NEUROSURGICAL ANESTHESIA Section Editor Donald S. Prough
Influence of Hydroxyethyl Starch on Coagulation in Patients During the Perioperative Period Yves Claes, MD, Jan Van Hemelrijck, MD, Mark Van Gerven, MD, Jozef k n o u t , PhD, Jos Vermylen, MD, PhD, Burchard Weidler, MD, PhD, and Hugo Van &en, MD, PhD Department of Anesthesiology and Center for Thrombosis and Vascular Research, University Hospitals of the Katholieke Universiteit, Leuven, Belgium
The perioperative use of hydroxyethyl starch (HES) has been implicated as a possible cause of intracranial bleeding. The purpose of this study was to compare the influence on blood coagulation of the isovolemic replacement of 1-L blood loss with either 6% HES (molecular weight [MW]average: 450,000) or 5% human albumin during neurosurgery or lower abdominal surgery. Twenty patients scheduled for brain tumor surgery and 20 patients undergoing transabdominal hysterectomy were studied. The activated partial thromboplastin time, prothrombin time, fibrinogen concentration, factor VIII coagulant, von Willebrand factor antigen, platelet count, and the activated clotting time were compared after induction of anesthesia, after administration of 500 and 1000 mL
H
ydroxyethyl starch (HES), a complex polysaccharide (molecular weight [MW] average: 450,000), is available in 6% aqueous solution for use as a plasma substitute. Desirable characteristics of HES are its oncotic properties (l),the long duration of its hemodynamic effects due to its long half-life ( 2 4 , the absence of any risk of transmitting infectious diseases, the low incidence of anaphylactoid reactions (8,9), and the low cost compared with human albumin (10,ll); however, the effect of HES on hemostasis is still controversial. Although the effect on hemostasis appears to be minimal when HES is used in recommended amounts (25 mLkg body wt) (9,12), several authors have demonstrated an important influence on the factor VIII complex, resulting in a von Willebrand-like syndrome (13-15). Several cases of intracranial coagulopathies associated with the use of HES have been reported in neurosurgical patients when HES was used for peroperative intravascular volume replacement during Accepted for publication February 20, 1992. Address correspondence to Dr. Van Aken, Department of Anesthesiology, University Hospitals of the Katholieke Universiteit Leuven, Herestraat 49, B-3000 Leuven, Belgium.
24
Anesth Analg 1992;75:24-30
of colloid solution, and 24 and 48 h postoperatively. All measured coagulation variables remained within physiologic range. Changes in coagulation indices were identical in neurosurgical and hysterectomy patients, except for a larger increase in fibrinogen concentration 24 and 48 h after hysterectomy. The acute phase reaction of factor VIII coagulant and von Willebrand factor, which plays a role in postoperative hypercoagulability, was attenuated by the use of HES.We conclude that isovolemicreplacement of I-L blood loss with either 6% HES (MW average: 450,000) or 5% human albumin does not interfere with normal hemostasis during and after neurosurgery or lower abdominal surgery. (Anesth Analg 1992;75:24-30)
aneurysm surgery (16,17) or to treat vasospasm by hypervolemic hernodilution (15,18).The common factor in these case reports was the use of large amounts of HES (>lo00 mL) or administration during several days. All the patients developed a von Willebrandlike and a hemophilia-like syndrome, characterized by a marked decrease in factor VIII coagulant (VII1:C) activity and von Willebrand factor antigen (vWF:Ag); the activated partial thromboplastin time (aPTT) was markedly prolonged. Using thromboelastography, Siege1et al. (19) have demonstrated the effects of HES (and dextran) on coagulation in vitro: both fibrin formation and clot strength were impaired. Stump et al. (14) demonstrated the effect of HES on factor VIII and a m by administering 1000 mL of a 6% HES solution to healthy volunteers after removal of 500 mL of blood; the reported decrease in factor VIII and the prolongation of the aMT were greater than would be expected if they were simply the result of hemodilution. It may be that patients undergoing brain surgery are especially vulnerable to coagulation disorders, possibly as a result of defibrination after brain tissue destruction (20). Any drug with an effect on fibrin clot formation may aggravate the problem. The purpose of the present study was to evaluate 01992 by the International Anesthesia Research society 0003-2999/!32/$5.00
ANESTH ANALG
NEUROSURGICAL ANESTHESIA CLAES ET AL. HYDROXYETHYL STARCH AND COAGULATION
1992;75:2440
the influence of 1000 mL of HES (MW average: 450,000) on per- and postoperative coagulation and to determine whether this influence is different in patients undergoing intracranial and intraabdominal procedures. One-thousand milliliters of HES is the amount of the colloid that is often used clinically during the penoperative period. This dose is below the recommended upper limit dose of 25 mL/kg, which has been shown not to cause any clinically si@icant alteration in coagulation.
25
Table 1. Demographic Variables of 20 Neurosurgery and 20 Hysterectomy Patients Study group Hysterectomy HES
HA Neurosurgery HES
HA
Age (Y4 (mean value 2
SD)
Gender distribution (femalelmale)
47 -c 9 47 2 10
1010 10/0
49 2 15
6/3 8/3
49 2 15
HES,6%hydroxyethyl starch; HA, 5% human albumin.
Methods The study was approved by the Ethical Committee for Research in Patients of the Katholieke Universiteit Leuven, and written, informed consent was obtained for all patients. Twenty patients scheduled for brain tumor surgery and 20 patients scheduled for transabdominal hysterectomy were studied. The expected blood loss for each surgical procedure was between 500 and lo00 mL. All patients were ASA physical status classification I or 11, were between 18 and 65 yr of age, and had a body weight between 60 and 75 kg. Exclusion criteria were preexisting coagulopathies; abnormal preoperative coagulation saeening tests (prothrombin time [PT], a m , platelet count); intake of drugs affecting hemostasis within 2 wk preoperatively (coumarin, heparin, acetylsalicylic acid, estrogens); as well as liver (bilirubin > 1 mg/dL, alkaline phosphatase > 135 U/L) or kidney (serum creatinine > 1.4 mg/dL) dysfunction (21). Within each group, patients were randomly assigned to receive lo00 mL of fluid for volume replacement, either as a 6% HES (Plasmasteril, Fresenius, Germany) or a 5% human albumin solution in 0.9% NaC1. Intravascular volume replacement was performed isovolumetrically, and right atrium pressures were continuously measured. If necessary, a hematocrit of 35% was restored during the first 12 h after surgery by administration of erythrocyte concentrates. The anesthetic technique included thiopental, fentanyl, pancuronium, and isoflurane in all patients. The lungs of all patients were mechanically ventilated with a 60% nitrous oxide in oxygen mixture. In each patient, venous blood samples were obtained from a nonheparinized central venous catheter. Sampling times were immediately after induction of anesthesia (sample l), after administration of 500 mL (sample 2 ) and 10oO mL (sample 3) of the colloid solution, and 24 (sample 4) and 48 h (sample 5) postoperatively. The following coagulation tests were performed on a atrated blood sample at each sampling time: aPTT, IT,thrombin time, fibrinogen, factor W C , and the vW:Ag. Factor vIII:C was measured in a one-stage assay adapted to an auto-
mated coagulometer ACL-810 (IL, Milan, Italy) using severe hemophilia A plasma and a micronized silica aP?T reagent (IL, Milan, Italy). Pooled plasma, obtained from 20 healthy donors and calibrated against the second international standard for factor VIII and vWF activities in plasma, was used as a reference in all assays. The vWF:Ag was measured with an enzyme-linked immunosorbent assay (ELISA) based on two murine monoclonal antibodies raised against human vWF (22). Platelet count was performed on an anticoagulated blood sample (EDTA). An activated clotting time was measured with the Hemochron 400 (International Technodyne Corporation, Metuchen, N.J.) on a fresh blood sample. Data are reported as mean values and standard deviation (sD). One-way analysis of variance with Bonferroni's correction and Tukey's test for multiple comparisons were used for statistical analysis, and P < 0.05 was considered significant.
Results Although the four treatment groups were identical with regard to age and weight, the hysterectomy groups obviously included female patients only, unlike the neurosurgery groups (Table 1). The results of the coagulation tests are summarized in Tables 2-9. The aPIT did not change significantly during the study period (Table 2). Although a significant decrease in PT, expressed as percent of normal (P < 0.001), was observed after administration of 1 L of colloid solution, as well as 24 h postoperatively, the PT returned to normal within 48 h, and the changes were identical in all treatment groups (Table 3). In all patients, a significant increase in fibrinogen concentration (P < 0.001) was observed 24 and 48 h postoperatively (Table 4). The increase in fibrinogen concentration was more pronounced in gynecologic than in neurosurgical patients; however, no difference could be observed between patients receiving HES and human albumin in each surgery group. A
26
ANESTH ANALG
NEUROSURGICAL ANESTHESIA CLAES ET AL. HYDROXYETHYL STARCH AND COAGULATION
1992;75:2430
Table 2. Activated Partial Thromboplastin Time (s) Hysterectomy group
Neurosurgical group
Sample time
HES (n = 10)
HA (n = 10)
HES (n = 11)
(n = 9)
After induction 0.5 L of colloid 1 L of colloid 24 h postoperatively 48 h postoperatively
31.2 f 2.7 31.7 f 6.2 32.8 f 7 32.6 f 4.5 33.8 f 5.5
31.5 f 2.1 31.5 f 2.9 31.1 f 3.6 29.7 f 4.7 30.3 f 5
31.2 f 4.8 31.7 f 3.9 32.4 f 5.6 32.6 f 3.5 30.1 +. 2.3
29.7 f 5.2 29.6 f 4.9 29.5 f 4.9 28.3 f 3 27.4 f 2.1
HA
HES, 6%hydroxyethyl starch; HA, 5% human albumin. Data are expressed as mean values -C SD.
Table 3. Prothrombin Time (%) Hysterectomy group
Neurosurgical group
HES
HA
Sample time
(n = 10)
(n = 10)
HES (n = 11)
After induction 0.5 L of colloid 1 L of colloid 24 h postoperatively 48 h postoperatively
86.3 f 16.7 77.4 f 14.7" 64.5f 15.8" 61.5 f 16" 74.6 f 27
81.6 f 15.1 73.5 f 14.7" 67.8 f 15.8" 66.1 f 16.3" 79.4 f 17.8
84.9 f 17.2 78.7 f 17.1" 75.8 f 18.4" 70.9 f 19.6" 78.7 f 17.8
HA (n = 9) ~
~~~
82.7 f 19.6 73.6 f 16" 69.1 f 18.1" 67.2 f 14.7" 88.2 f 14.6
HES, 6%hydroxyethyl starch; HA, 5% human albumin. Data are presented as mean values 2 SD. "Change over time, P < 0.001;no significant differences between groups.
Table 4. Fibrinogen (g/L) Neurosurgical group
Hysterectomy group Sample time
HES (n = 10)
HA (n = 10)
HES (n = 11)
After induction 0.5 L of colloid 1 L of colloid 24 h postoperatively 48 h postoperatively
2.20 f 0.47 2.01 f 0.39 1.972 0.46 3.37 f 0.81" 4.60 f 1.25"
2.58 f 0.60 2.33 f 0.58 2.03 f 0.64 3.71 f 1.25" 5.63 f 1.16"
2.65 f 0.79 2.40 f 0.67 2.58 f 0.71 3.17 f 0.55" 4.24 f 0.999
1.98 f 1.07 1.75 f 0.80 1.72 f 0.87 2.33 f 0.66" 3.07 f 1.17"
HES, 6% hydroxyethyl starch; HA, 5% human albumin. Data are presented as mean values & SD. "Change over time, P < 0.001;difference between groups, P < 0.01.
shortening of the thrombin time was observed at sample times 3, 4, and 5 in patients receiving HES (Table 5). Factors VIII:C and vWF:Ag were significantly increased (P < 0.001) 24 and 48 h after surgery in the groups receiving human albumin (Tables 6 and 7). In contrast, no significant changes in factor VII1:C and vWF:Ag were observed for all measurement times when HES was administered. Table 8 shows the individual values of factor VII1:C after the administration of 1 L of colloid. Factor VII1:C was 4 0 % in one gynecologic patient (Table 8). Although a significant decrease in platelet count was observed at the end of the operation (sample 3), the variability among patients was high, and there
was no difference between treatment groups (Table 9). The activated clotting time values were similar for all measurement times and treatment groups; however, they were characterized by a large interpatient variability (Table 10). None of the patients lost an abnormally large quantity of blood or experienced a clinically perceptible coagulation disorder.
Discussion We were unable to demonstrate any difference in the changes in coagulation between neurosurgical and gynecologic patients receiving 1 L of a 6% HES (MW average: 450,000) solution for isovolemic intravascu-
NEUROSURGICAL ANESTHESIA CLAES ET AL. HYDROXYETHYL STARCH AND COAGULATION
ANESTH ANALG 1992:752440
27
Table 5. Thrombin Time (s) Neurosurgical group
Hysterectomy group
HA
HES (n = 10)
Sample time
(n = 10)
HA
HES = 11)
(n = 9)
22.0 f 2.1 18.2 2 1.9 16.0 t l.Yb 16.8 t 2 . F b 17.2 t 1.7"*b
23.8 f 2.4 23.0 2 2.6 21.1 f 2.5b 18.5 f 2.7b 19.4 ? 2.3''
(n
________~~
After induction 0.5 L of colloid 1 L of colloid 24 h postoperatively 48 h postoperatively
21.1 f 2.4 21.9 2 1.5 14.9 f l.Fb 16.8 2 2.ffb 17.5 2 2 . P b
21.6 f 1.84 20.8 f 2.0 20.4 f 2.3b 19.7 2 3.4b 20.1 f 2.4b
HES, 6% hydroxyethyl starch; HA, 5% human albumin. Data are presented as mean values ? SD. 'Change over time, P < 0.001.
Werences between groups, P < 0.01.
Table 6. Factor VIII Coagulant Activity (%) ~~
Neurosurgical group
Hysterectomy group
HES (n = 10)
HA
HES
Samule time
(n = 10)
(n = 11)
HA (n = 9)
After induction 0.5 L of colloid 1 L of colloid 24 h postoperatively 48 h postoperatively
129 f 56 148 f 58 113 2 50 109 lt 66# 147 f 89"
100 f 57 117 2 100 144 f 139 226 lWb 224 f 106",b
93 f 36 106 f 52 87 f 29 97 f 41" 136 r 60"
226 4 109 228 f 82 216 f 83 383 2 14Pb 428 f 14Fb
*
HES, 6% hydroxyethyl starch; HA, 5% human albumin. Data are presented as mean values 2 SD. "Difference between groups, P < 0.001. *Changeover time, P < 0.001.
Table 7. von Willebrand Factor Antigen (%) Neurosurgical group
Hysterectomy group sample time
HES (n = 10)
HA (n = 10)
After induction 0.5 L of colloid 1 L of colloid 24 h postoperatively 48 h postoperatively
144 f 47 201 f 249 122 f 53 130 2 57" 147 2 78"
112 53 111 f 51 106 f 53 187 f 5 P b 256 f 32",b
*
HA
HES (n = 11)
(n = 9)
125 f 41 111 f 52 104 2 33 100 f 51" 130 f 61"
144266 118 f 24 118 2 36 170 f 4CPb 212 f 67,'
HES,6% hydroxyethyl starch; HA, 5% human albumin. Data are presented as mean values 2 SD. "Change over time, P < 0.001. Werence between groups, P < 0.01.
lar volume replacement. Moreover, all the changes in coagulation that could be observed in patients treated with 1 L of a 6% HES or a 5% human albumin solution were within physiologic limits and did not interfere with normal hemostasis. Therefore, 1 L of a 6% HES or a 5% human albumin solution appears to be devoid of any negative influence on coagulation in surgical patients; however, these conclusions may not be valid if larger amounts of colloids are administered. There may also be a difference between neurosurgical patients and patients undergoing other types of surgery in these circumstances.
Although the gender distribution of both surgery groups was obviously different, patients treated with estrogens were excluded. No major differences have been reported between coagulation in male and female patients. Moreover, the gender distribution of patients receiving HES or human albumin was identical. An important observation appears to be the effect of HES on the factor VIII complex. Bleeding in a neurosurgical patient is obviously disastrous. Therefore, it is important that coagulant levels do not decrease below dangerous limits. In treating patients
28
ANESTH ANALG
NEUROSURGICAL ANESTHESIA CLAES ET AL. HYDROXYETHYL STARCH AND COAGULATION
1992:75:2m
Table 8. Individual Values of Factor VIII Coagulant Activity (%) After Administration of 1 L of Colloid Hysterectomy group HES HA
Patient
No. ~~~~
~
1 2 3 4 5 6 7 8 9 10 11
~~~
78 62 79 55 38 143 63 218 158 79
values in patients receiving 1L of a 6%HES solution. An increase in factor VIII activity plays a major role in postoperative hypercoagulability (12,23). We speculate that the peroperative administration of 1 L of HES may exert a protective influence against postoperative thromboembolism by inhibiting the postoperative increase in factor VIII activity and by the rheologic effect of hemodilution. This speculation is supported by the observations of Heilmann et al. (24), who found a lower incidence of deep venous thrombosis after cesarean section in patients who received 1500 mL of HES compared with patients treated with low-dose heparin. The effect of HES on the factor VIII complex is similar to the effect of dextran (25); however, the antihemostatic effect of dextran appears to be more pronounced. Although HES and dextran are both complex polysaccharides, their molecular structures are different, and it has not yet been demonstrated that their effects on coagulation on the molecular scale are identical. When compared to baseline, a significant decrease in PT values, expressed in percent, was observed after administration of 1000 mL of colloid, and 24 h after operation. This observation was probably the result of dilution of the coagulation factors by approximately 25%. The PT returned to baseline values within 48 h. A significant increase in fibrinogen concentration was observed in all groups 24 and 48 h after surgery. This has been previously reported (23,26) as being the result of the acute phase reaction. Of note, the increase in fibrinogen appears to be more pronounced in patients undergoing hysterectomy than in neurosurgical patients, probably due to a more extensive surgical trauma. This may contribute to the high incidence of postoperative thromboembolic accidents after lower abdominal surgery. The significant shortening of the thrombin times after infusion of 1L of HES, reflecting an accelerated conversion of fibrinogen to fibrin, may indicate the influence of HES on the polymerization of fibrin monomers. This observation is comparable to the
Neurosurgery group HES HA
~
100 58 110 169 67 200 105 241 67 91
77
111
98 183 111 86 125 99
234 97 71 92 118 77 106 60
54
97 73 79
HES, 6% hydroxyethyl starch; HA, 5% human albumin.
with hemophilia A undergoing neurosurgery, most hematologists would maintain factor VII1:C activity >40%-50%. Although no difference was observed between the groups, factor VII1:C decreased to 38%of the preoperative level after the administration of 1 L of HES in one patient undergoing hysterectomy. The lowest level of factor VII1:C observed in the neurosurgical patients was 54%. The probability that a clinically important (factor VII1:C < 50%)difference in factor VII1:C between HES- and human albumintreated neurosurgical patients would have been detected by the test procedure is 92%, because the calculated @errorfor the observed factor VII1:C value of 87% k 29%after 1 L of HES and for the number of patients studied is 0.084. Factor VIII activity levels
Influence of Hydroxyethyl Starch on Coagulation in Patients During the Perioperative Period Yves Claes, MD, Jan Van Hemelrijck, MD, Mark Van Gerven, MD, Jozef k n o u t , PhD, Jos Vermylen, MD, PhD, Burchard Weidler, MD, PhD, and Hugo Van &en, MD, PhD Department of Anesthesiology and Center for Thrombosis and Vascular Research, University Hospitals of the Katholieke Universiteit, Leuven, Belgium
The perioperative use of hydroxyethyl starch (HES) has been implicated as a possible cause of intracranial bleeding. The purpose of this study was to compare the influence on blood coagulation of the isovolemic replacement of 1-L blood loss with either 6% HES (molecular weight [MW]average: 450,000) or 5% human albumin during neurosurgery or lower abdominal surgery. Twenty patients scheduled for brain tumor surgery and 20 patients undergoing transabdominal hysterectomy were studied. The activated partial thromboplastin time, prothrombin time, fibrinogen concentration, factor VIII coagulant, von Willebrand factor antigen, platelet count, and the activated clotting time were compared after induction of anesthesia, after administration of 500 and 1000 mL
H
ydroxyethyl starch (HES), a complex polysaccharide (molecular weight [MW] average: 450,000), is available in 6% aqueous solution for use as a plasma substitute. Desirable characteristics of HES are its oncotic properties (l),the long duration of its hemodynamic effects due to its long half-life ( 2 4 , the absence of any risk of transmitting infectious diseases, the low incidence of anaphylactoid reactions (8,9), and the low cost compared with human albumin (10,ll); however, the effect of HES on hemostasis is still controversial. Although the effect on hemostasis appears to be minimal when HES is used in recommended amounts (25 mLkg body wt) (9,12), several authors have demonstrated an important influence on the factor VIII complex, resulting in a von Willebrand-like syndrome (13-15). Several cases of intracranial coagulopathies associated with the use of HES have been reported in neurosurgical patients when HES was used for peroperative intravascular volume replacement during Accepted for publication February 20, 1992. Address correspondence to Dr. Van Aken, Department of Anesthesiology, University Hospitals of the Katholieke Universiteit Leuven, Herestraat 49, B-3000 Leuven, Belgium.
24
Anesth Analg 1992;75:24-30
of colloid solution, and 24 and 48 h postoperatively. All measured coagulation variables remained within physiologic range. Changes in coagulation indices were identical in neurosurgical and hysterectomy patients, except for a larger increase in fibrinogen concentration 24 and 48 h after hysterectomy. The acute phase reaction of factor VIII coagulant and von Willebrand factor, which plays a role in postoperative hypercoagulability, was attenuated by the use of HES.We conclude that isovolemicreplacement of I-L blood loss with either 6% HES (MW average: 450,000) or 5% human albumin does not interfere with normal hemostasis during and after neurosurgery or lower abdominal surgery. (Anesth Analg 1992;75:24-30)
aneurysm surgery (16,17) or to treat vasospasm by hypervolemic hernodilution (15,18).The common factor in these case reports was the use of large amounts of HES (>lo00 mL) or administration during several days. All the patients developed a von Willebrandlike and a hemophilia-like syndrome, characterized by a marked decrease in factor VIII coagulant (VII1:C) activity and von Willebrand factor antigen (vWF:Ag); the activated partial thromboplastin time (aPTT) was markedly prolonged. Using thromboelastography, Siege1et al. (19) have demonstrated the effects of HES (and dextran) on coagulation in vitro: both fibrin formation and clot strength were impaired. Stump et al. (14) demonstrated the effect of HES on factor VIII and a m by administering 1000 mL of a 6% HES solution to healthy volunteers after removal of 500 mL of blood; the reported decrease in factor VIII and the prolongation of the aMT were greater than would be expected if they were simply the result of hemodilution. It may be that patients undergoing brain surgery are especially vulnerable to coagulation disorders, possibly as a result of defibrination after brain tissue destruction (20). Any drug with an effect on fibrin clot formation may aggravate the problem. The purpose of the present study was to evaluate 01992 by the International Anesthesia Research society 0003-2999/!32/$5.00
ANESTH ANALG
NEUROSURGICAL ANESTHESIA CLAES ET AL. HYDROXYETHYL STARCH AND COAGULATION
1992;75:2440
the influence of 1000 mL of HES (MW average: 450,000) on per- and postoperative coagulation and to determine whether this influence is different in patients undergoing intracranial and intraabdominal procedures. One-thousand milliliters of HES is the amount of the colloid that is often used clinically during the penoperative period. This dose is below the recommended upper limit dose of 25 mL/kg, which has been shown not to cause any clinically si@icant alteration in coagulation.
25
Table 1. Demographic Variables of 20 Neurosurgery and 20 Hysterectomy Patients Study group Hysterectomy HES
HA Neurosurgery HES
HA
Age (Y4 (mean value 2
SD)
Gender distribution (femalelmale)
47 -c 9 47 2 10
1010 10/0
49 2 15
6/3 8/3
49 2 15
HES,6%hydroxyethyl starch; HA, 5% human albumin.
Methods The study was approved by the Ethical Committee for Research in Patients of the Katholieke Universiteit Leuven, and written, informed consent was obtained for all patients. Twenty patients scheduled for brain tumor surgery and 20 patients scheduled for transabdominal hysterectomy were studied. The expected blood loss for each surgical procedure was between 500 and lo00 mL. All patients were ASA physical status classification I or 11, were between 18 and 65 yr of age, and had a body weight between 60 and 75 kg. Exclusion criteria were preexisting coagulopathies; abnormal preoperative coagulation saeening tests (prothrombin time [PT], a m , platelet count); intake of drugs affecting hemostasis within 2 wk preoperatively (coumarin, heparin, acetylsalicylic acid, estrogens); as well as liver (bilirubin > 1 mg/dL, alkaline phosphatase > 135 U/L) or kidney (serum creatinine > 1.4 mg/dL) dysfunction (21). Within each group, patients were randomly assigned to receive lo00 mL of fluid for volume replacement, either as a 6% HES (Plasmasteril, Fresenius, Germany) or a 5% human albumin solution in 0.9% NaC1. Intravascular volume replacement was performed isovolumetrically, and right atrium pressures were continuously measured. If necessary, a hematocrit of 35% was restored during the first 12 h after surgery by administration of erythrocyte concentrates. The anesthetic technique included thiopental, fentanyl, pancuronium, and isoflurane in all patients. The lungs of all patients were mechanically ventilated with a 60% nitrous oxide in oxygen mixture. In each patient, venous blood samples were obtained from a nonheparinized central venous catheter. Sampling times were immediately after induction of anesthesia (sample l), after administration of 500 mL (sample 2 ) and 10oO mL (sample 3) of the colloid solution, and 24 (sample 4) and 48 h (sample 5) postoperatively. The following coagulation tests were performed on a atrated blood sample at each sampling time: aPTT, IT,thrombin time, fibrinogen, factor W C , and the vW:Ag. Factor vIII:C was measured in a one-stage assay adapted to an auto-
mated coagulometer ACL-810 (IL, Milan, Italy) using severe hemophilia A plasma and a micronized silica aP?T reagent (IL, Milan, Italy). Pooled plasma, obtained from 20 healthy donors and calibrated against the second international standard for factor VIII and vWF activities in plasma, was used as a reference in all assays. The vWF:Ag was measured with an enzyme-linked immunosorbent assay (ELISA) based on two murine monoclonal antibodies raised against human vWF (22). Platelet count was performed on an anticoagulated blood sample (EDTA). An activated clotting time was measured with the Hemochron 400 (International Technodyne Corporation, Metuchen, N.J.) on a fresh blood sample. Data are reported as mean values and standard deviation (sD). One-way analysis of variance with Bonferroni's correction and Tukey's test for multiple comparisons were used for statistical analysis, and P < 0.05 was considered significant.
Results Although the four treatment groups were identical with regard to age and weight, the hysterectomy groups obviously included female patients only, unlike the neurosurgery groups (Table 1). The results of the coagulation tests are summarized in Tables 2-9. The aPIT did not change significantly during the study period (Table 2). Although a significant decrease in PT, expressed as percent of normal (P < 0.001), was observed after administration of 1 L of colloid solution, as well as 24 h postoperatively, the PT returned to normal within 48 h, and the changes were identical in all treatment groups (Table 3). In all patients, a significant increase in fibrinogen concentration (P < 0.001) was observed 24 and 48 h postoperatively (Table 4). The increase in fibrinogen concentration was more pronounced in gynecologic than in neurosurgical patients; however, no difference could be observed between patients receiving HES and human albumin in each surgery group. A
26
ANESTH ANALG
NEUROSURGICAL ANESTHESIA CLAES ET AL. HYDROXYETHYL STARCH AND COAGULATION
1992;75:2430
Table 2. Activated Partial Thromboplastin Time (s) Hysterectomy group
Neurosurgical group
Sample time
HES (n = 10)
HA (n = 10)
HES (n = 11)
(n = 9)
After induction 0.5 L of colloid 1 L of colloid 24 h postoperatively 48 h postoperatively
31.2 f 2.7 31.7 f 6.2 32.8 f 7 32.6 f 4.5 33.8 f 5.5
31.5 f 2.1 31.5 f 2.9 31.1 f 3.6 29.7 f 4.7 30.3 f 5
31.2 f 4.8 31.7 f 3.9 32.4 f 5.6 32.6 f 3.5 30.1 +. 2.3
29.7 f 5.2 29.6 f 4.9 29.5 f 4.9 28.3 f 3 27.4 f 2.1
HA
HES, 6%hydroxyethyl starch; HA, 5% human albumin. Data are expressed as mean values -C SD.
Table 3. Prothrombin Time (%) Hysterectomy group
Neurosurgical group
HES
HA
Sample time
(n = 10)
(n = 10)
HES (n = 11)
After induction 0.5 L of colloid 1 L of colloid 24 h postoperatively 48 h postoperatively
86.3 f 16.7 77.4 f 14.7" 64.5f 15.8" 61.5 f 16" 74.6 f 27
81.6 f 15.1 73.5 f 14.7" 67.8 f 15.8" 66.1 f 16.3" 79.4 f 17.8
84.9 f 17.2 78.7 f 17.1" 75.8 f 18.4" 70.9 f 19.6" 78.7 f 17.8
HA (n = 9) ~
~~~
82.7 f 19.6 73.6 f 16" 69.1 f 18.1" 67.2 f 14.7" 88.2 f 14.6
HES, 6%hydroxyethyl starch; HA, 5% human albumin. Data are presented as mean values 2 SD. "Change over time, P < 0.001;no significant differences between groups.
Table 4. Fibrinogen (g/L) Neurosurgical group
Hysterectomy group Sample time
HES (n = 10)
HA (n = 10)
HES (n = 11)
After induction 0.5 L of colloid 1 L of colloid 24 h postoperatively 48 h postoperatively
2.20 f 0.47 2.01 f 0.39 1.972 0.46 3.37 f 0.81" 4.60 f 1.25"
2.58 f 0.60 2.33 f 0.58 2.03 f 0.64 3.71 f 1.25" 5.63 f 1.16"
2.65 f 0.79 2.40 f 0.67 2.58 f 0.71 3.17 f 0.55" 4.24 f 0.999
1.98 f 1.07 1.75 f 0.80 1.72 f 0.87 2.33 f 0.66" 3.07 f 1.17"
HES, 6% hydroxyethyl starch; HA, 5% human albumin. Data are presented as mean values & SD. "Change over time, P < 0.001;difference between groups, P < 0.01.
shortening of the thrombin time was observed at sample times 3, 4, and 5 in patients receiving HES (Table 5). Factors VIII:C and vWF:Ag were significantly increased (P < 0.001) 24 and 48 h after surgery in the groups receiving human albumin (Tables 6 and 7). In contrast, no significant changes in factor VII1:C and vWF:Ag were observed for all measurement times when HES was administered. Table 8 shows the individual values of factor VII1:C after the administration of 1 L of colloid. Factor VII1:C was 4 0 % in one gynecologic patient (Table 8). Although a significant decrease in platelet count was observed at the end of the operation (sample 3), the variability among patients was high, and there
was no difference between treatment groups (Table 9). The activated clotting time values were similar for all measurement times and treatment groups; however, they were characterized by a large interpatient variability (Table 10). None of the patients lost an abnormally large quantity of blood or experienced a clinically perceptible coagulation disorder.
Discussion We were unable to demonstrate any difference in the changes in coagulation between neurosurgical and gynecologic patients receiving 1 L of a 6% HES (MW average: 450,000) solution for isovolemic intravascu-
NEUROSURGICAL ANESTHESIA CLAES ET AL. HYDROXYETHYL STARCH AND COAGULATION
ANESTH ANALG 1992:752440
27
Table 5. Thrombin Time (s) Neurosurgical group
Hysterectomy group
HA
HES (n = 10)
Sample time
(n = 10)
HA
HES = 11)
(n = 9)
22.0 f 2.1 18.2 2 1.9 16.0 t l.Yb 16.8 t 2 . F b 17.2 t 1.7"*b
23.8 f 2.4 23.0 2 2.6 21.1 f 2.5b 18.5 f 2.7b 19.4 ? 2.3''
(n
________~~
After induction 0.5 L of colloid 1 L of colloid 24 h postoperatively 48 h postoperatively
21.1 f 2.4 21.9 2 1.5 14.9 f l.Fb 16.8 2 2.ffb 17.5 2 2 . P b
21.6 f 1.84 20.8 f 2.0 20.4 f 2.3b 19.7 2 3.4b 20.1 f 2.4b
HES, 6% hydroxyethyl starch; HA, 5% human albumin. Data are presented as mean values ? SD. 'Change over time, P < 0.001.
Werences between groups, P < 0.01.
Table 6. Factor VIII Coagulant Activity (%) ~~
Neurosurgical group
Hysterectomy group
HES (n = 10)
HA
HES
Samule time
(n = 10)
(n = 11)
HA (n = 9)
After induction 0.5 L of colloid 1 L of colloid 24 h postoperatively 48 h postoperatively
129 f 56 148 f 58 113 2 50 109 lt 66# 147 f 89"
100 f 57 117 2 100 144 f 139 226 lWb 224 f 106",b
93 f 36 106 f 52 87 f 29 97 f 41" 136 r 60"
226 4 109 228 f 82 216 f 83 383 2 14Pb 428 f 14Fb
*
HES, 6% hydroxyethyl starch; HA, 5% human albumin. Data are presented as mean values 2 SD. "Difference between groups, P < 0.001. *Changeover time, P < 0.001.
Table 7. von Willebrand Factor Antigen (%) Neurosurgical group
Hysterectomy group sample time
HES (n = 10)
HA (n = 10)
After induction 0.5 L of colloid 1 L of colloid 24 h postoperatively 48 h postoperatively
144 f 47 201 f 249 122 f 53 130 2 57" 147 2 78"
112 53 111 f 51 106 f 53 187 f 5 P b 256 f 32",b
*
HA
HES (n = 11)
(n = 9)
125 f 41 111 f 52 104 2 33 100 f 51" 130 f 61"
144266 118 f 24 118 2 36 170 f 4CPb 212 f 67,'
HES,6% hydroxyethyl starch; HA, 5% human albumin. Data are presented as mean values 2 SD. "Change over time, P < 0.001. Werence between groups, P < 0.01.
lar volume replacement. Moreover, all the changes in coagulation that could be observed in patients treated with 1 L of a 6% HES or a 5% human albumin solution were within physiologic limits and did not interfere with normal hemostasis. Therefore, 1 L of a 6% HES or a 5% human albumin solution appears to be devoid of any negative influence on coagulation in surgical patients; however, these conclusions may not be valid if larger amounts of colloids are administered. There may also be a difference between neurosurgical patients and patients undergoing other types of surgery in these circumstances.
Although the gender distribution of both surgery groups was obviously different, patients treated with estrogens were excluded. No major differences have been reported between coagulation in male and female patients. Moreover, the gender distribution of patients receiving HES or human albumin was identical. An important observation appears to be the effect of HES on the factor VIII complex. Bleeding in a neurosurgical patient is obviously disastrous. Therefore, it is important that coagulant levels do not decrease below dangerous limits. In treating patients
28
ANESTH ANALG
NEUROSURGICAL ANESTHESIA CLAES ET AL. HYDROXYETHYL STARCH AND COAGULATION
1992:75:2m
Table 8. Individual Values of Factor VIII Coagulant Activity (%) After Administration of 1 L of Colloid Hysterectomy group HES HA
Patient
No. ~~~~
~
1 2 3 4 5 6 7 8 9 10 11
~~~
78 62 79 55 38 143 63 218 158 79
values in patients receiving 1L of a 6%HES solution. An increase in factor VIII activity plays a major role in postoperative hypercoagulability (12,23). We speculate that the peroperative administration of 1 L of HES may exert a protective influence against postoperative thromboembolism by inhibiting the postoperative increase in factor VIII activity and by the rheologic effect of hemodilution. This speculation is supported by the observations of Heilmann et al. (24), who found a lower incidence of deep venous thrombosis after cesarean section in patients who received 1500 mL of HES compared with patients treated with low-dose heparin. The effect of HES on the factor VIII complex is similar to the effect of dextran (25); however, the antihemostatic effect of dextran appears to be more pronounced. Although HES and dextran are both complex polysaccharides, their molecular structures are different, and it has not yet been demonstrated that their effects on coagulation on the molecular scale are identical. When compared to baseline, a significant decrease in PT values, expressed in percent, was observed after administration of 1000 mL of colloid, and 24 h after operation. This observation was probably the result of dilution of the coagulation factors by approximately 25%. The PT returned to baseline values within 48 h. A significant increase in fibrinogen concentration was observed in all groups 24 and 48 h after surgery. This has been previously reported (23,26) as being the result of the acute phase reaction. Of note, the increase in fibrinogen appears to be more pronounced in patients undergoing hysterectomy than in neurosurgical patients, probably due to a more extensive surgical trauma. This may contribute to the high incidence of postoperative thromboembolic accidents after lower abdominal surgery. The significant shortening of the thrombin times after infusion of 1L of HES, reflecting an accelerated conversion of fibrinogen to fibrin, may indicate the influence of HES on the polymerization of fibrin monomers. This observation is comparable to the
Neurosurgery group HES HA
~
100 58 110 169 67 200 105 241 67 91
77
111
98 183 111 86 125 99
234 97 71 92 118 77 106 60
54
97 73 79
HES, 6% hydroxyethyl starch; HA, 5% human albumin.
with hemophilia A undergoing neurosurgery, most hematologists would maintain factor VII1:C activity >40%-50%. Although no difference was observed between the groups, factor VII1:C decreased to 38%of the preoperative level after the administration of 1 L of HES in one patient undergoing hysterectomy. The lowest level of factor VII1:C observed in the neurosurgical patients was 54%. The probability that a clinically important (factor VII1:C < 50%)difference in factor VII1:C between HES- and human albumintreated neurosurgical patients would have been detected by the test procedure is 92%, because the calculated @errorfor the observed factor VII1:C value of 87% k 29%after 1 L of HES and for the number of patients studied is 0.084. Factor VIII activity levels
