Systemic effects of collagen-impregnated aortoiliac Dacron vascular prostheses on platelet activation and fibrin formation J. c . A. De M o l Van Otterloo, M D , J. H. Van Bockel, M D , E. D. Ponfoort, E. Briet, M D , E. J. P. B r o m m e r , M D , J. H e r m a n s , P h D , M. R. Daha, M D ,
Leiden, The Netherlands To minimize intraoperative blood loss a watertight knitted Dacron aortoiliac prosthesis has been developed by impregnation with bovine collagen. A potential disadvantage is that collagen may be associated with an increase in thrombus formation. We conducted a prospective randomized trial to study the systemic effects of collagen-impregnated prostheses and of aortoiliac operation as such on the coagulation mechanism during the first 10 days after operation. Forty-one patients randomly received either a collagenimpregnated (n = 20) or a nonimpregnated prosthesis (n = 21). Twelve patients who underwent cholecystectomies served as controls. Three markers of the coagulation mechanism were monitored: [3-thromboglobulin, fibrinopeptide A, and fibrin/fibrinogen degradation products. We found no significant differences in median J3-thromboglobulin, fibrinopeptide A, and fibrin/fibrinogen degradation product levels between patients in the collagen-impregnated prosthesis group and patients in the nonimpregnated prosthesis group. This indicates that collagen does not stimulate the coagulation cascade any more than conventional Dacron protheses do. In a comparison of patients who underwent aortoiliac reconstruction and patients who underwent cholecystectomies, the results indicated a significant increased platelet activation and fibrin metabolism in the aortoiliac reconstruction group compared with the control group. Finally, we observed a significantly' higher preoperative fibrin metabolism in patients with vascular disease than in control subjects. This difference is attributable to the high preoperative fibrin/fibrinogen degradation product values in patients with aortic aneurysms. (J VASe SuR~ 1990;14:59-66.)
Vascular prostheses have been successfully applied in aortoiliac reconstructive surgery for more than 30 years. Different types of materials have been used to optimize results. Dacron is usually applied in the aortoiliac segment, since it has been demonstrated to provide excellent long-term patency with a low incidence of complications. However, both animal studies 1-* and clinical studies s:3 have demonstrated that Dacron is hiighly thrombogenic. MoreFrom the Departments of Surgery (Drs. De Mol Van Otterloo~ Van Bockel, Ponfoort, Hermans), Haematology (Dr. Briet), and Nephrology (Dr. Daha), UniversityHospital Leiden, and the Gaubius Institute (Dr. Brommer) Leiden, and the Department of Surgery (Dr. De Mol Van Otterloo), Westeinde Hospital The Netherlands. Supported by a grant from Meadox MedicalsInc., United States and the Stopler Company,Utrecht, The Netherlands. Reprint requests: J. H. Van Bockel,MD, Departmentof Surgery, University Hospital, P.O. Box 9600, 2300 RC, Leiden, The Netherlands. 24/1/28409
over, it has been demonstrated that platelet activation at the blood-prosthesis interface continues for up to 9 years after implantation, a'7 This quality is one of the major factors that restrict successful application of Dacron to vessel segments that have high blood flow. The capacity of knilxed Dacron prostheses to increase thrombus formation has been clinically studied by a number of investigators by both indium-11 ilabeled platelct imaging techniques s-7'9-:3 and by evaluation of systemic parameters of platelet activation. :°':2'::: indium-labeled platelet imaging is considered to be the best method with which to study platelet activation at the blood-prosthesis interface.l:'12 However, recently it has been demonstrated that indium can bind directly to prosthetic material, which calls into question the reliability of these imaging studies./4 A second disadvantage of platelet imaging is that during the operative procedure and the first few days afterward, imaging is technically not feasible. 13 Therefore only limited data 59
60 De Mol Van Otterloo et al.
are available with regard to the effect of Dacron on the coagulation system in general and on platelet accumulation at the blood-prosthesis interface in particular during these first postoperative days. Moreover, little detailed information is available on the effect of aortoiliac surgery, as such, on the coagulation system in this period. Recently, Dacron aortoiliac prostheses have been impregnated with bovine collagen to obviate the need for preclotting while the superior heating characteristics of knitted Dacron are retained) 927 Advantages are that the amount of operative blood loss may be reduced and that the prosthesis can be applied under circumstances when adequate preclotting is not possible (e.g., in heparinized patients, for ruptured aneurysm resection). However, a potential disadvantage is that a collagen-impregnated Dacron prosthesis may be more thrombogenic than its substrate. The purpose of our study was to evaluate and quantitate the systemic effects of collagenimpregnated knitted Dacron prostheses and of aortoiliac surgery on the coagulation mechanism by evaluation of parameters of platelet activation, fibrin formation, and fibrin breakdown during the first 10 days after operation. PATIENTS A N D M E T H O D S
From October 1985 to September 1988, we studied systemic effects on platelet activation and fibrin metabolism after implantation of aortoiliac Dacron vascular prostheses. Thirty-eight men and three women, with a mean age of 64 years (range, 42 to 80 years) who underwent aortoiliac reconstruction were included in the study. Patients who had systemic diseases (e.g., diabetes mellitus, cancer, specific thrombotic disorders) were excluded, since these diseases may affect platelet activation and the metabolism of fibrin. Patients who were admitted for elective aortoiliac reconstruction were stratified into groups for either arteriosclerotic aneurysmal disease or occlusive disease, since we anticipated that the type of disease could influence the coagulation mechanism. Thus 26 patients were operated on for aortic aneurysms, and 15 were operated on for occlusive vascular disease. Within each stratum, patients were randomized into one of two treatment groups. The first group consisted of 20 patients in whom collagenimpregnated prostheses (the Microvel double velour with Hemashield prosthesis, Meadox Medicals, Inc., Oakland, N.J.) were implanted, whereas the second group of 21 patients received nonimpregnated pros-
)rournal of VASCULAR SURGERY
theses (microvel double velour prosthesis, Meadox Medicals Inc.) (Table I). The collagen-impregnated prosthesis is similar to the nonimpregnated prosthesis, but the knitted fabric is impregnated with cross-linked type I bovine collagen. 16'24 As control subjects, we studied 12 patients who had elective cholecystectomies (five men and seven women) with a mean age of 55 years (range, 30 to 76 years). This enabled us to study the results in patients who were undergoing abdominal surgery without implantation of an aortoiliac prosthesis. Patients were included in the study only after their informed consent had been obtained according to the procedure that was established by the Committee on Medical Ethics of the University Hospital Leiden. The perioperative anticoagulant therapy was standardized to enable a reliable comparison of results between the two prosthesis groups and the control subjects. Patients who underwent cholecystectomies also received this standard anticoagulant regimen. This regimen included administration of heparin. Heparin prophylaxis (5000 IU) was administered subcutaneously on the evening before operations. During operations, before cross-clamping the aorta, heparin (70 IU/kg body weight) was administered intravenously. Protamine sulphate was not used during or after the operations. Postoperatively, patients who had a vascular procedure received heparin (5000 IU twice daily) and acenocoumarol. Heparin was discontinued as soon as acenocoumarol therapy resulted in a Trombo test time in the range of 120 to 140 seconds. Patients who underwent cholecystectomies received no acenocoumarol therapy, and postoperative heparin prophylaxis (5000 IU twice daily) was discontinued as soon as the patients were fully mobilized. Three separate markers of the coagulation mechanism were studied, [3-thromboglobulin (BTG), fibrinopeptide-A (FPA), and fibrin/fibrinogen degradation products. Plasma levels of [3-thromboglobulin (BTG), a platelet degranulation product, are released from the c~-granules in the platelet during platelet activation. Plasma BTG level is a wellestablished parameter of platelet activation and precedes thrombus formation. ~2"lsa8 [3-Thromboglobulin levels were carefully measured after collection of 5 ml of blood in a glass tube (Radiochemicai Centre, Amersham, England) contained ethylenediaminetetraacetic acid and theophylline. The sample was gently stirred, not shaken, and placed in ice water before centrifugation for 30 minutes at 2000g (4 ° C), within an hour ofvenipuncture. The top 500 mm 3 of plasma was stored in plastic tubes at - 7 0 ° C.
Volume 14 Number 1 July 1991
[3-Thromboglobulin was determined with a commercially available sandwich enzyme-linked immunosorbent assay kit (Asserachrom [3-thromboglobulin, Diagnostica-Stago, Asnieres, France). Fibrinopeptide A, a cleavage product of the thrombin-mediated conversion of fibrinogen to fibrin during thrombus fbrmation, was determined. Fibrinopeptide A is an indirect parameter of fibrin formation) s's° Fibrinopeptide A levels were measured after collection of !5 ml of blood in a calibrated polystyrene tube that contained 0.25 ml 0.15 mol/L sodium chloride with 250 IU heparin and 250 IU aprotinin, which was placed in ice water. The samples were centrifuged for 30 minutes at 2000 a (40 C) within an hour of venipuncture. Approximately 2.4 ml of supernatant was stored in plastic tubes at - 7 0 ° C. Fibrinopeptide A levels were determined with a commercially available radioimmunoassay kit (RIA-mat, Malfinckroclt Inc., Diagnostic Products D, St. Louis, Mo.) Finally, the plasma levels of total degradation products of both fibrin and fibrinogen (TDP) were determined. In a steady-state situation, degradation of fibrin occurs in a delicate equilibrium with its formation. The level of TDP is then a valid parameter of fibrin metabolism. However, during fibrin deposition, the TDP level is a valid parameter of the catabolic arm of fibrin metabolism, namely fibrin breakdown. Total degradation product levels were measured after collection of 5 ml blood into calibrated polystyrene tubes that contained buffered sodium citrate and aprotinin (final concentration 40 IU/ml), which were placed in ice water. The samples were centrifuged fi~r 30 minutes at 2000 g (4 ° C) within an hour of venipuncmre. Approximately 2.4 ml supernatant was stored in plastic tubes at - 70 ° C. Total degradation ]products were determined with a sandwich enzyme-linked immunosorbent assay test, with the use of monoclonal antibodies that are reactive with a specific neoantigen determinant of fibrin/fibrinogen degradation products. 17 On the basis of the results of a pilot study, blood samples were taken the day before operation (day - 1 ) ; during operation, after recirculation and before closure of the wound (day 0); and on postoperative days 2, 7, and 10. Total degradation product levels were not determined on postoperative day 10. All blood samples were obtained after careful venipuncture, which was performed by the same investigator (trial coordinator), with Wassermann needles of 1.0 mm in diameter. To avoid platelet damage by suction, nonevacuated tubes were used.
Effects of collagen-coated Dacron grafts on platdets/fibrin
61
Table I. Forty-one aortoiliac reconstructions after randomization for type of prosthesis and stratification for vascular disease Number of patients Type of disease
Collagen -
Collagen +
Total
Aneurysmal disease Occlusive disease Cholecystectomies (controls)
14 7
12 8
26 15 12
In addition, factors that may influence the coagNation mechanism during operation were monitored. Four such factors were recognized. The first factor was the amount of prosthetic material that was implanted, since this may be related to the extent of the activation of the coagulation mechanism. Either tubes or bifurcated prostheses (aortobiiliac, aortoiliofemoral, aortobifemoral) were used. The second factor was the volume of blood loss during surgery and the volume of donor blood transfusion, which were recorded for each patient. The third factor was the presence of limb ischemia in patients who were undergoing vascular surgery; this factor was assessed by determination of the duration of aortic crossclamping. Finally, the fourth factor was the presence of superficial postoperative thrombophlebitis, which is known to be a potent initiator of the coagulation mechanism and therefore could potentially distort the results (Table II)) 8 Differences of medians were evaluated with the Mann-Whitney U test, one-way or two-way analysis of variance, as appropriate. Contingency tables were evaluated with the chi-square and Fisher tests. Wilcoxon's test was ttsed to evaluate longitudinal differences. Statistical evaluations were considered to be significant ifp < 0.05. RESULTS Graft occlusion did not occur, neither in patients with collagen-impregnated prostheses nor in patients with nonimpregnated prostheses. Furthermore, we found no, significant differences in median BTG, FPA, or TDP levels betweeen patients in the collagenimpregnated prosthesis group and patients in the nonimpregnated prosthesis group (Fig. 1). This suggests that collagen does not stimulate the coagulation cascade any more than conventional Dacron prostheses do. On the contrary, the data suggested that the collagen-impreguated prosthesis group compared favorably with the nonimpregnated prosthesis group in regard to platelet activation and fibrin
62
Journal of VASCULAR SURGERY
De A4ol Van Otterloo et al.
Table II. Distribution of factors that may influence the clotting mechanism in 41 aortoiliac procedures Number of patients Collagen - prosthesis
Collagen + prosthesis
Factors
(n = 21)
(n = 20)
Type of prosthesis Tube graft (no. of) Aortobiiliac Aortoiliofemoral Aortobifemoral Mean volume of blood loss (ml) Mean volume of donor blood transfusions (ml) Mean aortic cross clamping time (min) Superficial thrombophlebitis after operation (no. of patients)
6 9 4 2 2057 1123 45 4
1 9 3 7 1834 1062 51 3
formation (Fig. 1). Median BTG, FPA, and TDP levels were increased in the nonimpregnated prosthesis group at day 0. Moreover, a statistically significant increase (p = 0.04) in median TDP in the nonimpregnated prosthesis group (TDP: 2.3 mg/ml) as compared with the collagen-impregnated prosthesis group (1.4 mg/ml) indicated increased intraoperative fibrin metabolism in the former group. In a comparison of results between patients who were undergoing aortoiliac reconstructions and patients who were undergoing cholecystectomies, the findings indicated a significantly increased platelet activation in the aortoiliac reconstruction group compared with the control group (BTG: 144 IU/ml and 39 IU/ml, p = 0.0006, respectively) (Fig. 2). Furthermore, we fomad that intraoperative fibrin metabolism overall, as reflected by FPA and TDP levels, was significantly higher during aortoiliac reconstruction than during cholecystectomy (FPA: 49 ng/ml vs 21.8 ng/ml, p = 0.014, and TDP: 2.0 mg/ml vs 0.0 mg/ml, p < 0.0001, respectively) (Fig. 2). After operation, both BTG and FPA levels initially fell, so that by day 2, they had reached preoperative levels. However, the results at day 7 and day 10 after operation indicated prolonged platelet activation and fibrin metabolism in patients who had undergone aortoiliac reconstructions. Median BTG plasma levels were increased significantly in the aortoiliac reconstruction group compared with the control group (Fig. 2). Plasma FPA levels also remained elevated until day 10 after vascular operation, whereas in control subjects FPA had dropped to normal by this stage, which suggests a prolonged fibrin formation after aortoiliac surgery (day 10:8.1 ng/ml vs 0.0 ng/ml, respectively, p = 0.03). These findings were paralleled by the median TDP plasma levels. Total degradation product levels increased significantly more in patients in the aortoiliac recon-
struction group than in patients in the control group (Fig. 2). Preoperative data (day - 1 ) demonstrated that patients who had vascular operations had an increased fibrin metabolism compared with control subjects (Fig. 2). We found a statistically significant difference in median TDP level between patients with vascular disease and control subjects (1.0 mg/ml vs 0.0 mg/ml, p = 0.0004) (Fig. 2). In contrast, preoperative BTG and FPA levels in patients who had vascular operations and control subjects were similar. To evaluate the influence of potentially thrombogenic factors on the results in both the collagenimpregnated prosthesis and the nonimpregnated prosthesis groups, the effect of the amount of prosthetic material that was implanted was studied. Aortobiiliac, aortoiliofemoral, and aortobifemoral bifurcation prostheses were evenly distributed among patients with noncollagen and coUagenimpregnated prostheses (Table II). By chance, aortic tube prostheses were significantly more present in the nonimpregnated prosthesis group (Table II). However, the mean blood-prosthesis surface area did not differ significantly between the two groups (nonimpregnated prosthesis: 6775 m m 2 vs collagenimpregnated prosthesis: 7946 mm2; p = 0.14). The volume of blood loss in the two groups was similar (2057 ml vs 1834 ml,p = 0.43, Table II), as was the volume of blood that was administered (1123 ml vs 1062 ml, p = 0.68) (Table II). Another factor was the degree ofischemia, which was represented by the duration of aortic clamping. We found no significant difference in mean clamping time (45 minutes vs 51 minutes,p = 0.24) (Table II). Finally, the last factor that was evaluated was the occurrence of superficial thrombophlebitis, which stimulates BTG release. Again, no difference was observed (4 vs 3, p = not significant, Table II).
Volume 14 Number 1 July 1991
Effects of collagen-coatedDacron grafls on platelets/fibrin 63
NS
NS
p-O.02
N8
NS
NS
p-O.O006
NS
p-O.03
I>-0.03
B'rO (lU/n~)
erro (ILUml)
Is0.
100.
5o .
.
.
.
.
.
.
.
.
.
.
.
o
ct ni
cJ rJ
ct r~
• ~ ni
cl ni
o.
-1
0
2
7
I0
NS
NS
NS
02
v
NS
NS
NS
c
v
-1
day
c
v
c
0
p-0.014
v
¢
2
v
7
NS
c
10
NS
p-0.03
FPA (ngr/mi)
FPA (l~Ftn~)
80-
4o.
¢t ni
O"
el rd
ci ni
ci ni
ci ni
-1
0
2
7
NS
P-O.04
NS
v
o10
v
day
c -1
p-O.O004
NS
vc
vc
0
2
p
Leiden, The Netherlands To minimize intraoperative blood loss a watertight knitted Dacron aortoiliac prosthesis has been developed by impregnation with bovine collagen. A potential disadvantage is that collagen may be associated with an increase in thrombus formation. We conducted a prospective randomized trial to study the systemic effects of collagen-impregnated prostheses and of aortoiliac operation as such on the coagulation mechanism during the first 10 days after operation. Forty-one patients randomly received either a collagenimpregnated (n = 20) or a nonimpregnated prosthesis (n = 21). Twelve patients who underwent cholecystectomies served as controls. Three markers of the coagulation mechanism were monitored: [3-thromboglobulin, fibrinopeptide A, and fibrin/fibrinogen degradation products. We found no significant differences in median J3-thromboglobulin, fibrinopeptide A, and fibrin/fibrinogen degradation product levels between patients in the collagen-impregnated prosthesis group and patients in the nonimpregnated prosthesis group. This indicates that collagen does not stimulate the coagulation cascade any more than conventional Dacron protheses do. In a comparison of patients who underwent aortoiliac reconstruction and patients who underwent cholecystectomies, the results indicated a significant increased platelet activation and fibrin metabolism in the aortoiliac reconstruction group compared with the control group. Finally, we observed a significantly' higher preoperative fibrin metabolism in patients with vascular disease than in control subjects. This difference is attributable to the high preoperative fibrin/fibrinogen degradation product values in patients with aortic aneurysms. (J VASe SuR~ 1990;14:59-66.)
Vascular prostheses have been successfully applied in aortoiliac reconstructive surgery for more than 30 years. Different types of materials have been used to optimize results. Dacron is usually applied in the aortoiliac segment, since it has been demonstrated to provide excellent long-term patency with a low incidence of complications. However, both animal studies 1-* and clinical studies s:3 have demonstrated that Dacron is hiighly thrombogenic. MoreFrom the Departments of Surgery (Drs. De Mol Van Otterloo~ Van Bockel, Ponfoort, Hermans), Haematology (Dr. Briet), and Nephrology (Dr. Daha), UniversityHospital Leiden, and the Gaubius Institute (Dr. Brommer) Leiden, and the Department of Surgery (Dr. De Mol Van Otterloo), Westeinde Hospital The Netherlands. Supported by a grant from Meadox MedicalsInc., United States and the Stopler Company,Utrecht, The Netherlands. Reprint requests: J. H. Van Bockel,MD, Departmentof Surgery, University Hospital, P.O. Box 9600, 2300 RC, Leiden, The Netherlands. 24/1/28409
over, it has been demonstrated that platelet activation at the blood-prosthesis interface continues for up to 9 years after implantation, a'7 This quality is one of the major factors that restrict successful application of Dacron to vessel segments that have high blood flow. The capacity of knilxed Dacron prostheses to increase thrombus formation has been clinically studied by a number of investigators by both indium-11 ilabeled platelct imaging techniques s-7'9-:3 and by evaluation of systemic parameters of platelet activation. :°':2'::: indium-labeled platelet imaging is considered to be the best method with which to study platelet activation at the blood-prosthesis interface.l:'12 However, recently it has been demonstrated that indium can bind directly to prosthetic material, which calls into question the reliability of these imaging studies./4 A second disadvantage of platelet imaging is that during the operative procedure and the first few days afterward, imaging is technically not feasible. 13 Therefore only limited data 59
60 De Mol Van Otterloo et al.
are available with regard to the effect of Dacron on the coagulation system in general and on platelet accumulation at the blood-prosthesis interface in particular during these first postoperative days. Moreover, little detailed information is available on the effect of aortoiliac surgery, as such, on the coagulation system in this period. Recently, Dacron aortoiliac prostheses have been impregnated with bovine collagen to obviate the need for preclotting while the superior heating characteristics of knitted Dacron are retained) 927 Advantages are that the amount of operative blood loss may be reduced and that the prosthesis can be applied under circumstances when adequate preclotting is not possible (e.g., in heparinized patients, for ruptured aneurysm resection). However, a potential disadvantage is that a collagen-impregnated Dacron prosthesis may be more thrombogenic than its substrate. The purpose of our study was to evaluate and quantitate the systemic effects of collagenimpregnated knitted Dacron prostheses and of aortoiliac surgery on the coagulation mechanism by evaluation of parameters of platelet activation, fibrin formation, and fibrin breakdown during the first 10 days after operation. PATIENTS A N D M E T H O D S
From October 1985 to September 1988, we studied systemic effects on platelet activation and fibrin metabolism after implantation of aortoiliac Dacron vascular prostheses. Thirty-eight men and three women, with a mean age of 64 years (range, 42 to 80 years) who underwent aortoiliac reconstruction were included in the study. Patients who had systemic diseases (e.g., diabetes mellitus, cancer, specific thrombotic disorders) were excluded, since these diseases may affect platelet activation and the metabolism of fibrin. Patients who were admitted for elective aortoiliac reconstruction were stratified into groups for either arteriosclerotic aneurysmal disease or occlusive disease, since we anticipated that the type of disease could influence the coagulation mechanism. Thus 26 patients were operated on for aortic aneurysms, and 15 were operated on for occlusive vascular disease. Within each stratum, patients were randomized into one of two treatment groups. The first group consisted of 20 patients in whom collagenimpregnated prostheses (the Microvel double velour with Hemashield prosthesis, Meadox Medicals, Inc., Oakland, N.J.) were implanted, whereas the second group of 21 patients received nonimpregnated pros-
)rournal of VASCULAR SURGERY
theses (microvel double velour prosthesis, Meadox Medicals Inc.) (Table I). The collagen-impregnated prosthesis is similar to the nonimpregnated prosthesis, but the knitted fabric is impregnated with cross-linked type I bovine collagen. 16'24 As control subjects, we studied 12 patients who had elective cholecystectomies (five men and seven women) with a mean age of 55 years (range, 30 to 76 years). This enabled us to study the results in patients who were undergoing abdominal surgery without implantation of an aortoiliac prosthesis. Patients were included in the study only after their informed consent had been obtained according to the procedure that was established by the Committee on Medical Ethics of the University Hospital Leiden. The perioperative anticoagulant therapy was standardized to enable a reliable comparison of results between the two prosthesis groups and the control subjects. Patients who underwent cholecystectomies also received this standard anticoagulant regimen. This regimen included administration of heparin. Heparin prophylaxis (5000 IU) was administered subcutaneously on the evening before operations. During operations, before cross-clamping the aorta, heparin (70 IU/kg body weight) was administered intravenously. Protamine sulphate was not used during or after the operations. Postoperatively, patients who had a vascular procedure received heparin (5000 IU twice daily) and acenocoumarol. Heparin was discontinued as soon as acenocoumarol therapy resulted in a Trombo test time in the range of 120 to 140 seconds. Patients who underwent cholecystectomies received no acenocoumarol therapy, and postoperative heparin prophylaxis (5000 IU twice daily) was discontinued as soon as the patients were fully mobilized. Three separate markers of the coagulation mechanism were studied, [3-thromboglobulin (BTG), fibrinopeptide-A (FPA), and fibrin/fibrinogen degradation products. Plasma levels of [3-thromboglobulin (BTG), a platelet degranulation product, are released from the c~-granules in the platelet during platelet activation. Plasma BTG level is a wellestablished parameter of platelet activation and precedes thrombus formation. ~2"lsa8 [3-Thromboglobulin levels were carefully measured after collection of 5 ml of blood in a glass tube (Radiochemicai Centre, Amersham, England) contained ethylenediaminetetraacetic acid and theophylline. The sample was gently stirred, not shaken, and placed in ice water before centrifugation for 30 minutes at 2000g (4 ° C), within an hour ofvenipuncture. The top 500 mm 3 of plasma was stored in plastic tubes at - 7 0 ° C.
Volume 14 Number 1 July 1991
[3-Thromboglobulin was determined with a commercially available sandwich enzyme-linked immunosorbent assay kit (Asserachrom [3-thromboglobulin, Diagnostica-Stago, Asnieres, France). Fibrinopeptide A, a cleavage product of the thrombin-mediated conversion of fibrinogen to fibrin during thrombus fbrmation, was determined. Fibrinopeptide A is an indirect parameter of fibrin formation) s's° Fibrinopeptide A levels were measured after collection of !5 ml of blood in a calibrated polystyrene tube that contained 0.25 ml 0.15 mol/L sodium chloride with 250 IU heparin and 250 IU aprotinin, which was placed in ice water. The samples were centrifuged for 30 minutes at 2000 a (40 C) within an hour of venipuncture. Approximately 2.4 ml of supernatant was stored in plastic tubes at - 7 0 ° C. Fibrinopeptide A levels were determined with a commercially available radioimmunoassay kit (RIA-mat, Malfinckroclt Inc., Diagnostic Products D, St. Louis, Mo.) Finally, the plasma levels of total degradation products of both fibrin and fibrinogen (TDP) were determined. In a steady-state situation, degradation of fibrin occurs in a delicate equilibrium with its formation. The level of TDP is then a valid parameter of fibrin metabolism. However, during fibrin deposition, the TDP level is a valid parameter of the catabolic arm of fibrin metabolism, namely fibrin breakdown. Total degradation product levels were measured after collection of 5 ml blood into calibrated polystyrene tubes that contained buffered sodium citrate and aprotinin (final concentration 40 IU/ml), which were placed in ice water. The samples were centrifuged fi~r 30 minutes at 2000 g (4 ° C) within an hour of venipuncmre. Approximately 2.4 ml supernatant was stored in plastic tubes at - 70 ° C. Total degradation ]products were determined with a sandwich enzyme-linked immunosorbent assay test, with the use of monoclonal antibodies that are reactive with a specific neoantigen determinant of fibrin/fibrinogen degradation products. 17 On the basis of the results of a pilot study, blood samples were taken the day before operation (day - 1 ) ; during operation, after recirculation and before closure of the wound (day 0); and on postoperative days 2, 7, and 10. Total degradation product levels were not determined on postoperative day 10. All blood samples were obtained after careful venipuncture, which was performed by the same investigator (trial coordinator), with Wassermann needles of 1.0 mm in diameter. To avoid platelet damage by suction, nonevacuated tubes were used.
Effects of collagen-coated Dacron grafts on platdets/fibrin
61
Table I. Forty-one aortoiliac reconstructions after randomization for type of prosthesis and stratification for vascular disease Number of patients Type of disease
Collagen -
Collagen +
Total
Aneurysmal disease Occlusive disease Cholecystectomies (controls)
14 7
12 8
26 15 12
In addition, factors that may influence the coagNation mechanism during operation were monitored. Four such factors were recognized. The first factor was the amount of prosthetic material that was implanted, since this may be related to the extent of the activation of the coagulation mechanism. Either tubes or bifurcated prostheses (aortobiiliac, aortoiliofemoral, aortobifemoral) were used. The second factor was the volume of blood loss during surgery and the volume of donor blood transfusion, which were recorded for each patient. The third factor was the presence of limb ischemia in patients who were undergoing vascular surgery; this factor was assessed by determination of the duration of aortic crossclamping. Finally, the fourth factor was the presence of superficial postoperative thrombophlebitis, which is known to be a potent initiator of the coagulation mechanism and therefore could potentially distort the results (Table II)) 8 Differences of medians were evaluated with the Mann-Whitney U test, one-way or two-way analysis of variance, as appropriate. Contingency tables were evaluated with the chi-square and Fisher tests. Wilcoxon's test was ttsed to evaluate longitudinal differences. Statistical evaluations were considered to be significant ifp < 0.05. RESULTS Graft occlusion did not occur, neither in patients with collagen-impregnated prostheses nor in patients with nonimpregnated prostheses. Furthermore, we found no, significant differences in median BTG, FPA, or TDP levels betweeen patients in the collagenimpregnated prosthesis group and patients in the nonimpregnated prosthesis group (Fig. 1). This suggests that collagen does not stimulate the coagulation cascade any more than conventional Dacron prostheses do. On the contrary, the data suggested that the collagen-impreguated prosthesis group compared favorably with the nonimpregnated prosthesis group in regard to platelet activation and fibrin
62
Journal of VASCULAR SURGERY
De A4ol Van Otterloo et al.
Table II. Distribution of factors that may influence the clotting mechanism in 41 aortoiliac procedures Number of patients Collagen - prosthesis
Collagen + prosthesis
Factors
(n = 21)
(n = 20)
Type of prosthesis Tube graft (no. of) Aortobiiliac Aortoiliofemoral Aortobifemoral Mean volume of blood loss (ml) Mean volume of donor blood transfusions (ml) Mean aortic cross clamping time (min) Superficial thrombophlebitis after operation (no. of patients)
6 9 4 2 2057 1123 45 4
1 9 3 7 1834 1062 51 3
formation (Fig. 1). Median BTG, FPA, and TDP levels were increased in the nonimpregnated prosthesis group at day 0. Moreover, a statistically significant increase (p = 0.04) in median TDP in the nonimpregnated prosthesis group (TDP: 2.3 mg/ml) as compared with the collagen-impregnated prosthesis group (1.4 mg/ml) indicated increased intraoperative fibrin metabolism in the former group. In a comparison of results between patients who were undergoing aortoiliac reconstructions and patients who were undergoing cholecystectomies, the findings indicated a significantly increased platelet activation in the aortoiliac reconstruction group compared with the control group (BTG: 144 IU/ml and 39 IU/ml, p = 0.0006, respectively) (Fig. 2). Furthermore, we fomad that intraoperative fibrin metabolism overall, as reflected by FPA and TDP levels, was significantly higher during aortoiliac reconstruction than during cholecystectomy (FPA: 49 ng/ml vs 21.8 ng/ml, p = 0.014, and TDP: 2.0 mg/ml vs 0.0 mg/ml, p < 0.0001, respectively) (Fig. 2). After operation, both BTG and FPA levels initially fell, so that by day 2, they had reached preoperative levels. However, the results at day 7 and day 10 after operation indicated prolonged platelet activation and fibrin metabolism in patients who had undergone aortoiliac reconstructions. Median BTG plasma levels were increased significantly in the aortoiliac reconstruction group compared with the control group (Fig. 2). Plasma FPA levels also remained elevated until day 10 after vascular operation, whereas in control subjects FPA had dropped to normal by this stage, which suggests a prolonged fibrin formation after aortoiliac surgery (day 10:8.1 ng/ml vs 0.0 ng/ml, respectively, p = 0.03). These findings were paralleled by the median TDP plasma levels. Total degradation product levels increased significantly more in patients in the aortoiliac recon-
struction group than in patients in the control group (Fig. 2). Preoperative data (day - 1 ) demonstrated that patients who had vascular operations had an increased fibrin metabolism compared with control subjects (Fig. 2). We found a statistically significant difference in median TDP level between patients with vascular disease and control subjects (1.0 mg/ml vs 0.0 mg/ml, p = 0.0004) (Fig. 2). In contrast, preoperative BTG and FPA levels in patients who had vascular operations and control subjects were similar. To evaluate the influence of potentially thrombogenic factors on the results in both the collagenimpregnated prosthesis and the nonimpregnated prosthesis groups, the effect of the amount of prosthetic material that was implanted was studied. Aortobiiliac, aortoiliofemoral, and aortobifemoral bifurcation prostheses were evenly distributed among patients with noncollagen and coUagenimpregnated prostheses (Table II). By chance, aortic tube prostheses were significantly more present in the nonimpregnated prosthesis group (Table II). However, the mean blood-prosthesis surface area did not differ significantly between the two groups (nonimpregnated prosthesis: 6775 m m 2 vs collagenimpregnated prosthesis: 7946 mm2; p = 0.14). The volume of blood loss in the two groups was similar (2057 ml vs 1834 ml,p = 0.43, Table II), as was the volume of blood that was administered (1123 ml vs 1062 ml, p = 0.68) (Table II). Another factor was the degree ofischemia, which was represented by the duration of aortic clamping. We found no significant difference in mean clamping time (45 minutes vs 51 minutes,p = 0.24) (Table II). Finally, the last factor that was evaluated was the occurrence of superficial thrombophlebitis, which stimulates BTG release. Again, no difference was observed (4 vs 3, p = not significant, Table II).
Volume 14 Number 1 July 1991
Effects of collagen-coatedDacron grafls on platelets/fibrin 63
NS
NS
p-O.02
N8
NS
NS
p-O.O006
NS
p-O.03
I>-0.03
B'rO (lU/n~)
erro (ILUml)
Is0.
100.
5o .
.
.
.
.
.
.
.
.
.
.
.
o
ct ni
cJ rJ
ct r~
• ~ ni
cl ni
o.
-1
0
2
7
I0
NS
NS
NS
02
v
NS
NS
NS
c
v
-1
day
c
v
c
0
p-0.014
v
¢
2
v
7
NS
c
10
NS
p-0.03
FPA (ngr/mi)
FPA (l~Ftn~)
80-
4o.
¢t ni
O"
el rd
ci ni
ci ni
ci ni
-1
0
2
7
NS
P-O.04
NS
v
o10
v
day
c -1
p-O.O004
NS
vc
vc
0
2
p
