Ann Thorac Surg 1999;68:479-485
© 1999 The Society of Thoracic Surgeons
Original Articles
A sprayable hemostat containing fibrillar collagen, bovine thrombin, and autologous plasma
Jeff J. Prior, PhDa,
Donald G. Wallace, PhDa,
Andrew Harner, MAa,
Noël Powers, BSa
a Cohesion Technologies, Palo Alto, California, USA
Address reprint requests to Dr Prior, Cohesion Technologies, Inc, 2500 Faber Place, Palo Alto, CA 94303
e-mail: jprior{at}cson.com
 |
Abstract
|
|---|
Background. A new sprayable hemostat, CoStasis hemostatic device (Cohesion Technologies Inc, Palo Alto, CA), consisting of collagen, thrombin, and autologous plasma was tested versus fibrin sealant and collagen hemostatic sponges. Performance was also monitored as fibrinogen and platelets were depleted from the plasma. In addition, the strength of the CoStasis gel, its sealing ability, its fibrillar structure, and its platelet-aggregating ability were investigated.
Methods. Hemostatic performance was determined with an in vivo bleeding rabbit kidney and spleen model. Differential scanning calorimetry and electron microscopy were used to analyze collagen structure. Sealing ability was determined with a burst-test apparatus.
Results. In the in vivo model, CoStasis was superior to fibrin sealant and collagen sponges in achieving a rapid time to hemostasis. The formulation continued to perform well when either platelets or fibrinogen was depleted. CoStasis formed weaker gels than fibrin sealant and could withstand only modest pressures. The collagen in the formulation had a fibrillar structure that was shown to aggregate human platelets.
Conclusions. CoStasis, with the two platelet activators collagen and thrombin in addition to the thrombin-catalyzed formation of fibrin and the sealing properties of the soft gel, provides an excellent atraumatic hemostat.
|
Introduction
|
|---|
Recently, a collagen composite tissue adhesive was developed in which fibrillar collagen was added to a source of fibrinogen and thrombin [1]. It was shown that the collagen increased the viscosity of the formulation during application to surgical sites and reinforced the fibrin clot that formed. This work led to the development of the investigational product CoStasis, which is designed for use with autologous plasma (typically containing fibrinogen, 2 to 5 mg/mL) (Cohesion Technologies Inc, Palo Alto, CA) [2]. CoStasis is intended to be used as a sprayable atraumatic hemostat in the treatment of diffuse capillary bleeding. The formulation provides two potent platelet-activating agents, thrombin and collagen. In addition, the thrombin catalyzes the formation of the fibrin clot, and the collagen strengthens the gel that adheres to the bleeding site and provides a seal resistant to blood flow.
In this report, we describe the in vivo hemostatic performance of CoStasis in a bleeding rabbit kidney and spleen model, where the time to hemostasis after application of CoStasis was compared with the time to hemostasis after using Instat collagen sponge, an investigational fibrin sealant, and Tisseel commercial fibrin sealant. In the kidney model, the superior performance of CoStasis is demonstrated. The physical and chemical characteristics of CoStasis are described, and the collagen component’s activity in platelet aggregation is demonstrated.
|
Material and methods
|
|---|
The collagen component of CoStasis was purified from bovine corium; it was solubilized with pepsin and reprecipitated in 20 mmol/L sodium phosphate, pH 7.2 [3]. Bovine topical thrombin was purchased from Jones Medical Industries (St. Louis, MO). CoStasis is composed of fibrillar collagen at 20 mg/mL, 40 mmol/L calcium chloride, and thrombin, 500 NIH units L/mL (component 1) and is mixed with an equal volume of plasma (component 2) during application.
Plasma was obtained for in vivo experiments by centrifuging fresh citrated rabbit blood (0.32% sodium citrate, wt/vol) at 1,380 g for 2 minutes to yield platelet-rich plasma. Rabbit plasma contains approximately 2 mg of fibrinogen per milliliter [4]. For in vivo experiments testing depleted plasma, platelet-rich plasma was spun at 1,380 g for 45 minutes to yield platelet-poor plasma. Platelet depletion was verified by hemocytometer count to give a reduction of 98.4% in the platelet-poor plasma. The two preparations were mixed 1:1 to yield platelets at 50% of normal levels. Fibrinogen-depleted plasma was made as follows: platelet-poor plasma was made as just described, and the platelet pellet was saved to be mixed with the plasma after removal of the fibrinogen. Fibrinogen was precipitated out of the plasma by heating to 53° to 56°C for 3 minutes as previously described [5]. After the fibrinogen was pelletted, the fibrinogen-depleted supernatant was mixed with platelets as needed. The fibrinogen-depleted plasma was mixed with normal plasma to yield plasma at 50% of normal levels. Fibrinogen levels were determined using a BBL fibrometer (Becton Dickinson, Cockeysville, MD) by the method described in procedure No. 880, Sigma Diagnostics (St. Louis, MO). Fibrinogen levels were 300 to 330 mg/dL in normal rabbit plasma, 150 mg/dL in the 50%-depleted plasma, and not measurable in the fibrinogen-depleted plasma.
The investigational fibrin sealant was prepared from reagents supplied by Haemacure (Kirkland, PQ, Canada). Tisseel was from Immuno, AG (Vienna, Austria). Human fibrinogen was used at concentrations of 60 or 125 mg/mL. Clotting of the fibrinogen was achieved by rapid mixing with an equal volume of solution containing either 160 U/mL of human thrombin or 500 U/mL of bovine thrombin in 40 mmol/L CaCl2. Instat collagen sponge was from Johnson & Johnson (Arlington, TX). Fibrin sealant and CoStasis were applied using a two-syringe system, which allowed the thrombin-containing syringe and the fibrinogen-containing syringe to be depressed simultaneously and the fluids mixed through a Y connector and sprayed on the bleeding site.
In vivo rabbit hemostasis model
Adult New Zealand White rabbits were anesthetized with an intramuscular injection of ketamine hydrochloride (35 mg/kg) and xylazine hydrochloride (3 mg/kg). Anesthesia was maintained during the procedure by inhalation of isoflurane and oxygen. The abdomen was shaved, and a midline laparotomy was performed to expose the viscera. An incision 2 mm deep by 15 mm long was made in the kidney and in the spleen. Gauze was used to lightly wipe up the blood for approximately 3 to 5 seconds immediately after the incision was made and was then removed just prior to application of the test material. The light pressure on the gauze was applied by the same individual in the same manner for all treatments. The time interval from application of material until hemostasis was reached was determined. Animals received humane care in compliance with the "Guide for the Care and Use of Laboratory Animals" published by the National Institutes of Health (NIH publication 85-23, revised 1985).
Statistical analyses of mean times to hemostasis were performed using Wilcoxon sign rank tests of time-to-event data in a Kaplan-Meier analysis to determine p values. Observations of time to hemostasis exceeding 10 minutes were censored and included as 10 minutes for the purpose of these analyses. All analyses were performed using JMP version 3.0 (SAS Institute).
Electron microscopy
For electron microscopy, component 1 of CoStasis (collagen plus thrombin) was diluted 1:100, 1:1000, and 1:10000 on ice. Drops of sample were placed on Parafilm in a Petri dish and kept on ice. Butvar/carbon–coated grids were floated on top of the drops of sample for 15 minutes. Grids were wicked with filter paper, allowed to dry, and then placed sample side down on beads of 1% phosphotungstic acid (pH 7.0) for 10 minutes. Grids were wicked dry and viewed on a Philips CM10 transmission electron microscope operating at 80 kV.
Differential scanning calorimetry
Differential scanning calorimetry was performed on a Mettler DSC 20 and TC10 TA processor (Hightstown, NJ). Samples containing 0.1 to 0.3 mg of collagen in approximately 12 to 15 µL were sealed in aluminum pans and heated at 10°C per minute. A reference pan containing only buffer was prepared and heated at the same time in the instrument.
Rheometry
Rheometry was performed in a Rheometrics fluids spectrometer (model 8400; Piscataway, NJ) using parallel- plate geometry. CoStasis was prepared as already described except the plasma component was acidified to pH 5.0, which prevented gelation and allowed time for degassing and loading the sample. The sample was then submerged in neutral phosphate buffer, and the time course of gelation was monitored. The dynamic elastic modulus, or G', was measured at an oscillation rate of 1 radian/s and a strain of 1% [6]. The gelled sample was then exposed to a frequency sweep (oscillation rate, 0.1 to 100 radian/s) and a strain sweep (strain, 1% to 50%). Two control formulations were also prepared; one, called thrombin-plasma, had the collagen omitted from the composition just described, and the other, called collagen-plasma, contained collagen and plasma but no thrombin or calcium. Finally, the rheometry of a fibrin sealant was examined. In this case, gelation was delayed by using a low concentration of thrombin (6 U/mL).
Platelet aggregation
Platelet aggregation was performed with a Sienco dual sample aggregation meter (Morrison, CO). Platelet-rich plasma was isolated by centrifuging blood from human donors for 2 minutes at 1,380 g. The platelet-rich plasma was mixed with fibrillar collagen, and the time required to reach one half of the maximum aggregation was determined for each sample and used as a measure of platelet-aggregating ability. Statistical analysis of the time required to reach half of the maximum aggregation was by analysis of variance using a Tukey-Kramer analysis of all data pairs.
Burst tests
Burst tests were performed on a custom apparatus that consisted of a pressure gauge (model PG5000; PSI-Tronix, Tulare, CA) connected to a circular sample plate with a central orifice 2 mm in diameter. To simulate a tissue surface, the sample plate had a circular sheet of coarse-fibered collagen, prepared from pulverized cow hide, and a matching central hole, fastened to it by a gasket seal. After the collagen sheet was dampened with saline solution, the test formulation was sprayed onto the surface to seal the hole. Pressurized saline solution was then applied by a syringe pump (Sage Instruments model 355; Orion Research, Cambridge, MA) at a flow rate of 5 mL/min, and the pressure at which saline solution penetrated the test formulation was recorded.
|
Results
|
|---|
Hemostatic performance of CoStasis hemostatic device
CoStasis was compared with an investigational fibrin sealant with 60 mg/mL of fibrinogen and 160 U/mL of thrombin, Tisseel commercial fibrin sealant with 125 mg/mL of fibrinogen and 500 U/mL of thrombin, and Instat collagen sponge using a bleeding rabbit kidney and spleen model. As seen in Table 1, CoStasis had the fastest mean and median times to hemostasis in the kidney of the four materials tested and was the only material to achieve complete hemostasis in 100% of the sites in 2 minutes or less. CoStasis was found to be significantly faster (p 
0.05, Wilcoxon) than the other materials. In the spleen, CoStasis again was found to have significantly shorter bleeding times (p 0.05, Wilcoxon) than the other formulations (Table 2). The fibrin sealant, which contained lower levels of fibrinogen and thrombin, achieved hemostasis in less than half of the sites treated, and therefore a median time to hemostasis could not be calculated. The mean was calculated using a value of 10 minutes for the data exceeding the 10-minute cutoff time.
The CoStasis hemostatic device mixes an equal volume of the collagen and thrombin component with the patient’s plasma, which provides a source of fibrinogen and platelets; it was therefore important to analyze performance under conditions where fibrinogen and platelet concentrations might be lower than normal. Experiments were designed to test the efficacy of CoStasis when platelets were reduced to 50% of normal and when they were completely depleted. In a separate experimental series, platelet levels were normal, but fibrinogen was reduced to 50% of normal and was completely depleted. A final experimental series explored the effect of depleting both fibrinogen and platelets.
On kidney sites when platelets were depleted by 50%, the hemostatic performance was equivalent to that using plasma with normal platelet levels (Table 3, experiments 1 and 2). Although not significantly different, the hemostatic performance using plasma with 100%-depleted platelets was slower than that using normal plasma (see Table 3, experiment 3). Both the 50%-depleted and completely platelet-depleted plasmas achieved hemostasis in all the kidney sites in less than 2 minutes, a significantly greater proportion than with Instat (p 0.04, Wilcoxon).
View this table:
[in this window]
[in a new window]
|
Table 3. Hemostatic Performance of CoStasis Hemostatic Device in Kidney Sites Under Conditions of Deficient Plasmaa
|
|
Formulations containing normal platelet levels but with fibrinogen at 50% of normal or totally depleted had mean times to hemostasis equivalent to formulations with normal fibrinogen levels and achieved hemostasis in all kidney sites in less than 2 minutes (see Table 3, experiments 4 and 5). Both formulations had a hemostatic performance significantly faster than that of Instat (p < 0.01), and the proportion of sites achieving hemostasis in 2 minutes was significantly greater as well (p < 0.04, Wilcoxon).
When both platelets and fibrinogen were depleted, the mean time to hemostasis was significantly slower than when normal plasma was used (p = 0.001, Wilcoxon) and not significantly different from the mean time with Instat (see Table 3, experiment 6). The depleted plasma achieved hemostasis in only 30% of kidney sites versus 100% with normal plasma. When platelets were completely depleted but fibrinogen was only 50% depleted, the time to hemostasis was still significantly slower than normal (p < 0.04) and not significantly different from that of Instat (see Table 3, experiment 7). In this case, hemostasis was achieved in 70% of the sites, intermediate between the completely depleted and normal values.
Material consisting of the collagen and thrombin component alone, without plasma, was also tested (see Table 3, experiment 8). This material had a hemostatic performance significantly slower than that of the normal formulation (p < 0.0005, Wilcoxon) but not significantly different from that of Instat.
The data for hemostatic performance in the spleen are presented in Table 4. Because these data were highly variable, none of the values were significantly different. However, depletion of platelets and fibrinogen generally correlated with a decrease in the percentage of sites achieving hemostasis in 10 minutes or less and an increase in mean time to hemostasis. The poorest-performing materials in the spleen were the formulation with complete depletion of platelets and the formulation with no plasma component. In both of these cases, only 20% of the sites achieved hemostasis in less than or equal to 10 minutes, and the average time to hemostasis increased to greater than 500 seconds. The variability of the spleen data is also apparent when it is noted that the increase in mean time to hemostasis for CoStasis made with normal plasma in this experiment is longer than the mean time for CoStasis and Instat from the previous experiment (see Table 2).
View this table:
[in this window]
[in a new window]
|
Table 4. Hemostatic Performance of CoStasis Hemostatic Device in Spleen Sites Under Conditions of Depleted Plasmaa
|
|
Collagen structural analysis
Because the fibrillar nature of collagen is required for platelet activation [7], the structure of the collagen in CoStasis was analyzed. The melting profile of the collagen, as obtained by differential scanning calorimetry, was characteristic of a mixture of fibrils of varying diameters. The main transition appeared near 50°C and was tentatively identified as being caused by intermediate-sized fibrils, ranging from 50 to 300 nm in diameter. A distinct shoulder was present at about 44°C and probably represents so-called microfibrils, ranging from 3 to about 20 nm in diameter [8]. Samples were examined by electron microscopy, and the diameters and identities of the collagen fibril assignments were in agreement with the assignments just presented (Fig 1). When CoStasis was tested in the animal model with denatured collagen replacing the fibrillar collagen, inferior performance resulted, further indicating the benefit of fibrillar collagen (data not shown).
View larger version (175K):
[in this window]
[in a new window]
|
Fig 1. Electron microscopy of collagen component. Sample was prepared as described in Material and Methods section. The sample contained many small fibrils 3 to 10 nm in diameter as well as larger fibrils, banded and nonbanded. The large fibril shown here has a maximum diameter of about 80 nm. (x72,000 before 56% reduction.)
|
|
Analysis of platelet-aggregating ability of collagen
The collagen in CoStasis was tested in the absence of thrombin to determine if it was able to aggregate human platelets. The collagen suspension was found to aggregate platelets at concentrations of 0.1, 0.05, and 0.01 mg/mL, with times necessary to reach one half of the maximum aggregation of 2.2, 2.8, and 4.6 minutes, respectively. The aggregation was significantly slower at 0.01 mg/mL (p < 0.006, t test). Formulations that appeared microfibrillar by differential scanning calorimetry (as a result of adjusting salt concentrations) also were capable of aggregating platelets, but controls containing only calcium and sodium chloride were inactive.
Because thrombin is a strong platelet-aggregating agent [9], it was used as a positive control to verify that the collagen was fully aggregating platelets. At 0.2 U/mL (sufficient to activate platelets, binding both high and moderate affinity receptors [10]), the degree of platelet aggregation was similar to that of the CoStasis collagen.
Physical characterization of CoStasis gel
Rheometry was used to follow the gelling response and the strength (as determined by the G' [11]) of the final gel at 25°C. The strengths of a fibrin sealant, CoStasis, a thrombin-plasma gel (lacking collagen), and a collagen-plasma gel (without thrombin or calcium) were compared. As seen in Table 5, CoStasis formed a stronger gel than the thrombin-plasma gel and the collagen-plasma gel; the CoStasis gel and the thrombin-plasma gel were resistant to strain. Fibrin sealant, with much higher fibrinogen levels, formed gels three orders of magnitude higher in G' compared with CoStasis and also resistant to strain.
View this table:
[in this window]
[in a new window]
|
Table 5. Rheometry of CoStasis, Fibrin Sealant, Collagen-Plasma, and Thrombin-Plasma (Elastic Modulus dynes/cm2)a,b
|
|
Table 6 compares the ability of various sealant formulas to seal a fluid leak with use of a burst-test apparatus. CoStasis (collagen-thrombin-plasma) and thrombin-plasma were comparable; both failed at relatively low pressures. Fibrin sealant failed at about fivefold higher pressures.
|
Comment
|
|---|
Earlier workers investigated the effect of combining collagen with a fibrin sealant and found an improvement in hemostatic ability. Stark and de Leval [12] applied fibrin sealant to a collagen fleece and observed that for hemostasis in operations on children with complex heart defects, fibrin sealant (Tisseel)–collagen fleece composite was the best technique in most cases. Jakob and colleagues [13] put fibrin sealant–soaked porcine collagen on complex liver injuries in an anticoagulated rat model and concluded that fibrin sealant in combination with porcine collagen was superior to the use of either agent alone in regard to continuous parenchymal bleeding. Finally, Sakon and associates [14] developed a reliable procedure for intraoperative hemostasis and prevention of postoperative bleeding in the resection of cirrhotic livers of patients with hepatocellular carcinoma. The procedure used a combination of microcrystalline collagen powder and fibrinogen tissue adhesive.
Here we used a formulation containing fibrillar collagen in suspension with thrombin and calcium chloride and mixed with the patient’s platelet-rich plasma. CoStasis was compared with Tisseel fibrin sealant, currently used in Europe for hemostasis, with Instat collagen sponge, used in the United States as a hemostat, and with an investigational fibrin sealant. The bleeding kidney model data demonstrated that CoStasis achieved hemostasis significantly faster than fibrin sealant and Instat; the spleen data were more variable and therefore inconclusive. CoStasis was the only material to achieve hemostasis in all of the kidneys in less than 2 minutes, our goal for a fast-acting hemostat.
The use of plasma as the fibrinogen component of a fibrin sealant, or fibrin gel, has been investigated by previous workers who compared it with fibrin glue from cryoprecipitate. Oz and colleagues [15] made a fibrin gel from platelet-rich plasma and 500 U/mL of thrombin and in a bleeding rabbit liver and spleen model, found no significant difference in hemostasis between it and cryoprecipitate-prepared fibrin glue. Hartman and co-workers [16] used the patient’s plasma to make fibrin gel at operation and reported that it provided hemostasis that was at least as good as that of heterologous plasma glue in 40 cardiac surgical patients; postoperative chest tube outputs were similar, with a trend toward less output when the patient’s own plasma was used.
Because the fibrinogen is from the patient’s own plasma (prepared with a sterile plasma collection system, supplied by the manufacturer, in about 5 minutes), CoStasis addresses the risk of hepatitis, acquired immunodeficiency syndrome, and other transmissible disease agents in blood. CoStasis provides a superior hemostat without the risks associated with fibrin sealants derived from pooled human blood (although these risks are reduced with Tisseel which is treated to inactivate viruses).
CoStasis was found to remain effective when either platelets or fibrinogen were depleted by 50%. Even at 100% depletion of either component, hemostasis in each kidney tested was achieved in less than 2 minutes. It was only in experiments where both fibrinogen and platelets were depleted that significantly slower times to hemostasis were observed, although not significantly different from those of the Instat collagen sponge. The data suggest that, compared with currently available hemostats, CoStasis should perform well in patients with low levels of fibrinogen or platelets.
When the collagen in CoStasis was tested for platelet aggregation, the activity did not decrease until the collagen concentration was less than 0.05 mg/mL. The collagen in the CoStasis product is 10 mg/mL when it is delivered to the bleeding site, thus assuring the presence of ample collagen to aggregate platelets. CoStasis also contains 250 U/mL of thrombin, one of the strongest platelet aggregators [10]; thus, the CoStasis formulation benefits from two powerful platelet aggregators that, at the bleeding site, causes the formation of a platelet a plug, which releases coagulation factors, and provides a procoagulant surface.
CoStasis formed a gel with a higher G' than corresponding gels made when collagen was absent (thrombin-plasma). Further, the CoStasis G' was greater than the sum of the moduli of the collagen gel alone and the thrombin-plasma gel. The latter observation suggests that collagen and fibrin networks have a synergistic effect on each other, as noted previously [1]. The fact that fibrin sealant exhibited a much higher G' value implies that a hard gel is not necessarily better. This has been observed by Basu and coworkers [17], who found that fibrin sealant made from cryoprecipitate was as effective in controlling bleeding from aortic and atrial suture lines and from the epicardium as a commercial fibrin sealant (130 mg/mL of protein concentrate) in a canine model.
Burst data demonstrated that CoStasis and plasma gels had lower burst strengths than fibrin sealant, which is consistent with the other results described. In addition, the burst pressure of CoStasis was lower than capillary blood pressure (25 mm Hg) [18]. This finding implies that the mechanism of hemostatic action by CoStasis is not simply to seal over a wound; rather, the collagen-strengthened gel adheres to the bleeding site and thus provides some tamponade hemostatic effect; as blood migrates into the gel, thrombin is present to catalyze fibrin clot formation, and collagen and thrombin initiate platelet activation and aggregation.
|
Acknowledgments
|
|---|
We thank Vivek Shenoy and Dr Joel Rosenblatt, Collagen Corporation, Palo Alto, CA, for advice on rheometry; Prof Klaus Kühn, Max Planck Institut, Munich, Germany, and Dr David Birk, Boston, MA, for electron microscopy; and Dr Gregory M. Cruise, Cohesion Technologies, for some burst-test measurements. We also thank David Sierra and Jim Duronio at Cohesion Technologies for design of the burst test, and Dr Louis C. Sehl and David Sierra for extensive advice on coagulation proteins and sealants.
|
Footnotes
|
|---|
Drs Jeff J. Prior and Donald G. Wallace, and Ms Noël Powers, are employed by Cohesion Technologies Inc, which has a financial interest in materials discussed in this article. Mr Andrew Harner was employed by Cohesion Technologies at the time the work was performed.
|
References
|
|---|
-
Sierra D.H. Fibrin-collagen composite tissue adhesive. In: Sierra D.H., Saltz R., eds. Surgical adhesives and sealants. Lancaster, PA: Technomic Publishing Co, 1996:29-39.
-
Esnouf M.P. The thrombin-fibrinogen reaction and the stabilization of fibrin. In: Biggs R., ed. Human blood coagulation, haemostasis and thrombosis, 2nd ed. Oxford, Eng: Blackwell Scientific, 1976:32-41.
-
McPherson J.M., Wallace D.G., Sawamura S.J., et al. Collagen fibrillogenesis in vitro. Coll Relat Res 1985;5:119-135.[Medline]
-
Lewis J.H. Comparative hemostasis in vertebrates. New York: Plenum, 1996:186.
-
Lorand L., Credo R.B., Janus T.J. Factor XIII (fibrin-stabilizing factor). Methods Enzymol 1981;80:333-341.
-
Rosenblatt J., Devereux B., Wallace D.G. Dynamic rheological studies of hydrophobic interactions in injectable collagen biomaterials. J Appl Polymer Sci 1993;50:953-963.
-
Barnes M.J., Gordon J.L., MacIntyre D.E. Platelet-aggregating activity of type I and type III collagens from human aorta and chicken skin. Biochem J 1976;160:647-651.[Medline]
-
Wallace D.G., Condell R.A., Donovan J.W., Paivinen A., Rhee W.M., Wade S.B. Multiple denaturational transitions in fibrillar collagen. Biopolymers 1986;25:1875-1893.[Medline]
-
Watson S.P., Authi K.S. Platelets, a practical approach. Oxford, UK: IRL Press, 1996:319.
-
Colman R.W., Cook J.J., Niewiarowski S. Mechanisms of platelet aggregation. In: Colman R.W., Hirsh J., Marder V.J., Salzman E.W., eds. Hemostasis and thrombosis, 3rd ed. Philadelphia: JB Lippincott, 1994:508-523.
-
Ferry J.D. Viscoelastic properties of polymers, 3rd ed. New York: John Wiley, 1980:1-29.
-
Stark J., de Leval M. Experience with fibrin seal (Tisseel) in operations for congenital heart defects. Ann Thorac Surg 1984;38:411-413.[Abstract]
-
Jakob H., Campbell C.D., Stemberger A., et al. Combined application of heterologous collagen and fibrin sealant for liver injuries. J Surg Res 1984;36:571-577.[Medline]
-
Sakon M., Monden M., Gotoh M., et al. Use of microcrystalline collagen powder and fibrinogen tissue adhesive for hemostasis and prevention of rebleeding in patients with hepatocellular carcinoma associated with cirrhosis of the liver. Surg Gynecol Obstet 1989;168:453-454.[Medline]
-
Oz M.C., Jeevanandam V., Smith C.R., et al. Autologous fibrin glue from intraoperatively collected platelet-rich plasma. Ann Thorac Surg 1992;53:530-531.[Abstract]
-
Hartman A.R., Galanakis D.K., Honig M.P., Seifert F.C., Anagnostopoulos C.E. Autologous whole plasma fibrin gel. Intraoperative procurement. Arch Surg 1992;127:357-359.[Abstract/Free Full Text]
-
Basu S., Marini C.P., Bauman F.G., et al. Comparative study of biological glues. Ann Thorac Surg 1995;60:1255-1262.[Abstract/Free Full Text]
-
Price S.A., Wilson L.M. Pathophysiology, 2nd ed. New York: McGraw-Hill, 1982:317.
Accepted for publication March 4, 1999.
This article has been cited by other articles:
|
|
|
|
 
J. G. Schoenecker, R. K. Johnson, A. P. Lesher, J. D. Day, S. D. Love, M. R. Hoffman, T. L. Ortel, W. Parker, and J. H. Lawson
Exposure of Mice to Topical Bovine Thrombin Induces Systemic Autoimmunity
Am. J. Pathol.,
November 1, 2001;
159(5):
1957 - 1969.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
