Secondary Logo

Journal Logo

Is Perioperative Plasma Aprotinin Concentration More Predictable and Constant After a Weight-Related Dose Regimen?

Royston, David FRCA*,; Cardigan, Rebecca PhD†,; Gippner-Steppert, Cornelia PhD‡,; Jochum, Marianne PhD

doi: 10.1097/00000539-200104000-00007
Cardiovascular Anesthesia: Research Report
Free
SDC

To determine whether a weight-related dose had advantages over a fixed, large-dose regimen, we measured plasma concentrations of aprotinin by using an enzyme-linked immunosorbent assay method at set time points in 30 patients having heart surgery with cardiopulmonary bypass. A weight-related dose comprising a preincision bolus injection of 40,000 kallikrein-inhibiting units (KIU)/kg (5.6 mg/kg) with the same amount given in the oxygenator prime was compared with a large-dose regimen of 2 × 106 KIU (280 mg) preincision bolus and addition to prime, together with an infusion of 500,000 KIU/h (70 mg/h). Peak plasma concentration in the Weight-Related group was less variable than with the fixed-dose regimen. Forty percent of patients allocated to the fixed-dose regimen had an aprotinin concentration of more than 400 KIU/mL, compared with none in the Weight-Related group; this suggests a relative overdosing in the early surgical period in the Fixed-Dose group. There was great individual variability between patients in the time-concentration curves for aprotinin, with no difference between the two regimens. The weight-related dose regimen benefited by not requiring an intraoperative infusion while achieving the same plasma concentrations of aprotinin.

*Department of Anaesthesia, Harefield Hospital, Harefield, Middlesex, United Kingdom; †Department of Haematology, University College London Medical School, London, England, United Kingdom; and ‡Division of Clinical Biochemistry in the Department of Surgery, Klinikum Innenstadt der LM-Universität München, Munich, Germany

December 14, 2000.

Implications: Peak plasma concentrations of aprotinin were less variable with a weight-related dose schedule. This has implications for safety with regard to control of anticoagulation and cost in patients with small body mass. Plasma concentrations varied greatly with time between patients. This observation has implications for determining an optimal dose on the basis of aprotinin’s currently known mechanisms of action.

Address correspondence and reprint requests to Dr. Dave Royston, Department of Anaesthesia, Harefield Hospital, Harefield, Middlesex, UB9 6JH, UK. Address e-mail to Dave@tharg.demon.co.uk.

The use of aprotinin therapy is associated with a significant reduction in bleeding and with fewer requirements for blood and blood products. However, the currently recommended dosing regimen gives a fixed dose to all patients regardless of body mass, age, sex, or pathology. Concentrations of aprotinin, and especially peak concentrations, measured in the plasma after the administration of this regimen are highly variable (1–5). This may have ramifications for studies of the mechanism of the efficacy of this therapy and for safety, especially related to monitoring of anticoagulation. There are also cost implications of using a fixed, large dose in patients who have a relatively small body mass.

In this study, we investigated a weight-defined dose regimen. We sought to determine whether this approach would lead to more predictable plasma aprotinin concentrations than the fixed-dose regimen during the operative procedure. A secondary aim was to determine whether the use of a weight-related regimen would allow a simple and accurate prediction of the final plasma concentration of aprotinin.

Back to Top | Article Outline

Methods

With IRB approval, blood samples were drawn at the times shown in Table 1 from patients scheduled to have cardiac surgery where the use of large-dose aprotinin therapy was clinically indicated (reoperation, heart transplantation, combined revascularization and valve replacement, and complex valve surgery, such as the Ross procedure). Anesthetic technique included fentanyl 10–14 μg/kg, isoflurane, and propofol together with pancuronium for neuromuscular blockade. After tracheal intubation and the establishment of pressure monitoring and IV access with a radial artery and central venous catheter, respectively, the patients received a test dose of 10,000 kallikrein-inhibiting units (KIU) (1.4 mg) of aprotinin solution (Trasylol™; Bayer Pharmaceuticals, Newbury, UK).

Table 1

Table 1

Five minutes after the test dose, patients allocated to the Weight-Related Dose group received an initial dose of 40,000 KIU (5.6 mg) of aprotinin solution per kilogram of body weight. This bolus consisted of two aliquots of 20,000 KIU/kg, each given over a 10-min period. A blood sample for measurement of plasma aprotinin concentration was taken from the arterial catheter after each of these two aliquots had been infused. In addition, 40,000 KIU/kg of aprotinin was added to the prime of the oxygenator. Blood samples for assay of aprotinin concentration were also obtained 5 min after commencement of bypass, every 30 min during bypass, 5 min after separation from bypass, 15 min after protamine administration, and before transfer to the intensive care unit. An additional sample was obtained 2 min after release of the aortic cross-clamp.

For the fixed-dose regimen, patients received a dosing regimen as noted in the product information and as originally described (6). For these patients, after the test dose of 10,000 KIU, the initial dose was 2 × 106 KIU (280 mg) irrespective of the weight. This dose was given over a 20-min period, after which the first blood sample for determination of aprotinin concentration was withdrawn. A further 2 × 106 KIU of aprotinin solution was added to the prime of the oxygenator, and a continuous infusion of 500,000 KIU/h (70 mg/h) was administered until the patient was transferred to the intensive care unit.

Anticoagulation was achieved by using porcine heparin, with an initial dose of 300 IU/kg and a target celite-activated clotting time (ACT) of more than 1000 s. Surgery was conducted at a temperature of 28°C–30°C, with hyperkalemic cardioplegic arrest for myocardial preservation. Hollow-fiber membrane oxygenators were used in all patients.

The oxygenator prime contained 700 mL of Hartmann solution together with 100 mL of 20% mannitol and 500 mL of a gelatin-based colloid together with 5000 IU of porcine heparin. Hemofiltration was not used, and all patients had the contents of the pump-oxygenator system reinfused at the end of the perfusion period.

Blood samples were obtained with sodium citrate (final concentration 0.3 mmol), and the plasma was removed after centrifugation at 2000 g for 15 min. Samples were stored at −20°C until assay.

Aprotinin concentrations were measured immunologically by using a slightly modified competitive enzyme-linked immunosorbent assay originally described by Müller-Esterl (7). Briefly: microtiter plates were coated with an affinity-purified rabbit antibody against aprotinin (200 μL, 4 mg/L), diluted in carbonate buffer (15.9 mmol/L Na2CO3, 35.0 mmol/L NaHCO3, pH 9.6), and incubated at 4°C overnight. Excess antibody was removed by washing with phosphate-buffered saline (PBS)-Tween (150 mmol/L NaCl, 10 mmol/L NaH2PO4 × H2O, 10 mmol/L Na2HPO4 × 2H2O, 0.05% [wt/vol] Tween 20, pH 7.4). Aprotinin and samples were diluted in PBS-Tween containing 2% (wt/vol) bovine serum albumin to bring the aprotinin concentration into the working range of the assay (78 μg/L to 10 mg/L [0.5 to 660 KIU/mL]). Sample or standard (200 μL) were added to the wells of the microtiter plate, together with aprotinin conjugated to hydrogen peroxide oxidoreductase (200 μL, 20 μg/L), and incubated at 37°C for 2 h. Unbound antigen and conjugate were removed by washing with PBS-Tween. Bound aprotinin conjugate was visualized by incubation at room temperature for 20–25 min with 200 μL substrate solution (2% [wt/vol] diammonium 2,2′-azinobis-3-[ethyl-2,3-dihydrobenzothiazole]-6-sulfonate, 0.6% [vol/vol] 30% hydrogen peroxide, 100 mmol/L citrate, 100 mmol/L phosphate, pH 4.5). The concentration of conjugate was measured as increase in absorbance at 405 nm per unit time. The increased absorbance is inversely related to the concentration of aprotinin present in the sample. The concentration of aprotinin in micrograms was converted to KIU by using the formula aprotinin (KIU) = 6.6 × aprotinin (μg). The working range of, and standard used for, the assay was 0.5 to 660 KIU/mL (78 μg/L to 10 mg/L). The intraassay coefficient of variation was 5.8%, and the interassay variation was 6.4% for the plasma concentrations under study. Statistical analysis was by analysis of variance, Student’s t-test, and linear least-squares regression, as appropriate.

Back to Top | Article Outline

Results

Twenty patients (17 men and 3 women) received the fixed dose, and 10 (8 men and 2 women) received the weight-related regimen. There was no difference in bypass times (mean [range]) between groups (73 [51–105] min for fixed and 78 [49–123] min for weight-related dosing) (Table 1). Chest tube drainage in the first 12 postoperative hours was also similar at 287 (120–470) mL and 317 (50–560) mL, respectively. Two patients in each group had a hemoglobin of <8 g/dL in the intensive care unit and received 2 U of packed red blood cells. No patient received hemostatic factors. None of the patients died in the first 30 postoperative days. Four patients (three fixed-dose and one weight-related) had a peak postoperative plasma creatinine concentration of more than 200 mmol. All these patients had heart transplantation and were taking inhibitors of angiotensin-converting enzyme inhibitors before surgery and cyclosporin after surgery.

Figure 1 shows the peak plasma concentration of aprotinin for each patient by dosing regimen. The peak concentration was observed after the initial loading dose in two patients in the Fixed-Dose group and at 5 min on bypass in all other patients. The time (mean [range]) between the end of the initial dose and the 5 min on bypass sample was 68.5 (25–100) min for patients allocated to the fixed-dose regimen, which was not significantly different from the 73 (31–128) min in the weight-related-dose patients. Mean (sem) plasma concentration in the Fixed-Dose group of 333 (21) KIU/mL was not significantly different from the 292 (19) KIU/mL measured in the Weight-Related group (P = 0.167). However, there was a larger variability in peak plasma concentrations for the fixed-dose patients, as shown by a significant difference (P = 0.036) between the mean deviation in plasma aprotinin concentration from the group mean values of 83.4 (9.8) KIU/mL for the fixed dose and 56.4 (10.5) KIU/mL for the weight-related dose. In addition, there was a significant (P < 0.001) negative correlation between the peak plasma concentration and the patient’s body weight with the fixed-dose regimen. This correlation disappeared when the weight-related dose was used (Fig. 2).

Figure 1

Figure 1

Figure 2

Figure 2

Although the range of peak concentrations was less variable with the weight-related dose, there was no benefit of this drug regimen in producing a more predictable plasma concentration throughout the remainder of the operative period. Figure 3 shows mean (sem) plasma concentrations plotted against the sample time point. There were no significant differences between the concentrations of aprotinin for the two dose regimens at any time point. The mean deviation from the mean was not significantly different at any time point. For a specific patient, there was no simple relationship between the plasma concentrations achieved during the administration of the weight-related initial dose (Fig. 4). The plasma concentration achieved after the first bolus of 20,000 KIU/kg varied between 57 and 209 KIU/mL. There was also no relationship between the increase in plasma concentrations achieved after a second bolus of 20,000 KIU/kg and that attained after the first bolus of 20,000 KIU/kg. Similarly, there was no relationship between the decrease in plasma concentration of aprotinin with time during the period of cardiopulmonary bypass (Fig. 5). There was no obvious relationship between these two variables for either regimen. In particular, the use of an infusion did not postpone or attenuate the decrease of plasma aprotinin with time.

Figure 3

Figure 3

Figure 4

Figure 4

Figure 5

Figure 5

The concentrations of aprotinin at the end of the period of bypass are shown in Figure 6. Three patients had a plasma aprotinin concentration <50 KIU/mL, 12 had a plasma concentration >125 KIU/mL, and none had a plasma concentration >200 KIU/mL at the end of bypass.

Figure 6

Figure 6

Back to Top | Article Outline

Discussion

This study has shown that the peak plasma concentration of aprotinin achieved with a fixed-dose regimen is more variable than with a weight-related regimen. The study has also confirmed the considerable variability in the plasma concentrations achieved for this drug when measured in patients before surgery (3) or with a non–weight-related regimen during the perioperative period (1,2,4). In these previous studies, individual plasma concentrations, at specified time points, showed up to a 10-fold difference in absolute values for aprotinin concentration.

The first question to address is whether there is any simple explanation for the large variability in plasma concentrations measured in this and other studies. Aprotinin concentration might be altered by a number of different processes. These include factors related to the initial uptake period, the redistribution period, and the drug elimination.

There was no temporal relationship between aprotinin concentration and the period from the initiation of dosing to extracorporeal circulation to explain the variability. Aprotinin is mainly lost from the plasma by binding to endothelium. Earlier experiments in laboratory animals showed that about 70% of the injected dose of the drug was rapidly lost from the circulation (8,9). Aprotinin has an affinity for the microvasculature within the kidney, especially for the peptidase on the brush border (10). However, only a fraction of the bound aprotinin is lost into the urine, suggesting that there is a metabolic or redistribution component, or both, in the elimination process. The complex binding, pharmacokinetics, and pharmacodynamics of aprotinin are mirrored by those of heparin, which, like aprotinin, is found in the mast cell (11,12). Heparin clearance is nonlinear, and elimination occurs by two separate processes: a rapid saturable phase caused by dose-dependent binding to endothelial cells and macrophages (13,14) and a slower phase of elimination caused by renal excretion. This complex mechanism of elimination means that as the dose of heparin is increased, the elimination half-life appears to lengthen. A bolus of 25 U/kg has an apparent half-life of 30 minutes, which increases to 60 minutes with a dose of 100 U/kg. This half-life duration is further increased to 150 minutes when the bolus dose is 400 U/kg (15,16). Aprotinin may behave in the same idiosyncratic way.

Our current inability to define the precise site of redistribution of aprotinin may also explain the unpredictability of decreases in plasma concentrations throughout the period of bypass. It is recognized that there are shifts of fluid during the bypass procedure that are compensated by the addition of fluid throughout the period of perfusion. The total volume of circulating blood is kept within relatively tight boundaries by maintaining the volume of the venous reservoir. However, this does not account for any changes in extravascular volume, into which it is assumed that aprotinin can diffuse freely (10). However, there may be quantitative differences between the leakage of aprotinin into the tissues because of alteration in disease severity between patient groups. For example, patients with cardiac failure requiring heart transplantation (three patients in the Weight-Related and four in the Fixed-Dose groups) or with valve disease (five patients in the Weight-Related and seven in the Fixed-Dose groups) may be more likely to have tissue edema as a result of their underlying decompensated disease process than patients requiring reoperative myocardial revascularization. Whether this will affect subsequent aprotinin distribution and metabolism is unknown and unstudied. To examine this hypothesis will require studies in larger numbers of patients than investigated here.

There are two caveats to the observations in this study regarding the plasma aprotinin concentrations throughout bypass. First, the observations were made during bypass periods that may be regarded by some practitioners as relatively short. The maximum period of extracorporeal circulation was 123 minutes. However, the average bypass time was only a few minutes different from the 79 minutes reported by Beath et al. (1) for their high-risk patients. Extrapolation of the results of this study to longer periods of bypass should be undertaken with caution. Second, we did not use any filtration system during the perioperative period that may alter plasma aprotinin concentrations (17).

The next question to address is whether the observed variability has any relevance to patient outcome. It could be that, as with heparin, plasma aprotinin concentration is unrelated to biological activity. Three issues are of interest related to safety, mechanism of action and the potential for giving a relative overdose of the drug. Less variability in the early operative peak plasma concentration may have safety implications in relation to control of anticoagulation. A number of studies have shown that the duration of the ACTs is increased in the presence of aprotinin therapy and that this increase is less with kaolin compared with celite activation (18–21). These observations were made in blood samples to which aprotinin had been added to obtain a plasma concentration of 200 KIU/mL, the concentration originally targeted for its antiinflammatory potential (5,22). However, at larger concentrations of aprotinin, the duration of the kaolin-activated test is also increased substantially. For example, Zucker et al. (23) showed that the duration of the kaolin ACT was about 100 seconds longer in the presence of aprotinin at concentrations more than 400 KIU/mL (with heparin at 3 IU/mL). Such aprotinin concentrations were achieved in 8 of the 30 patients in these studies, all in the Fixed-Dose group. If a target kaolin ACT of 400 seconds had been used as an index of adequate anticoagulation, then the “true,” or heparin alone, ACT may have been only 300 seconds. This value would normally be interpreted as less than optimal anticoagulation. It is for this reason that we chose to administer heparin to achieve a target ACT of 1000 seconds, irrespective of the activator used.

There are no data to show a relationship between plasma concentrations of aprotinin and tissue or organ toxicity. The 50% lethal dose for aprotinin in mice and rats is reported as 2.5 × 106 KIU/kg (350 mg/kg) (10). The variability in plasma concentrations, shown here for the weight-related and earlier with the dose-related regimen (1), will make direct associations between tissue injury and a specific or threshold plasma aprotinin concentration difficult.

There are no robust explanations for the mechanism of action of aprotinin. The fixed-dose regimen was derived as a means of establishing an antiinflammatory aprotinin concentration (5,24). The target plasma concentration of 200 KIU/mL was thought to be sufficient to inhibit kallikrein by approximately 50%. Inhibition of kallikrein activity has been demonstrated in recirculating blood at these concentrations (25), but not during clinical heart surgery with aprotinin doses of up to 60,000 KIU/kg (26). This may reflect a greater degree of kallikrein activation during surgery, and 90% inhibition of kallikrein requires concentrations closer to 500 KIU/mL (10).

The amount of aprotinin to inhibit free plasmin by 50% in the laboratory setting is approximately 50 KIU/mL. More than 90% inhibition of plasmin requires concentrations nearer 125 KIU/mL (10). The plasma concentrations achieved at the end of bypass (Fig. 6) showed great variability across this putative band of inhibiting concentrations. In particular, no patient maintained a plasma concentration of more than 200 KIU/mL to the end of the bypass, and three had concentrations <50 KIU/mL. As reported previously (5), there was no predictive or significant relationship between concentrations of aprotinin at either the peak or end of bypass or surgery and the postoperative chest tube drainage, used as a surrogate marker of efficacy (data not shown).

Despite the lack of a relationship between plasma aprotinin and chest tube drainage, there is an apparent dose-response effect for efficacy when the total aprotinin dose is considered (26,27), the so-called large-dose regimen being close to the 100% effective dose for reducing the need for hemostatic products (27). Because no patient received hemostatic products, it could be argued that plasma concentrations in excess of 400 KIU/mL were possibly not of any additional benefit, but the current study was not designed to test this hypothesis. We had anticipated a predictive relationship between total dose and plasma concentrations with the weight-related dose. This would have led to the possibility of individual smaller, but possibly as effective, dosing of aprotinin. However, this relationship was not observed (Fig. 4).

An apparent relative overdose in the early surgical period in 40% of the patients with the fixed-dose regimen has certain fiscal implications in places where the drug may be more expensive, such as North America, or where patients have a lower body weight, as in Asian populations. In this study, the patient weighing 45 kg was allocated to the weight-related regimen and received a total dose of 3.6 × 106 KIU of aprotinin. This is 50% of the 7.2 × 106 KIU the patient would have received with the fixed-dose regimen (2 × 106 KIU initial dose plus 2 × 106 KIU in the prime and 0.5 KIU/h for 6.5 hours). Expressed differently, this represents the prescription and administration of 7 rather than 15 of the 50-milliliter ampule, or 4 compared with 8 of the 100-milliliter ampule.

In conclusion, this study has shown that the administration of aprotinin on a weight-related basis produces less variability in peak plasma concentration but does not overcome the variability in plasma concentrations achieved throughout the bypass period. The reduced peak concentration should be associated with more confidence in the appropriate control of anticoagulation. The variability in the time-concentration curves, even with a weight-related dose, suggests that large clinical trials will be required to address the precise mechanism of action of aprotinin and the optimal dose for a specific patient population.

Back to Top | Article Outline

References

1. Beath SM, Nuttall GA, Fass DN, et al. Plasma aprotinin concentrations during cardiac surgery: full- versus half-dose regimens. Anesth Analg 2000; 91: 257–64.
2. Bennett-Guerrero E, Sorohan JG, Howell ST, et al. Maintenance of therapeutic plasma aprotinin levels during prolonged cardiopulmonary bypass using a large-dose regimen. Anesth Analg 1996; 83: 1189–92.
3. Levy JH, Bailey JM, Salmenpera M. Pharmacokinetics of aprotinin in preoperative cardiac surgical patients. Anesthesiology 1994; 80: 1013–8.
4. O’Connor CJ, Brown DV, Avramov M, et al. The impact of renal dysfunction on aprotinin pharmacokinetics during cardiopulmonary bypass. Anesth Analg 1999; 89: 1101–7.
5. Royston D. High-dose aprotinin therapy: a review of the first five years’ experience. J Cardiothorac Vasc Anesth 1992; 6: 76–100.
6. Royston D, Bidstrup BP, Taylor KM, Sapsford RN. Effect of aprotinin on need for blood transfusion after repeat open-heart surgery. Lancet 1987; 2 (8571): 1289–91.
7. Müller-Esterl W. Drugs monitored during therapy: aprotinin, drugs and pesticides. In: Bergmeyer H, ed. Methods of enzymatic analysis. Weinheim: VerlagsgesellschaftmbH, 1986: 246–56.
8. Beller FK, Epstein MD, Kaller H. Distribution, half life time and placental transfer of the protease inhibitor Trasylol. Thromb Diath Haemorrh 1966; 16: 302–10.
9. Habermann E, Arndts D, Just M, et al. Behavior of Trasylol in the body as a model for the pharmacokinetics of basic polypeptides. Med Welt 1973; 24: 1163–7.
10. Fritz H, Wunderer G. Biochemistry and applications of aprotinin, the kallikrein inhibitor from bovine organs. Arzneimittelforschung 1983; 33: 479–94.
11. Fritz H, Kruck J, Russe I, Liebich HG. Immunofluorescence studies indicate that the basic trypsin-kallikrein-inhibitor of bovine organs (Trasylol) originates from mast cells. Hoppe Seylers Z Physiol Chem 1979; 360: 437–44.
12. Shikimi T, Kobayashi T. Production of antibody to aprotinin and location of this compound in bovine tissue. J Pharmacobiodyn 1980; 3: 400–6.
13. Glimelius B, Busch C, Hook M. Binding of heparin on the surface of cultured human endothelial cells. Thromb Res 1978; 12: 773–82.
14. Mahadoo J, Heibert L, Jaques LB. Vascular sequestration of heparin. Thromb Res 1978; 12: 79–90.
15. Bjornsson TD, Wolfram KM, Kitchell BB. Heparin kinetics determined by three assay methods. Clin Pharmacol Ther 1982; 31: 104–13.
16. de Swart CA, Nijmeyer B, Roelofs JM, Sixma JJ. Kinetics of intravenously administered heparin in normal humans. Blood 1982; 60: 1251–8.
17. Van Norman GA, Patel MA, Chandler W, Vocelka C. Effects of hemofiltration on serum aprotinin levels in patients undergoing cardiopulmonary bypass. J Cardiothorac Vasc Anesth 2000; 14: 253–6.
18. Wang JS, Lin CY, Hung WT, et al. In vitro effects of aprotinin on activated clotting time measured with different activators. J Thorac Cardiovasc Surg 1992; 104: 1135–40.
19. Feindt P, Seyfert UT, Volkmer I, et al. Celite and kaolin produce differing activated clotting times during cardiopulmonary bypass under aprotinin therapy. Thorac Cardiovasc Surg 1994; 42: 218–21.
20. Dietrich W, Jochum M. Effect of celite and kaolin on activated clotting time in the presence of aprotinin: activated clotting time is reduced by binding of aprotinin to kaolin. J Thorac Cardiovasc Surg 1995; 109: 177–8.
21. Jobes DR. Safety issues in heparin and protamine administration for extracorporeal circulation. J Cardiothorac Vasc Anesth 1998; 12 (2 Suppl 1):17–20.
22. Royston D. The serine antiprotease aprotinin (Trasylol): a novel approach to reducing postoperative bleeding. Blood Coagul Fibrinolysis 1990; 1: 55–69.
23. Zucker M, Walker C, Jobes D, LaDuca F. Comparison of celite and kaolin based heparin and protamine dosing assays during cardiac surgery: the in vitro effects of aprotinin. J Extracorporeal Technol 1995; 27: 201–7.
24. Royston D. Preventing the inflammatory response to open-heart surgery: the role of aprotinin and other protease inhibitors. Int J Cardiol 1996; 53 (Suppl): S11–37.
25. Wachtfogel YT, Kucich U, Hack CE, et al. Aprotinin inhibits the contact, neutrophil, and platelet activation systems during simulated extracorporeal perfusion. J Thorac Cardiovasc Surg 1993; 106: 1–9.
26. Fuhrer G, Gallimore MJ, Heller W, Hoffmeister HE. Aprotinin in cardiopulmonary bypass: effects on the Hageman factor (FXII)—Kallikrein system and blood loss. Blood Coagul Fibrinolysis 1992; 3: 99–104.
27. Royston D. Aprotinin versus lysine analogues: the debate continues. Ann Thorac Surg 1998; 65 (4 Suppl):S9–19.
© 2001 International Anesthesia Research Society