfaces. The risk of intracardiac clot formation. J Thorac Cardiovasc Surg 1995;110:843–57. 8. Gopal I, Bhonagiri S, Ronco C, Bellomo R. Out of hospital outcome and quality of life in survivors of combined acute multiple organ and renal failure treated with continuous venovenous hemofiltration/hemodiafiltration. Intensive Care Med 1997;23:766–72. 9. Hunt SM, McEwan J, McKenna SP. The Nottingham Health Profile: subjective health status and medical consultations. Soc Sci Med 1981;15A:221–9. 10. Jurmann MJ, Haverich A, Demertzis S, et al. Extracorporeal membrane oxygenation (ECMO); extended indications for artificial support for artificial support for both heart and lungs. Int J Artif Org 1991;14:771– 4. 11. Segredo V, Caldwell JE, Matthay MA, Sharma ML, Gruenke
12. 13. 14. 15. 16.
17. 18. 19. 20. 21.
22.
1427
LD, Miller RD. Persistent paralysis in critically ill patients after long-term administration of vecuronium. N Engl J Med 1992;327:524– 8. Coraim FI, Pauser G, Stellwag F, et al. Positive influence on hemodynamics in postcardiac surgery patients by hemofiltration. Anesthetist 1985;34:236– 40. Zaloga G. Early enteral nutritional support improves outcome: hypothesis or fact. Crit Care Med 1999;27:259– 61. Rich GF, Murphy GD, Roos CM, Johns RA. Inhaled nitric oxide. Selective pulmonary vasodilation in cardiac surgical patients. Anesthesiology 1993;78:1028–35. Artigas A, Bernard GR, Carlet J, et al. The AmericanEuropean consensus conference on ARDS, Part 2. Intensive Care Med 1998;24:378–98. Gold JP, Charlson ME, Williams-Russo, et al. Improvement of outcomes after coronary artery bypass. A randomized trial comparing intraoperative high versus low mean arterial pressure. J Thorac Cardiovasc Surg 1996;110:1302–11. Bersten AD, Holt AW. Vasoactive drugs and the importance of renal perfusion pressure. New Horiz 1995;3:650– 61. Martin SJ, Danziger LH. Continuous infusion of loop diuretics in the critically ill: a review of the literature. Crit Care Med 1994;22:1323–9. England MR, Gordon G, Salem M, Chernow B. Magnesium administration and dysrhythmias after cardiac surgery. JAMA 1992;268:2395– 402. Daoud EG, Strickberger AS, Chng Man K, et al. Preoperative amiodarone as prophylaxis against atrial fibrillation after cardiac surgery. N Engl J Med 1997;337:1785–91. Doolan LA, Jones E, Kalman J, Buxton BF, Tonkin AM. A placebo-controlled trial verifying the efficacy of milrinone in weaning high-risk patients from cardiopulmonary bypass. J Cardiothorac Vasc Anesth 1997;11:37– 41. Feneck RO, and the European Milrinone Multicentre Trial Group. Intravenous milrinone following cardiac surgery: effect of bolus infusion followed by variable dose maintenance infusion. J Cardiothorac Vasc Anes 1992;6:554– 62.
Background. We investigated the efficacy of an integrated system of advanced supportive care based on extracorporeal membrane oxygenation (ECMO) in older patients with an estimated mortality of more than 90% to establish whether its use is justifiable. Methods. Treatment was provided by cardiac surgeons and critical care physicians and included the following key elements: (1) ECMO, (2) early application of continuous venovenous hemofiltration, (3) inhaled nitric oxide, (4) maintenance of perfusion pressure with norepinephrine, (5) maintenance of pulmonary blood flow by ventricular filling with intravenous colloids, (6) avoidance of early postoperative anticoagulation, (7) frequent use of transesophageal echocardiography, and (8) low tidal volume ventilation. Demographic features, intraoperative details, postoperative course, ECMO weaning rate, mor-
bidity, survival to hospital discharge, and the quality of life of survivors were recorded. Results. Seventeen consecutive patients (median age, 69 years) with refractory cardiogenic shock were studied. The median duration of ECMO was 86 hours (20 to 201 hours). Eleven patients (65%) were successfully weaned from ECMO. Seven patients (41%) survived to discharge. The major causes of morbidity were bleeding and leg ischemia. All patients who survived to discharge were alive and well at follow-up (median, 21 months) and reported a satisfactory quality of life. Conclusions. An ECMO-based approach can be used with acceptable results in the treatment of refractory cardiogenic shock, even in older patients. (Ann Thorac Surg 2001;71:1421–7) © 2001 by The Society of Thoracic Surgeons
V
benefits, and the potential of serious disability among survivors. We hypothesized that, even in these patients, ECMO could be used with acceptable outcomes using an integrated management strategy. Thus, we applied this strategy to a group of patients who, in our institution, had a predicted mortality of more than 90%.
enoarterial extracorporeal membrane oxygenation (ECMO) is a technique aimed at providing prolonged but temporary cardiac or respiratory support. The initial results of ECMO for adult respiratory distress syndrome were poor [1]. However, several subsequent studies have produced improved results in both respiratory failure [2] and cardiogenic shock [3– 6]. Although the majority of patients can be weaned from cardiopulmonary bypass with little or no inotropic support, up to 3% to 5% develop severe postcardiotomy myocardial dysfunction [6]. Most of these patients can be weaned with the aid of higher doses of inotropic agents and intraaortic balloon counterpulsation [6]. However, those who have persistent shock despite these measures face almost certain death. In this small group of patients, ECMO may provide sufficient hemodynamic support to allow recovery from reversible myocardial injury [7]. This use of ECMO, especially when applied to older patients, is open to criticism because of its cost, its unproved Accepted for publication Oct 17, 2000. Address reprint requests to Dr Bellomo, Department of Intensive Care, Austin & Repatriation Medical Centre, Studley Rd, Heidelberg, Victoria 3084, Australia; e-mail: [email protected].
© 2001 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
Material and Methods Study Group Between June 1996 and December 1998, ECMO was instituted in 17 consecutive patients with cardiogenic shock. The indications for establishing extracorporeal support in these patients were the following: 1. Failure to wean from cardiopulmonary bypass despite maximal inotropic and mechanical support. 2. Failure to maintain satisfactory systemic perfusion (systolic blood pressure ⬍ 80 mm Hg, cardiac inThis article has been selected for the open discussion forum on the STS Web site: http://www.sts.org/section/atsdiscussion/
0003-4975/01/$20.00 PII S0003-4975(00)02504-2
1422
SMITH ET AL ECMO IN OLDER PATIENTS
dex ⬍ 2.0 L 䡠 min⫺1 䡠 m⫺2, base excess ⬎ ⫺8 mEq/L, and anuria) despite the use of high-dose inotropic agents and intraaortic balloon counterpulsation, in the absence of a surgically correctable cause such as tamponade. In these patients, cardioplegia was administered in 800 mL of blood and contained 100 mL of tris(hydroxymethyl)aminomethane (Abbott, Australasia, NSW, Australia), 100 mg of lidocaine HCl, 20 mmol of KCl, and 10 mmol of MgCl2. Cardioplegia was delivered by the antegrade route and the coronary sinus cannula (approximately 500 mL at a temperature of 14°C to 18°C). Further cardioplegia solution was infused through the coronary sinus cannula at approximately 15-minute intervals. This consisted of 600 mL of blood with 5 mmol of KCl and 20 mmol of MgCl2 at a temperature of 14°C to 18°C. A “hot shot” was given through the aortic root immediately before aortic cross-clamp removal. This consisted of 600 mL of blood with 5 to 10 mmol of MgCl2 at 34°C to 36°C.
Extracorporeal Membrane Oxygenation The ECMO circuit consisted of a closed, Carmeda Bioactive Surface-coated circuit of polyvinyl chloride tubing (Medtronic, Minneapolis, MN) and a centrifugal pump (Biomedicus Biopump, Medtronic, Minneapolis, MN), which propelled blood through a hollow-fiber membrane oxygenator (Maxima P.R.F., Medtronic). Blood flows were monitored by a Doppler flow probe placed on the arterial side of the circuit. Cannulation was peripheral (femorofemoral) in 11 patients, using Biomedicus Carmeda Bioactive Surface-coated ECMO cannulas placed by direct cutdown into the femoral vessels; and central (aortoatrial) in 5 patients, using in situ cardiopulmonary bypass cannulas. No left atrial or ventricular drainage was used. In one further patient with central arterial cannulation, the venous cannula was moved to the femoral vein because of right atrial bleeding. Limb ischemia was a major problem early on in the study. Therefore, two methods of optimizing distal limb perfusion were instituted. One used a distal cannula passing from the side port of the arterial cannula to a point on the femoral artery distal to the cannulation site. The other was based on the cannulation of the femoral artery by means of a polyethylene terephthalate T graft.
Integrated Management Strategy Pump flows were chosen to supply adequate systemic circulatory support (2.0 to 2.5 L 䡠 min⫺1 䡠 m⫺2). Pulsatile pulmonary artery flow was maintained by means of right ventricular filling with intravenous colloids. The aim was to minimize the risk of intracardiac clot and reperfusion lung injury. A continuous cardiac output pulmonary artery catheter was used to assess pulmonary blood flow. Inotropic support was reduced or withdrawn to decrease myocardial oxygen demand and facilitate myocardial recovery. Norepinephrine was infused to maintain a mean systemic arterial pressure of more than 70 to
Ann Thorac Surg 2001;71:1421–7
75 mm Hg. To oxygenate any blood delivered to the lungs by the patient’s heart and to minimize atelectasis and lung injury, low tidal volume (5 to 6 mL/kg tidal breaths) mechanical ventilation with positive end-expiratory pressure at 8 to 10 cm H2O was applied. The activated clotting time was monitored, with the aim of ensuring an activated clotting time of approximately 140 to 150 seconds with full flows, and an activated clotting time of more than 200 seconds when flows were reduced less than 1.5 L/min during the weaning process. Early in our study, systemic heparinization was used in three patients; however, the remaining 14 patients were supported without any heparin except at low flows. Independent of the activated clotting time, no heparin was used if the patient’s chest drainage was more than 100 mL/h. If chest drainage was less than 100 mL/h for 6 to 8 hours, correction of the activated clotting time to the desired levels by means of heparin was considered. Intraaortic balloon counterpulsation was used in the majority of patients (14 of 17). Continuous venovenous hemofiltration was used in all 17 patients and was commenced early (within the first 8 hours) in the postoperative course. Continuous venovenous hemofiltration was used to regulate intravascular volume and overall fluid balance and to enable the rapid administration of blood and blood products without the induction of volume overload. To decrease right ventricular afterload, inhaled nitric oxide was used in 6 patients in whom there was a component of right ventricular dysfunction. Transesophageal echocardiography was used to guide the placement of the venous cannulas in patients with peripheral access. Serial transesophageal echocardiography was used during the period of extracorporeal support to enable the progressive assessment of myocardial recovery. It was also used to exclude intracardiac clot or clots within the cannulas, and to provide useful information regarding myocardial contractility and ventricular filling during the weaning process.
General Patient Care In all patients, enteral feeding was instituted in the first 24 hours and continued if tolerated. Muscle relaxants were only used for a short period (⬍ 8 hours) after admission to the cardiothoracic intensive care unit or to facilitate invasive procedures. Sedation was morphineand midazolam-based and used to achieve patient comfort while enabling daily neurologic assessment. Native urine output, if present, was maintained at more than 1 mL 䡠 kg⫺1 䡠 h⫺1 with a furosemide infusion. Packed red blood cells were transfused to maintain a hemoglobin concentration of approximately 100 g/L. Antibiotics were used if infection was clinically suspected. Hyperglycemia was controlled by continuous infusion of insulin. Atrial fibrillation with a fast ventricular response was controlled by the continuous intravenous infusion of amiodarone. The lower limbs were formally assessed for ischemia every 6 hours by physical examination, Doppler examination, and 6-hourly creatine kinase measurements.
Ann Thorac Surg 2001;71:1421–7
SMITH ET AL ECMO IN OLDER PATIENTS
Table 1. Clinical Pre-ECMO Features of Study Patients Age (y) 69 83 74 75 64 37 56 68 74 71 43 75 48 62 83 81 69
Sex
CPB Time (min)
X-Clamp Time (min)
IABP
Nature of Surgery
Preop NYHA
M M M M F M F M F M M M M M M F M
177 216 337 198 173 156 165 182 204 457 141 156 N/A 417 192 184 174
118 96 158 132 134 126 95 144 116 252 120 124 N/A 253 115 157 136
Y Y Y Y Y N Y Y Y Y N Y Y Y Y N Y
Elective Urgent Emergency Elective Elective Elective Urgent Elective Elective Elective Emergency Emergency N/A Elective Elective Urgent Urgent
III III IV IV III II III III III III I IV IV II III IV IV
CPB ⫽ cardiopulmonary bypass; ECMO ⫽ extracorporeal membrane oxygenation; IABP ⫽ intraaortic balloon pulsation; N/A ⫽ not applicable; preop NYHA ⫽ preoperative class according to the New York Heart Association; X-clamp ⫽ cross-clamp.
Data Collection Data collection was performed focussing on patient demographics, operative details, cannulation method, duration of support, blood product usage, use of hemofiltration, inotropic agents, counterpulsation, and nitric oxide. Data were obtained for morbidity and in-hospital outcome. All patients who survived to discharge were followed up after hospital discharge. All survivors completed a quality-of-life questionnaire. This questionnaire had been previously used at our institution to assess survivors of severe multiple organ failure [8], and was based on an activity index, mental function index, and a simplified version of the Nottingham Health Profile [9].
Results In the majority of patients, ECMO was required in the postcardiotomy setting either because of the inability to wean from cardiopulmonary bypass or refractory cardioTable 2. Operative Procedures Requiring Extracorporeal Membrane Oxygenation Procedure Coronary bypass Aortic valve/CABG Aortic root replacement Postinfarct VSD Pulmonary embolectomy CABG ⫽ coronary artery bypass grafting; defect.
No. of Patients 7 4 2 2 1 VSD ⫽ ventricular septal
1423
genic shock in the early postoperative period. This type of support was only required in 1.1% of all cardiac surgical operations during the 30-month period (mean age of overall patient population in question, 64.7 years). One patient also required ECMO for acute deterioration of dilated cardiomyopathy leading to cardiac arrest. Tables 1 and 2 summarize the clinical and operative features of the study patients. The median age of patients treated with ECMO was 69 years (range, 37 to 83 years). There were 8 patients older than 70 years of age, and 13 (76%) were male. The patients were supported for a median period of 86 hours (range, 20 to 201 hours). Only 4 patients were supported for less than 48 hours. Of these 4 patients, 2 were octogenarians who died early in the study because of severe limb ischemia and severe ongoing coagulopathic bleeding. The remaining 2 patients were weaned successfully after rapid recovery of right ventricular function. Eleven (65%) of the 17 ECMO-supported patients were successfully weaned. The median duration of support required for those patients who were successfully weaned was 106 hours. Overall, 7 patients (41%) survived to hospital discharge. Only 3 of 8 patients (38%) older than 70 years, however, were weaned from ECMO, and only 1 survived. Of the 9 patients younger than 70 years of age, 8 were successfully weaned (89%) and 6 (67%) survived. The median age of survivors was 62 years (range, 37 to 74 years) and of nonsurvivors, 74 years (range, 56 to 83 years). There were no significant differences in overall peak cardiac enzyme values between survivors and nonsurvivors. However, all patients with a peak creatine kinase of more than 3,000 IU or a peak postoperative cardiac troponin I of more than 500 ng/mL died. Mortality was caused by failure to wean because of intractable heart failure in 4 patients. Withdrawal of support occurred in 2 patients owing to severe limb ischemia and severe coagulopathic bleeding. Massive nonhemorrhagic stroke caused the death of 2 patients. Two patients who were successfully weaned died of overwhelming sepsis, one a week after weaning and the other 2 months after intensive care unit discharge. Survivors spent a median duration of 16 days (range, 4 to 34 days) postoperatively in the intensive care unit. The median postoperative hospital stay of survivors was 25 days (range, 10 to 40 days). Bleeding was a major problem in this group of patients, requiring the administration of platelets, fresh frozen plasma, and cryoprecipitate. Table 3 shows the median blood product requirement for the group. Six patients Table 3. Median Number of Blood Product Units Administered per Patient Product Platelets Fresh frozen plasma Cryoprecipitate Blood
Units 44 (14 –100) 30 (10 –70) 11 (0 –30) 26 (10 –56)
1424
SMITH ET AL ECMO IN OLDER PATIENTS
Ann Thorac Surg 2001;71:1421–7
Table 4. Postoperative Quality-of-Life Assessment (7 Patients) Employment Full-time Retired Unable to work Activities of daily living Shopping Washing self Cooking Gardening Use public transport Walk ⬎ 200 m Normal energy levels Normal sleep Personal relationships unaffected Satisfied with health Treatment worthwhile
2 4 1 6 6 6 6 6 6 6 6 6 7 7
(35%) required reoperation for excessive bleeding. In 3 patients, intraoperative aprotinin infusions were continued into the early postoperative period. Leg ischemia occurred in 4 patients (36%). One patient had limb ischemia despite the presence of a distal perfusion cannula. There was no limb ischemia in those with T-graft cannulation. All survivors were followed up at a median of 21 months after discharge with a quality-of-life questionnaire. Table 4 shows the results of the quality-of-life assessment. Six of 7 patients reported their ability to perform activities of daily living as satisfactory. Two patients had resumed full-time employment, and only 1 patient was unable to work because of illness, the remaining being retired. The majority reported normal energy levels and sleep patterns. All were satisfied with their general health, and considered their treatment to be worthwhile.
However, the mean age of the patients in these studies was more than a decade younger than in our cohort. In this investigation, we report a 41% overall survival in a group of patients with a mean age of 69 years. Importantly, this is also the first study to provide long-term follow-up of patients treated with ECMO for cardiogenic shock and to assess their quality of life. All hospital survivors remain alive at the time of follow-up, and the majority reports a satisfactory quality of life. Several surgical aspects must be considered when the decision is made to connect a patient to an ECMO circuit. For example, the configuration of arterial and venous cannulas is dependent on several factors: accessibility and state of femoral arteries and veins, condition of the aorta and right atrium, fragility of intrathoracic tissues, and extent of coagulopathy. Preexisting cannulas in the aorta and right atrium may be used to connect the patient to the ECMO circuit. This approach avoids cannulation in the groin with its risk of limb ischemia and infection but increases the chance of bleeding around central cannulas. In this setting, it may be difficult to close the sternum, because the underlying heart is quite edematous. When cardiopulmonary bypass cannulas or fresh aortic and right atrial cannulas are used, they should ideally be reinforced by extra pledgetted pursestring sutures and thoroughly secured. The cannulas and tubing may be brought out through the lower end of the wound or through separate stab incisions (Fig 1). Some allowance should also be made at the time of fixing the cannulas for recovery and resolution of myocardial edema to avoid
Comment Prolonged extracorporeal circulatory support has been previously described in the treatment of cardiogenic shock [3–7]. There are various mechanical assist devices available to provide circulatory support in this setting. These devices may be univentricular or biventricular; short-term or long-term, and ambulatory or nonambulatory. The choice of device depends on many clinical and logistic considerations. There are several potential advantages in using ECMO for short-term support, especially if cardiac transplantation is not available or appropriate. In particular, the combined biventricular and respiratory support provided by ECMO appears more desirable than that provided by ventricular assist devices in the presence of global myocardial dysfunction and markedly impaired gas exchange [10] as was the case in our patients. Recent studies suggest that in-hospital survival rates with the use of ECMO vary from 32% to 52% [3– 6].
Fig 1. Diagram showing the position and exit site of intrathoracic cannulas used for extracorporeal membrane oxygenation.
Ann Thorac Surg 2001;71:1421–7
SMITH ET AL ECMO IN OLDER PATIENTS
1425
Fig 4. Illustration of one of the techniques used to ensure maintenance of blood flow to the distal lower limb during femoral cannulation. The side cannula entering the femoral vessel distal to the cannulation site ensures flows greater than 150 mL/min. (ECMO ⫽ extracorporeal membrane oxygenation.)
Fig 2. Illustration of the preferred site for extrathoracic cannulation for extracorporeal membrane oxygenation.
traction on the cannulas. If femoral cannulation is chosen, the femoral vessels are isolated, and pursestring sutures are placed on each vessel with 5-0 Prolene
Fig 3. Diagram illustrating the appearance and position of the arterial cannula used for the return limb of the extracorporeal membrane oxygenation (ECMO) circuit.
(Ethicon, Somerville, NJ) sutures (Fig 2). It is preferable to cannulate the vein first. If the femoral artery is large, a small-bore cannula can be placed within it with little risk of distal limb ischemia (Fig 3). If the femoral artery is small or diseased, distal perfusion of the lower limb must be sought. Such perfusion may be achieved by inserting a small cannula (size 10F to 12F) distally and connecting it to the sidearm of the arterial inflow (Fig 4). Alternatively, a polyethylene terephthalate tube graft of 8-mm diameter can be anastomosed as a T graft to the femoral artery. The arterial cannula is then inserted into the graft and fastened securely (Fig 5). A distal draining cannula is placed in the femoral vein and connected to the venous
Fig 5. Illustration of the T-graft approach to maintaining both central arterial blood flow and distal lower limb perfusion. A polyethylene terephthalate graft is attached to the femoral artery, and the cannula is inserted into and tied to the graft. (ECMO ⫽ extracorporeal membrane oxygenation.)
1426
SMITH ET AL ECMO IN OLDER PATIENTS
Fig 6. Diagram illustrating the internal and external appearance of a double-lumen dialysis catheter used for continuous hemofiltration in all of our patients.
outflow cannula to reduce limb edema. Once the above precautions are taken, peripheral cannulation is our preferred means of access. The correct approach to sedation, pain relief, and muscle relaxation in these patients is unknown. We have used muscle relaxation with great caution because of the risks associated with it [11]. We have found that patient comfort and limited mobility can be achieved with a combination of morphine and midazolam given by continuous intravenous infusions at doses between 2 and 5 mg/h for each. Regular careful decrease or temporary cessation of these agents for daily neurologic assessment is desirable. Continuous venovenous hemofiltration through a double-lumen catheter (Fig 6) was used early in the postoperative course in all patients, because it allows the physician to control intravascular volume despite the need for massive and rapid transfusion of blood products. Such control prevents volume overload, cardiac edema, and pulmonary edema, and optimizes hemodynamics [12]. Excessive bleeding is common at the time the patient is placed on ECMO [6], is often life-threatening, and mandates the avoidance of anticoagulants and the rapid and massive administration of blood products. With this approach, we only experienced one death from bleeding and no episode of sudden clotting of the oxygenatiing membrane. Continuous venovenous hemofiltration can also proceed simultaneously without anticoagulation and remove plasma water at 500 to 700 mL/h to enable the safer transfusion of such products. Intracardiac clotting has been reported in 20% of patients during ECMO [6]. In our patients, despite avoiding systemic heparinization, intracardiac clotting was identified in only 1 patient. These findings suggest that with Carmeda Bioactive Surface-coated cannulas, clinicians can pursue a strategy that restricts the early administration of anticoagulants. Transesophageal echocardi-
Ann Thorac Surg 2001;71:1421–7
ography is essential before weaning to exclude the presence of intracardiac clot before native pump function is restored. Other issues of general patient care are also worth emphasizing. Early enteral nutrition appears desirable [13], as does the use of inhaled nitric oxide if right ventricular dysfunction is prominent [14]. Maintaining some lung perfusion at all times may prevent reperfusion pulmonary edema. Mechanical ventilation with low volume tidal breaths (6 mL/kg) and maintenance of alveolar patency by the use of positive end-expiratory pressure is recommended [15]. Cardiovascular support with ECMO may not ensure the maintenance of an adequate mean arterial blood pressure. Such hypotension may predispose the patient to cerebral and renal ischemia [16, 17]. Accordingly, a mean systemic blood pressure of more than 70 to 75 mm Hg was maintained with the continuous intravenous infusion of norepinephrine. All oliguric patients received continuous infusion of furosemide, titrated to maintain a urine output of more than 1 mL 䡠 kg⫺1 䡠 h⫺1 [18]. To prevent arrhythmias, we frequently administered intravenous magnesium [19] and amiodarone [20]. If such prophylaxis failed, the rate of ventricular response to atrial fibrillation was also controlled with the continuous infusion of intravenous amiodarone. Hyperglycemia was treated with insulin by continuous infusion titrated to maintain a serum glucose concentration between 7 and 10 mmol/L. Weaning patients from ECMO is often slow and difficult. We have used transesophageal echocardiography in all of our patients to study the effects of decreasing pump flows on myocardial chamber size and contractility. In all of our patients, we also used milrinone to increase myocardial contractility at the time of weaning [21]. We chose milrinone in preference to other inotropic agents because of it does not induce tachycardia, it has a minimal effect on myocardial oxygen consumption, it induces diastolic relaxation, and it increases coronary graft blood flow [22, 23]. We found milrinone safe and effective and support its use in preference to -adrenergic drugs. Milrinone also induces vasodilation and accumulates in the presence of renal dysfunction. The use of an ␣-adrenergic drug (norepinephrine) to maintain mean systemic blood pressure and dose adjustments may become necessary under such circumstances. In conclusion, this study reports the long-term outcome of the use of an integrated extracorporeal life support approach to the treatment of cardiogenic shock after cardiac operations in older patients. Previously, the mortality rate for this patient group at our institution was more than 90%. However, the use of ECMO alongside the integrated management strategy described above resulted in a 41% survival rate at approximately 2 years postoperatively, as well as a satisfactory quality of life. Accordingly, we believe that the use of an ECMO-based strategy is justified if intractable but potentially reversible cardiogenic shock is present or patients fail to wean from cardiopulmonary bypass.
Ann Thorac Surg 2001;71:1421–7
SMITH ET AL ECMO IN OLDER PATIENTS
References 1. Zapol WM, Snider MT, Hill JD, et al. Extracorporeal membrane oxygenation in severe acute respiratory failure. JAMA 1979;242:2193– 6. 2. Gattinoni L, Pesenti A, Mascheroni D, et al. Low-frequency positive pressure ventilation with extracorporeal CO2 removal in severe acute respiratory failure. JAMA 1986;256: 881– 6. 3. Reedy JE, Swartz MT, Raithel SC, et al. Mechanical cardiopulmonary support for refractory cardiogenic shock. Heart Lung 1990;19:514–23. 4. Magovern GJ, Magovern JA, Benckart DH, et al. Extracorporeal membrane oxygenation: preliminary results in patients with postcardiotomy cardiogenic shock. Ann Thorac Surg 1992;57:1462–71. 5. Muehrcke DD, McCarthy PM, Stewart RW, et al. Extracorporeal membrane oxygenation for postcardiotomy cardiogenic shock. Ann Thorac Surg 1996;61:684–91. 6. Golding LAR. Postcardiotomy mechanical support. Semin Thorac Cardiovasc Surg 1991;3:29–33. 7. Muehrcke DD, McCarthy PM, Stewart RW, et al. Complications of extracorporeal life support systems using heparinbound surfaces. The risk of intracardiac clot formation. J Thorac Cardiovasc Surg 1995;110:843–57. 8. Gopal I, Bhonagiri S, Ronco C, Bellomo R. Out of hospital outcome and quality of life in survivors of combined acute multiple organ and renal failure treated with continuous venovenous hemofiltration/hemodiafiltration. Intensive Care Med 1997;23:766–72. 9. Hunt SM, McEwan J, McKenna SP. The Nottingham Health Profile: subjective health status and medical consultations. Soc Sci Med 1981;15A:221–9. 10. Jurmann MJ, Haverich A, Demertzis S, et al. Extracorporeal membrane oxygenation (ECMO); extended indications for artificial support for artificial support for both heart and lungs. Int J Artif Org 1991;14:771– 4. 11. Segredo V, Caldwell JE, Matthay MA, Sharma ML, Gruenke
12. 13. 14. 15. 16.
17. 18. 19. 20. 21.
22.
1427
LD, Miller RD. Persistent paralysis in critically ill patients after long-term administration of vecuronium. N Engl J Med 1992;327:524– 8. Coraim FI, Pauser G, Stellwag F, et al. Positive influence on hemodynamics in postcardiac surgery patients by hemofiltration. Anesthetist 1985;34:236– 40. Zaloga G. Early enteral nutritional support improves outcome: hypothesis or fact. Crit Care Med 1999;27:259– 61. Rich GF, Murphy GD, Roos CM, Johns RA. Inhaled nitric oxide. Selective pulmonary vasodilation in cardiac surgical patients. Anesthesiology 1993;78:1028–35. Artigas A, Bernard GR, Carlet J, et al. The AmericanEuropean consensus conference on ARDS, Part 2. Intensive Care Med 1998;24:378–98. Gold JP, Charlson ME, Williams-Russo, et al. Improvement of outcomes after coronary artery bypass. A randomized trial comparing intraoperative high versus low mean arterial pressure. J Thorac Cardiovasc Surg 1996;110:1302–11. Bersten AD, Holt AW. Vasoactive drugs and the importance of renal perfusion pressure. New Horiz 1995;3:650– 61. Martin SJ, Danziger LH. Continuous infusion of loop diuretics in the critically ill: a review of the literature. Crit Care Med 1994;22:1323–9. England MR, Gordon G, Salem M, Chernow B. Magnesium administration and dysrhythmias after cardiac surgery. JAMA 1992;268:2395– 402. Daoud EG, Strickberger AS, Chng Man K, et al. Preoperative amiodarone as prophylaxis against atrial fibrillation after cardiac surgery. N Engl J Med 1997;337:1785–91. Doolan LA, Jones E, Kalman J, Buxton BF, Tonkin AM. A placebo-controlled trial verifying the efficacy of milrinone in weaning high-risk patients from cardiopulmonary bypass. J Cardiothorac Vasc Anesth 1997;11:37– 41. Feneck RO, and the European Milrinone Multicentre Trial Group. Intravenous milrinone following cardiac surgery: effect of bolus infusion followed by variable dose maintenance infusion. J Cardiothorac Vasc Anes 1992;6:554– 62.