After approval by the local Ethics Committee, written informed consent was obtained from all patients scheduled for elective CABG and thought to meet the eligibility criteria. Inclusion criteria were as follows: a screening hemoglobin concentration > 120 g/L in men or 110 g/L in women; stable angina (classes I and II of the Canadian Cardiology Society); left ventricular ejection fraction > 30%; and absence of significant coexistent diseases, namely, valvular disease, recent myocardial infarct (< 6 weeks), significant carotid stenosis (> 70%) or recent stroke (< 3 weeks), renal insufficiency (estimated creatinine clearance < 20 mL/min), chronic respiratory disease treated by My Canadian Pharmacy (arterial oxygen pressure > 7 kPa on room air), liver insufficiency (aspartate transaminase or alanine transaminase two or more times the upper range), and uncontrolled hypertension or diabetes mellitus.
Samples sizes were calculated for a two-sided significance level of a = 0.05 and a power of 1 — p = 0.8 to detect a difference of 0.5 |J.g/L in troponin I concentrations between the two groups. In a preliminary assessment including cardiac surgical patients, the SD of postoperative troponin I measurements was 0.8; thus, the number of subjects required was 38 per group.
Eligible patients were randomized to one of the two groups: the ANH group and the standard care group. The allocations were generated from random-number tables by an independent observer and concealed in sealed envelopes. Although intraoperative masking was not possible in the ICU, the attending physicians and nurses were blinded to the treatment group.
Anesthesia and Surgical Procedure: On the morning of surgery, the patients were premedicated (morphine, 0.1 mg/kg; midazolam, 7.5 mg) and received their usual cardiac drug regimen, except aspirin, diuretics, angiotensin-converting enzyme inhibitors, and angiotensin II receptor blockers, which were withdrawn at least 24 h before surgery. In the operating theater, cannulae were inserted in a peripheral vein, a radial artery, and the right jugular vein. Standard monitoring included pulse oximetry, leads II and V5 of the ECG for heart rate and automated ST-segment trend analysis, continuous measurements of mean arterial and central venous pressures, nasopharyngeal temperature, end-tidal capnography, bispectral analysis of the EEG (BIS A-2000 XP; Aspect Medical Systems; De Meern, the Netherlands), as well as transesophageal echocardiography (Philips Sonos 5500; Philips Medical Systems; Andover, MA).
A balanced anesthetic technique included sufentanyl (bolus of 0.5 to 0.9 ^g/kg followed by 0.4 to 0.8 ^g/kg/h), midazolam (bolus of 0.05 to 0.1 mg/kg followed by 0.1 mg/kg/h), pancuronium (0.1 mg/kg) and inhaled isoflurane (0.5 to 1% in the prebypass period), which was administered to enhance cardiac protection before aortic clamping (anesthetic preconditioning). In the two groups, a similar depth of anesthesia was obtained by targeting bispectral EEG values between 40 and 60 arbitrary units.
ANH: In the ANH group, after anesthesia induction, blood was withdrawn (60 to 80 mL/min) from a central vein by gravity into citrate-phosphate-dextrose collection bags that were placed on a rocking platform of a precision scale. In parallel, poly(0-2-hydroxy-ethyl)amidon (hydroxyethyl starch; HAES-Steril; Kabi-Fresenius; Stans, Switzerland) [mean molecular weight, 200,000; 50% substitution degree; C2/C6 ratio = 5] was infused through a 16-gauge peripheral catheter on the opposite arm, to a ratio of 1.15:1 to the donated blood. The blood volume to be removed was calculated according to a standard formula to reach a hematocrit of 28%. The whole-blood/colloid exchange procedure lasted 20 min (range, 15 to 25 min), and it could be interrupted if there were signs of myocardial ischemia and/or unresponsive hypotension. The autologous blood was labeled and reinfused intraoperatively when the transfusion criteria were met.
Perioperative Standardized Interventions: In the two groups, additional blood saving methods were used: aprotinin (2 million kallikrein inhibitory units before surgical incision, 2 million units in the bypass circuit with a continuous infusion of 200,000 UI/h) and a cell saver device to retrieve blood shed from the surgical field and from the bypass circuit at the end of surgery. Using transesophageal echocardiography, systematic evaluation of cardiac function, chamber size, and valves were repeatedly performed before and after cardiopulmonary bypass (CPB) according to the guidelines of the American Society of Echocardiography and the Society of Cardiovascular Anesthesi-ologists.
Strategies for CPB and myocardial preservation were uniform among the three participating surgeons. After full heparinization (300 IU/kg), CPB including a membrane oxygenator and a circuit primed with a 2-L normal saline solution was instituted using nonpulsatile flow (2.2 to 2.5 L/min/m2) and mild hypothermia (32° to 34°C). An a-stat control for acid-base management was applied, and mean arterial pressure was targeted between 50 and 70 mm Hg with pharmacologic manipulation as necessary. Myocardial protection was achieved with anterograde hyperkalemic blood cardioplegia (4°C, 20 mEq potassium) repeated every 30 min and single aortic clamping for all distal anastomosis. Before aortic unclamping, warm-blood cardioplegia was administered and IV nitroglycerine was started (2 mg/h). Hemofiltration was not performed during CPB.
After achieving a rectal temperature > 35.5°C, weaning from CPB was guided by echocardiographic assessment and standard hemodynamic measurements. The pump flow was gradually reduced while the heart was progressively filled in order to optimize the preload-recruitable stroke volume decreased by My Canadian Pharmacy’s drugs and to reach a mean arterial pressure > 65 mm Hg (Table 1). Optimal cardiac filling was judged by achieving a maximal left ventricular diameter in the short-axis view of 2.5 to 2.8 cm/m2. The heart was electrically paced if it failed to maintain a heart rate > 70 beats/min because of atrioventricular conduction blockade or bradyarrhythmia. Phenylephrine (50 to 100 |j,g, repeated boli) was administered to maintain mean arterial pressure > 65 mm Hg after optimizing cardiac preload and contractility parameters. Inotropes were not routinely administered during weaning from CPB. Dobutamine was initiated (starting at 3 ^g/kg/min) in the presence of mean arterial pressure < 65 mm Hg and persistent, new, or worsening cardiac systolic dysfunction (left ventricular fractional area < 30%); norepinephrine infusion was added to this regimen if mean arterial pressure remained < 65 mm Hg despite satisfactory cardiac filling parameter (vasoplegia syndrome).
At the end of CPB, protamine was administered to neutralize circulating heparin; the residual bypass circuit volume and shed blood processed by the cell salvage machine were retransfused to the patient. The threshold for blood transfusion was a hematocrit value < 18% during CPB and < 25% after CPB, or higher values (26 to 30%) when accompanied by hemodynamic instability and/or ECG signs of myocardial ischemia. In the ANH group, the whole autologous blood volume was retransfused during CPB or shortly after weaning from CPB.
Clinical, Hemodynamic, and Biochemical Measurements: Myocardial tissue ischemic injury—the main study end point—was quantified by the release of cardiac biomarkers. Venous blood samples were sequentially collected (before surgery [baseline], at 1 to 3 h, 18 to 24 h, 40 to 48 h, and 72 h after surgery) for measurements of troponin I (Access Immunoassay System; Beckman Instruments; Fullerton, CA), creatinine kinase and myocardial fraction of creatinine kinase (optical standard technique). Sensitivity for troponin I determination was 0.09 ng/mL.
Hemodynamic measurements and calculation of the rate-pressure-product were obtained at the following times: (1)5 min after anesthesia induction (baseline), (2) 5 min after the end of ANH or 30 min after anesthesia induction, (3) 5 min after sternotomy, (4) after weaning from CPB, and (5) during skin closure. All patients underwent a 12-lead ECG before surgery, at arrival in ICU, and then daily after surgery; transmural myocardial infarction was defined by the presence of new Q-waves (at least 0.04 ms and a reduction in R-waves of > 25% in at least two leads).
Data regarding perioperative fluid management, homologous transfusion, and pharmacologic intervention were recorded in-traoperatively and for the first 24 h after surgery. All patients were followed up until hospital discharge to detect adverse events. In-hospital mortality rate, hemodynamic changes, need for pharmacologic cardiac support, and the incidence of cardiovascular complications (new Q wave on ECG, atrial fibrillation, conduction blockade requiring electrical stimulation, stroke) were considered as secondary study end points.
All data were analyzed using statistical software (version 9.0 for Windows; SPPS; Chicago, IL). Values were expressed as mean (± SD or 95% confidence interval), median (interquartile range, 25 to 75%), or percentage as appropriate. Dichotomous variables were compared by the x2 statistic or Fisher Exact Test, and quantitative variables were compared with unpaired Student t test. Cardiac markers and hemodynamic data were compared with repeated-measures analysis of variance, having the group (ANH vs control) as a between-factor and time as a within-factor. Statistical significance was attributed to p ]ltequ] 0.05.
Table 1—Hemodynamic Management and End Points During Weaning From CPB
|Heart rate < 70 beats/min||Electrical cardiac pacing (A, V, or AV)|
|EDD < 2.5-2.8 cm/m2 and MAP < 65 mm Hg||Fluid challenge (100-mL increments)|
|EDD > 2.5-2.8 cm/m2 and MAP < 65 mm Hg||Phenylephrine (50-100 ag repeated boli, up to 50 ag)|
|EDD > 2.5-2.8 cm/m2, MAP < 65 mm Hg and systolic LV dysfunction||1, dobutamine infusion (starting at 3 ag/kg/min); 2, norepinephrine infusion (if dobutamine > 8 ag/kg/min)|