| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
*Division of Cardiovascular Anesthesia and Intensive Care, Policlinico di Monza, Monza, Italy;
Division of Cardiac Surgery, San Raffaele Hospital, Milan, Italy;
Division of Cardiac Surgery, Policlinico di Monza, Monza, Italy; and
Coagulation Service and Thrombosis Research Unit, San Raffaele Hospital, Milan, Italy
Address correspondence and reprint requests to Valter Casati, MD, Division of Cardiovascular Anesthesia and Intensive Care, Policlinico di Monza, via Amati 111, Monza (20052), Italy. Address e-mail to valter.casati{at}policlinicodimonza.it
Abstract
We evaluated the blood-sparing effects of intraoperative moderate acute normovolemic hemodilution (ANH) combined with intraoperative tranexamic acid treatment and shed blood reinfusion in patients undergoing off-pump coronary artery bypass (OPCAB). One-hundred consecutive OPCAB patients (baseline hematocrit >34%) were prospectively randomized to tranexamic acid treatment (control group; 50 patients) or to tranexamic acid treatment plus normovolemic (1:1 replacement with colloids) withdrawal of 17% ± 2% of the circulating blood volume (ANH group; 50 patients). All patients had shed blood reinfused with intraoperative bleeding in excess of 250 mL. The requirement for allogeneic transfusions, based on strict a priori defined criteria, was the primary end point of the study. Hematochemical evaluations, bleeding, major complications, and other outcomes were also recorded. Demographics, baseline hematochemical data, and operative characteristics were similar in the two groups. Patients in the ANH group had a median of 850 mL of blood withdrawn and showed a lower intraoperative minimum hematocrit (31% vs 37%; P < 0.0001). Two patients in the ANH group versus 10 patients in the control group (odds ratio, 0.17; 95% confidence interval, 0.03–0.89; P = 0.028) required transfusion of a significantly smaller number of packed red blood cell units (5 vs 24; P < 0.001). Postoperative hematochemical variables, bleeding, and outcomes were similar in the two groups of patients. Moderate ANH, combined with tranexamic acid administration and on-demand shed blood reinfusion, may reduce allogeneic transfusion requirements in OPCAB patients.
IMPLICATIONS: We studied the blood-sparing effects of moderate acute normovolemic hemodilution (ANH) in 100 patients undergoing off-pump coronary surgery (OPCAB). Combined with tranexamic acid administration and shed blood reinfusion when the intraoperative bleeding exceeded 250 mL, ANH was effective in reducing the number of OPCAB patients who required allogeneic transfusions and the number of packed red blood cell units transfused.
Compared with coronary surgery performed with cardiopulmonary bypass (CPB), off-pump coronary artery bypass (OPCAB) surgery is associated with a reduced frequency of inflammatory and hemorrhagic disorders (1–3). Perioperative administration of tranexamic acid, a synthetic antifibrinolytic drug with hemostatic properties (4), may significantly reduce bleeding in OPCAB patients (5). In case of excessive intraoperative bleeding, the use of a cell salvage system, which collects and processes blood shed from the surgical field, may further reduce the need for transfusions (6). We now routinely use tranexamic acid and on-demand reinfusion of shed blood in OPCAB surgery. Still, the number of OPCAB patients who require allogeneic transfusions is significant (7,8).
Intraoperative acute normovolemic hemodilution (ANH) involves preoperative withdrawal of a certain amount of blood while isovolemia is maintained with the infusion of colloid and crystalloid solutions. The autologous blood is then reinfused, either intraoperatively, in case of significant bleeding, or after surgery, after obtaining adequate hemostasis (9).
ANH has been evaluated in CPB surgery with contradictory results (10–12). The aim of this study was to evaluate in nonanemic OPCAB patients the blood-sparing effects of moderate ANH [withdrawal of 17% ± 2% of the circulating blood volume (13) while maintaining the hematocrit at more than 30%] combined with intraoperative tranexamic acid administration and on-demand intraoperative reinfusion of shed blood.
Methods
Approximately 25% of OPCAB patients require allogeneic red blood cell transfusions (5,7,8). This trial was designed to investigate the efficacy of ANH in reducing this percentage nearer to 5%. To achieve significance with an 
error of 0.05 and 80% power, a minimum of 44 patients per group was required (with continuity correction). The study was approved by the IRB.
A total of 112 consecutive patients scheduled for OPCAB surgery were considered for enrollment in the study. Exclusion criteria were left main coronary artery stenosis; left ventricular ejection fraction less than 30%; anemia (hematocrit <34% and hemoglobin <11.5 g/dL); history of hematological disorders; advanced chronic renal failure (serum creatinine >2 mg/dL); active chronic hepatitis; or cirrhosis. Preoperative treatments with aspirin or heparin were not exclusion criteria. Twelve patients (six with preoperative anemia, three with left ventricular ejection fraction <30%, two with advanced chronic renal failure, and one with a history of hemorrhagic diathesis) met exclusion criteria. After informed, written consent was obtained, the remaining 100 patients were considered for randomization.
Balanced anesthesia with propofol, fentanyl, isoflurane, and pancuronium bromide was performed in all patients. All patients received intraoperative tranexamic acid as an IV bolus of 1 g 20 min before sternotomy, followed by a continuous infusion of 400 mg/h until the end of surgery (5). The blood shed from the surgical field was collected in a cardiotomy reservoir (Autotransfusion Reservoir; Dideco, Mirandola, Italy) and, in case of intraoperative bleeding more than 250 mL, reinfused after washing and concentration in a cell salvage circuit (Compact Advanced; Dideco). During surgery, isovolemia was maintained by the infusion of lactated Ringer’s solution and succinylated gelatin (1:1 ratio) and monitored by standard variables (heart rate, systemic arterial blood pressure, and central venous pressure). ST segment analysis of leads II and V5 of the electrocardiogram (ECG) was continuously monitored for potential signs of ischemia.
By using a computer-generated random-number sequence, the 100 patients were prospectively randomized to ANH or no treatment. In the ANH group (50 patients), the whole-blood volume targeted for removal was 17% ± 2% of the circulating volume, as calculated from a body-surface area nomogram (13) by the attending anesthesiologist. After the induction of anesthesia and before systemic heparinization, blood was drawn by gravity through a large-bore catheter (8.5F) placed into the internal jugular vein and was collected into sterile bags containing citrate phosphate dextrose (Fenwal; Baxter Healthcare Corp., Irvine, CA) by using a blood mixer and balance system (Easymix V3; Baxter). During blood withdrawal, 4% succinylated gelatin in 0.9% NaCl (Eufusin; Fresenius Kabi Potenza S.r.l., Tito Scalo, Italy) was infused at a 1:1 ratio. Each unit of blood was labeled with the patient’s data, kept in the operating room (OR), and agitated at room temperature by using the same blood mixer and balance system until reinfusion. Irrespective of hematocrit values, reinfusion of the harvested autologous blood was started after protamine administration and on-demand reinfusion of the shed blood. No unit of autologous blood was wasted.
All patients were operated on in our institution with standardized techniques. The skeletonized left internal mammary artery (LIMA) was isolated through an extrapleural approach in all but three patients per group undergoing coronary surgery reintervention who had had their LIMA harvested during the first operation. The saphenous vein alone was used in 2 patients per group; the radial artery was used in one patient in the control group, and the gastroepiploic artery was used in one patient in the ANH group. In five patients of both groups, the LIMA and the right internal mammary artery were isolated to obtain a complete arterial coronary revascularization. When indicated to complete coronary revascularization, a tract of the saphenous vein was isolated through a standard open approach in the remaining patients.
Heparin (1 mg/kg) was administered to obtain an activated clotting time >250 s, and, if needed, further 50-mg boluses were injected. No drugs were administered to control the heart rate. After pericardiotomy, the target site of the coronary artery to be grafted was stabilized by a mechanical device (CAB Super Slide Retractor Set; Baxter Healthcare Corp., Cardiovascular Group). After coronary revascularization, heparin was antagonized with protamine (1 mg/1 mg); if needed, additional protamine was injected to obtain an activated clotting time near baseline (with 10% tolerance).
Before chest closure, mediastinal drains were positioned, and low-grade suction was instituted. Starting on the first postoperative day, all patients received a combination of small-dose aspirin (100 mg/d) and nadroparin calcium (2850 anti-Xa U/d).
Hemoglobin, hematocrit, platelet count, prothrombin time, activated partial thromboplastin time, creatinine, creatine phosphokinase, and creatine phosphokinase myocardial band isoenzyme were measured in blood samples obtained before the induction of anesthesia; at arrival in the intensive care unit (ICU) after completion of the reinfusion of autologous blood; 24 and 48 h after surgery; and before discharge from the cardiosurgical unit. Additional monitoring was performed if required by the clinical situation. During surgery, hematocrit and hemoglobin were monitored in serial samples drawn for blood gas determinations (IL 1640; Instrumentation Laboratory SPA, Milan, Italy) after the induction of anesthesia, before heparin administration, every 20 min during coronary surgery, after protamine administration, and at the completion of surgery.
Criteria for allogeneic transfusions were established a priori. Packed red blood cells (PRBC) were transfused with hemoglobin <8 g/dL and hematocrit <24% after on-demand reinfusion of shed blood and, in the ANH group, after reinfusion of the harvested autologous blood. Fresh frozen plasma (FFP) was infused after protamine administration if the prothrombin time was 
1.5 times the baseline with microvascular bleeding. The criterion for transfusion of platelet concentrates was a platelet count <50,000/mm3 in the presence of microvascular bleeding. The blood loss was recorded during the first 24 h, and chest drains were removed with bleeding <100 mL over 4 h. Surgical reexploration was performed with a normal coagulation profile and bleeding more than 300 mL/h over the first 2 h after arrival in the ICU or more than 200 mL/h over four consecutive hours.
Outcomes and the following major complications were recorded: respiratory failure (need for mechanical ventilation for more than 48 h), myocardial infarction (new Q waves on ECG, creatine phosphokinase myocardial band isoenzyme and creatine phosphokinase ratio more than 10%, and troponin I >0.1 µg/L), acute renal failure (creatinine twice the baseline value or need for dialysis), venous thromboembolism, and minor (awakening, with agitation or delayed) and major (transient ischemic attack or stroke) neurological complications.
Statistical analysis was performed with SPSS 11.5 for Windows (SPSS Inc., Chicago, IL). The Kolmogorov-Smirnov test was used to assess the normality of the distribution of continuous variables. Normally distributed variables were compared between groups with the Student’s t-test (two tailed) and expressed as mean and SD. Nonnormally distributed variables were compared with the Mann-Whitney U-test and expressed as median and interquartile range (25th–75th percentile). Efficacy was analyzed by following the intention-to-treat rule. Proportions were compared with the 
2 test or Fisher’s exact test, as appropriate. The binomial distribution was used to calculate odds ratios and 95% confidence intervals.
Two-way analysis of variance for repeated measures was used to evaluate the main effect of treatment, time, and the interaction of time with treatment on hematochemical variables. Differences were considered statistically significant with P values <0.05.
Results
Demographic and baseline hematochemical data were similar in the 2 groups of patients (Tables 1 and 2). One patient in each group required conversion to CPB surgery because of hemodynamic instability during manipulation of the heart. Because of the intention-to-treat approach of the study, they were included in the statistical analysis. Operative data are shown in Table 3. A median of 850 mL of blood was withdrawn from patients in the ANH group, who received larger intraoperative amounts of colloid and crystalloid solutions when including those given to compensate for blood withdrawal. No patient in the ANH group experienced complications related to normovolemic blood withdrawal.
|
|
|
|
|
Outcomes and the incidence of major complications were similar in the two groups (Table 5). One patient in the control group with a preoperative history of severe chronic obstructive pulmonary disease died on postoperative Day 9 because of respiratory failure and septic shock due to pneumonia. With the one exception of creatinine levels, which were unchanged, the hematochemical variables considered changed significantly with time, but no significant differences emerged between groups or as a group-time interaction (data not shown).
|
With the introduction of OPCAB surgery, a decreased use of allogeneic blood products has been reported, but excess bleeding requiring transfusions remains a major concern (7,8,14). Although activation of fibrinolysis and platelet consumption are less in OPCAB than in CPB patients, net consumption of antithrombin and fibrinogen and blood loss occur to a similar degree with the two types of surgery (15). In a pilot study, the intraoperative infusion of tranexamic acid, a synthetic antifibrinolytic drug with proven hemostatic effect in patients submitted to cardiac surgery with CPB and widely studied at our institution (16–19), was also shown to reduce bleeding in OPCAB surgery (5); however, the limited number of patients included in that study did not allow evaluation of its effect on transfusion requirements. Intraoperative ANH has been investigated in on-pump cardiac surgery, with controversial results (10–12). This is the first study to evaluate the blood-sparing effects of moderate ANH in patients submitted to OPCAB surgery. An average of 850 mL of blood was withdrawn to reach a target hematocrit close to, but more than, 30%. The design of the study could not include a double-blind procedure, either during or after surgery: only occasionally was the reinfusion of the autologous blood completed within the OR and proceeding to the ICU. Therefore, we established precise a priori objective criteria for transfusion requirements.
The addition of moderate ANH to our standard blood-sparing protocol resulted in an apparent fivefold reduction in the number of patients transfused with donor blood. Accordingly, we observed a reduction in the number of PRBC units transfused. The only two patients who required transfusions in the ANH group experienced excessive bleeding that necessitated surgical reexploration; this revealed evidence of a clear surgical source. The same 2 patients required additional PRBC transfusions for the later onset of anemia. No other patient in the ANH group met the criteria for allogeneic transfusions, whereas 10 patients in the control group required transfusions. Also in this group, two patients required reexploration for surgical bleeding and were transfused with PRBC and FFP. The remaining eight patients received PRBC transfusions because of postoperative anemia.
Not surprisingly, the total blood loss was similar in the 2 groups. Most of the blood loss occurs during surgery or in the first postoperative hours, and the blood-sparing properties of ANH are mainly related to increased dilution of the intraoperative blood loss, leading to a smaller net loss of red blood cells (9). In addition, reinfusion of autologous blood avoids the excess postoperative hemodilution induced by the use of crystalloids and colloids. Our results are very similar to those of Nuttall et al. (11) with the association of tranexamic acid and ANH in on-pump patients.
Because of the relatively small degree of hemodilution and the attention paid to maintain normovolemia, no patient in this series experienced intraoperative myocardial ischemia, and the outcomes and postoperative complications did not differ between the two treatment groups. As suggested by some authors, a more profound hemodilution might lead to serious complications, such as metabolic acidosis, peripheral edema, pulmonary and neurological sequelae, and, especially in coronary patients, myocardial ischemia (20–24).
There is concern about the occurrence of procoagulant activity (potentially leading to an increased rate of graft thrombosis) after OPCAB surgery (25,26). As a remedy, it has been suggested to avoid heparin neutralization with protamine and to initiate adequate anticoagulation in the early postoperative period (25,27). This is not the advice of other authors, who reported similar rates of graft patency in patients operated on with or without CPB, with better results in terms of early outcome in OPCAB patients (28). Admittedly, the use of procoagulant maneuvers such as ANH (29) in association with tranexamic acid may further increase the potential for graft thrombosis in OPCAB patients. In our series, postoperative acute myocardial infarction was diagnosed, on the basis of ECG abnormalities and a significant increase in myocardial enzymes, in only one patient in each group. Early initiation of aspirin and low-molecular-weight heparin administration may have played a role in preventing the occurrence of signs of myocardial ischemia, but the safety of the combination of moderate ANH and tranexamic acid in OPCAB surgery requires confirmation in larger studies.
In conclusion, moderate ANH, combined with tranexamic acid administration and reinfusion of shed blood with intraoperative bleeding exceeding 250 mL, led to a significant reduction in the number of OPCAB patients who required donor transfusions. Larger studies are needed to confirm these results and eliminate the possibility of an increased risk of thrombotic complications.
References
-aminocaproic acid in primary cardiac surgery. Ann Thorac Surg 1999; 68: 2252–7.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
) and ANH (
) patients submitted to off-pump coronary artery bypass surgery. Results are presented with box plots, medians, and whiskers denoting the range of values occurring within the inner fences. Determinations were performed on venous blood collected after the induction of anesthesia (Baseline), at arrival in the intensive care unit (ICU), 24 and 48 h after surgery, and before patient discharge. The minimum intraoperative hematocrit values (Intra-op) were determined on arterial blood.
