Prosthesis and method for making same

Abstract

A tubular collagen vascular graft prosthesis is subjected to treatment so that its intimal surface has an increased net negative surface charge which prevents thrombosis. Treatment with succinic anhydride is shown as one way of accomplishing this. Valve grafts are likewise treated.

Description

BACKGROUND

It is known to prepare vascular graft prostheses from collagen materials. Such prostheses are in fact available commercially from Johnson & Johnson of New Brunswick, New Jersey.

Aside from the publications and articles to be mentioned hereafter, pertinent material will be found in the following:

Wesolowski, S. A.: The healing of vascular prostheses. Surgery 57:319, 1965.

Wesolowski, S. A., Fries, C. C., Domingo, R. T., Fox L. M., and Sawyer, P. N.: Fate of simple and compound arterial prostheses: Experimental and human observations. In: Fundamentals of Vascular Grafting. McGraw-Hill, New York, 1963, pp. 252-268.

Wesolowski, S. A., Hennigar, G. R., Fox, L. M., Fries, C. C., and Sauvage, L. R.: Factors contributing to long term failure in human vascular prosthetic grafts. Presented at Symposium on Late Results of Arterial Reconstruction. International Cardiovascular Society Meeting, Rome, September 1963. J. Cardiov. Surg. 5:44 1964.

Sawyer, P. N., and Pate, J. W.: A study of electrical potential differences across the normal aorta and aortic grafts of dogs. Research Report NM 007081, 10.06. Naval Medical Research Institute, Bethesda, Md., 1953.

Sawyer, P. N., and Pate, J. W.: Bioelectric phenomena as etiologic factors in intravascular thrombosis. Amer. J. Physiol. 175:103, 1953.

Williams, R. D., and Carey, L. C.: Studies in the production of standard venous thrombosis. Ann. Surg. 149:381, 1959.

Schwartz, S. I., and Robinson, J. W.: Prevention of thrombosis with the use of a negative electric current. Surg. Forum 12:46, 1961.

Sadd, J. R., Koepke, D. E., Daggett, R. L., Zarnsdorff, W. C., Young, W. P., and Gott, V. L.: Relative ability of different conductive surfaces to repel clot formation on intravascular prostheses. Surg. Forum 12:252, 1961.

Means, G. E., and Feeney, R. E.: Chemical Modification of Proteins. Holden-Day, Inc., San Francisco, Calif. 1971, pp. 144-148.

SUMMARY OF INVENTION

It is an object of the invention to provide an improved method for preparing a collagen vascular graft prosthesis.

Still another object of the invention is to provide an improved vascular graft prosthetic device which is less susceptible to thrombosis and the like.

Still another object of the invention is to provide an improved tubular vascular graft, the intimal surface of which cannot be recognized by blood platelets for purposes of accumulation.

The above and other objects of the invention are achieved by the provision of a method in accordance with the invention which comprises modifying the intimal surface of a vascular graft prosthesis by increasing the net negative surface charge of the intimal surface thereof.

Preferably the prosthesis will be of collagen and the intimal surface thereof will be treated by a succinylation reaction in order that the free amino groups of the collagen surface be covered.

Preferably the starting material of the graft will be a collagen prepared from a ficin digested bovine carotid artery material also known as a dialdehyde starch tanned collagen.

As will be shown, the graft is generally provided in the form of a collagen tube one end of which is closed and inserted into a fluid such as ethanol with a liquid chemical reactant being inserted into the lumen of the tube to increase the negative surface charge of the intimal surface thereof. The reactant may preferably be, as will be shown hereinafter, succinic anhydride in a basic solution.

More particularly a NaHCO.sub.3 buffer solution is inserted into the lumen in sequential additions of approximately 10 ml. aliquots to which about 0.1 gm. of crystals of succinic anhydride are respectively added.

The above and other objects and features of the invention will become apparent from the detailed description which follows hereinbelow.

Before, however, the detailed description of the invention is presented, there is next given a further brief summary as to how the study leading to the instant invention was undertaken.

Collagen vascular prostheses prepared by enzymatically digesting bovine carotid arteries with ficin was obtained. Segments of these vascular prostheses were modified chemically to alter their intimal surface electrochemical configuration and, more particularly, for producing a net positive surface charge, negative surface charge, or neutral surface charge. These modified vascular grafts, together with segments of the unmodified collagen vascular prostheses were implanted into the carotid arteries, external jugular veins, femoral arteries and femoral veins of mongrel dogs for periods of 1 minute, 15 minutes, 30 minutes and 2 hours. The grafts were then removed and examined macroscopically for the presence and degree of thrombotic occlusion. These same prostheses were further analyzed using a scanning electron microscope in an attempt to determine the precise nature of the occluding thrombi. It was found that by increasing the net negative surface charge density on the intimal surface of these vascular prostheses, thrombotic occlusion could be substantially averted, while creating a more positive prostheses-blood interface potential greatly accelerated the thrombotic events.

Segments of these vascular protheses were again implanted into the vascular system of mongrel dogs and in vivo arterial and venous streaming potential measurements were carried out. The polarity of the streaming potential reflects the interfacial potential between the vascular prosthesis and the blood flowing through the graft with a positive streaming potential at the downstream electrode indicating a net negative surface charge density on the intimal surface of the graft with respect to the blood elements. The arterial and venous in vivo streaming potentials obtained from the "negatively" altered collagen vascular prostheses were uniformly positive in polarity, while those streaming potentials obtained from the "positively" modified grafts were uniformly negative. The streaming potentials obtained from those grafts altered chemically to neutralize the intimal surface charges at physiological pH were also found to be positive in polarity but of a lesser magnitude than the negatively altered grafts. The unmodified ficin digested bovine carotid arteries had streaming potentials of a slightly negative polarity. These streaming potential measurements served to check the chemical modification procedures and assure that the electrical alteration of surface charge density occurred as predicted.

BRIEF DESCRIPTION OF DRAWING

FIG. 1 diagrammatically illustrates apparatus for practicing the invention;

FIG. 2 diagrammatically illustrates a technique for measuring streaming potential in accordance with the invention; and

FIG. 3 diagrammatically illustrates a portion of FIG. 2 on enlarged scale.

DETAILED DESCRIPTION

Evidence has been presented implying the formation of a complex between incomplete carbohydrate chains in collagen and glucosyltransferase present on the outer surface of platelets as the primary step resulting in platelet recognition of and subsequent adhesion to collagen. (Jamieson, G. A., Urban, L. C., and Barber, A. J.: Enzymatic basis for platelet: collagen adhesion as the primary step in haemostasis. Nature New Biology 234:5 (1971): Barber, A. J. and Jamieson, G. A.: Platelet collagen adhesion characterization of collagen glucosyltransferase of plasma membanes of human blood platelets. Biochim. Biophys. Acta 252:533 (1971)). The reaction involves the enzymatic coupling of glucose to galactosyl residues attached to hydroxylysine side chains incorporated into the collagen peptides. The glucose is supplied by platelets as uridinediphosphoglucose (UDPG) according to the following: ##SPC1##

Platelet aggregation activity of collagen in vitro has been shown to require the structural integrity of the 6-hydroxymethyl group of galactose and the .epsilon.-amino groups of lysine and hydroxylysine on the collagen molecule. It has been postulated that the substrate specificity of the platelet glucosyltransferase is responsible for these observed platelet aggregation dependent phenomena (Chesney, C., Harper, E., and Coleman, R. W.: Critical role of the carbohydrate side chains of collagen in platelet aggregation. HIEG 72-43 (1972); Wilner, G. D., Nossel, H. L., and LeRoy, E. C.: Aggregation of platelets by collagen. J. Clin. Invest. 47:2616 (1968) ).

The use of modified bovine arterial heterografts as vascular prostheses has previously been reported and their physical and behavioral characteristics as arterial grafts have been discussed in detail. In the present studies it was found that increasing the electronegativity of the inner surface of the prosthesis significantly contributed to a favorable graft performance.

Some of the above is discussed in the following:

Bothwell, J. W., Lord, G. H., Rosenberg, N., Burrowes, C. B., Wesolowski, S. A., and Sawyer, P. N.: Modified arterial heterografts: relationship of processing techniques to interface characteristics. In: Biophysical Mechanisms in Vascular Homeostasis and Intravascular Thrombosis, P. N. Sawyer, Ed. Appleton-Century-Crofts, New York, 1965, pp. 306-313; Rosenberg, N., Henderson, J., Douglas, J. F., Lord, G. H., and Gaughran, E. R. L.: Use of arterial implants prepared by enzymatic modification of arterial heterografts. II. Physical properties of the elastica and collagen components of the arterial wall. Arch. Surg. 74:89 (1957); Rosenberg, N., Henderson, J., Lord, G. H., and Bothwell, J. W.: Use of enzyme treated heterografts as segmental arterial substitutes. V. Influence of processing factors on strength and invasion by host. Arch. Surg. 85:192 (1962); Rosenberg, N., Henderson, J., Lord, G. H., and Bothwell, J. W.: An arterial prosthesis of heterologous vascular origin. JAMA 187:741 (1964); Rosenberg, N., Henderson, J., Lord, G. H., and Bothwell, J. W.: Collagen arterial prosthesis of heterologous vascular origin: physical properties and behavior as an arterial graft. In: Biophysical Mechanisms in Vascular Homeostasis and Intravascular Thrombosis, P. N. Sawyer, Ed. Appleton-Century-Crofts, New York, 1965, pp. 314-321.

Succinylation of solubilized collagen has been demonstrated to result in an approximately 95% conversion of .epsilon.-amino groups to free carboxyl groups (Gustavson, K. H.: Akiv. For Kemi 17: 541 (1961)). Furthermore, it has been shown that blockage of the free amino groups of collagen by deamination, N-acetylation or treatment with dinitroflurobenzene results in a greater than 90% reduction in platelet aggregating activity in vitro.

Investigation has now been conducted in order to elucidate the antiplatelet and antithrombogenic activity of chemically modified collagen by the in vivo evaluation of ficin digested bovine arterial heterograft vascular prostheses implanted into the vascular system of mongrel dogs.

For the chemical modification of ficin digested bovine carotid arteries, the apparatus devised for the chemical modification of ficin digested bovine carotid arteries is as appears in FIG. 1. It includes a receptacle 10 having an open mouth 12 through which extends the shaft 14 of an electric stirrer 16. A glass stopcock 18 is arranged in the lower end of a ficin digested bovine carotid artery 20 which is immersed, for example, in 40% ethanol.

In order to increase the net positive surface charge density on the intimal surface of the collagen vascular prostheses a carbodiimide-promoted amide formation reaction was employed to convert the major source of net negative charges (free carboxyl groups of aspartic and glutamic acid) to amide moieties. The procedure involves treatment of the protein with excess EDC (1-ethyl-3,3' -dimethylaminopro pylcarbodiimide HCl) and an amine (NH.sub.4 Cl) in the presence of a high concentration of a denaturant (urea). This reaction (Means, G.E., and Feeney, R.E.: Chemical Modification of Proteins. Holden-Day, Inc., San Francisco, Calif., 1971, pp. 144-148) has been used for the near quantitative conversion of protein carboxyl groups to amides for determining numbers of carboxyl groups in proteins (Hoare, D. G. and Koshland, D. E.: J. Biol. Chem. 242:2447 (1967)). The chemical reaction proceeds at room temperature as follows: ##EQU1## wherein:

100 ml of reaction solution at room temperature containing 0.5 M EDC (7.75 gm/100 ml), 7.5 M urea (45 gm/100 ml), and 5.0 M NH.sub.4 Cl (26.75 gm/100 ml) in triple distilled water were prepared (adjusted to pH5) and added into the lumen of the vascular prosthesis in 10 ml aliquots with the implant suspended in 40% ethanol as shown in FIG. 1. The reaction solution was changed every hour with the last aliquot remaining within the lumen overnight. At the conclusion of the 24 hour reaction period the vascular prosthesis was removed from the modification apparatus and placed in a pan of sterile saline where the long tubes were cut into segments of 3.5 cm for implantation studies or 10.0cm lengths for streaming potential measurements.

In order to increase the net negative surface charge density on the intimal surface of the collagen vascular prostheses acylation of the free amino groups (especially those of lysine and hydroxylysine) of the protein with succinic anhydride in a solution more basic than a PH of 7.0 was carried out. This reaction not only covers the major source of free positive charges, but also converts them into anionic residues (Means, G. E., and Feeney, R. E.: Chemical Modification of Proteins. Holden-Day, Inc., San Francisco, Calif., 1971, pp. 74,75). The chemical reaction proceeds as follows: ##SPC2##

A basic solution of 75 ml of 100% ethanol and 25 ml of 1 M NaHCO.sub.3 buffer solution was prepared and added into the lumen of the vascular prosthesis in 10 ml aliquots over a 5 hour period with the implant suspended in 40% ethanol as shown in FIG. 1. To each of these 10 ml aliquots of basic solution were added crystals of succinic anhydride. 1.0 gm of succinic anhydride was equally divided among the 10 aliquots. At the conclusion of the reaction period, the vascular prosthesis was removed from the modification apparatus and placed in a pan of sterile saline where the long tubes were cut into segments of 3.5 cm for implantation studies or 10.0 cm lengths for streaming potential measurements.

In order to neutralize the electrical surface charges on the intimal surface of the collagen vascular prostheses, a carbodiimide promoted internal amide formation reaction was employed to link together the free carboxyl and amino groups of adjacent peptide chains. Such a reaction removed the major sources of electronegativity (carboxyl groups) and electropositivity (free amino groups). The chemical reaction proceeds as follows: ##EQU3##

A 100 ml solution of 0.5 M EDC (7.75 gm/100 ml) in triple distilled water was prepared and added into the lumen of the vascular prosthesis in 10 ml aliquots with the implant suspended in 40% ethanol as shown in FIG. 1. The reaction solution was changed every hour with the last aliquot remaining within the lumen overnight. At the conclusion of the 24 hour reaction period, the vascular prosthesis was removed from the modification apparatus and placed in a pan of sterile saline where the long tubes were cut into segments of 3.5 cm for implantation studies or 10.0 cm lengths for streaming potential measurements.

As a starting material in the above, there was employed, for example, dialdehyde starch tanned collagen vascular prostheses including dialdehyde starch tanned bovine collagen vascular heterografts which were commercially available and obtained from Johnson & Johnson, New Brunswick, New Jersey.

Surgical implantation of modified collagen vascular prostheses involved carotid artery and jugular vein implantation. Mongrel dogs weighing between 10 and 28 Kg (ave. 18.6 Kg) were anesthetized with Nembutal (Sodium Pentabarbital), 0.5 cc/Kg, by intravenous injection. The animals were placed on the operating table in the supine position and an oblique cervical incision, paralleling the anterior border of the sternocleidomastoid muscle was made. The platysma muscle layer was incised along with the anterior layer of the deep cervical fascia and after achieving hemostasis these structures were retracted posteriorly to reveal the external jugular vein. At the angle of the mandible, the posterior auricular vein, the common facial vein and the retromandibular vein were identified and a mobilized segment of each was encircled with a loose ligature of umbilical tape. The posterior external jugular vein was ligated with 3-0 silk and divided near its entrance. The external jugular vein was then dissected free of its fascial attachments down to where it disappears behind the posterior border of the sternocleidomastoid muscle where another loose ligature of umbilical tape was secured. All small perforating branches of the external jugular vein were ligated with 3-0 silk and divided close to their entrance. The anterior border of the sternocleidomastoid muscle was retracted posteriorly to expose the dissection of the fascia and fibroareolar tissue layers that overlie the carotid sheath. The carotid sheath was incised and, after identification of the vagus nerve, a loose ligature of umbilical tape was made to encircle the carotid artery at the level of the third tracheal ring. The incision in the carotid sheath was continued cephalad to the bifurcation of the common carotid artery (or the superior thyroid artery) where another loose ligature of umbilical tape was secured.

Implantation of the respective modified collagen vascular prosthesis was carried out as follows: "Booties" were constructed of 2 cm segments of heavy rubber tubing and placed around the free ends of the loose ligatures of umbilical tape. These booties were then drawn down the ligatures to prevent flow in the respective vessel and held in place with Kelly clamps. Two transverse incisions approximately 2.5 cm apart were made in the respective blood vessel with a pair of fine curved Metzenbaum scissors. The vascular prostheses (3.5 cm in length for macroscopic studies and SEM evaluation and 10.0 cm lengths for streaming potential measurements), mounted over hydrochloric acid-cleaned stainless steel cannulas (1.5 cm in length) doubly ligatured with 3-0 silk were implanted via their free cannula ends into the rents in the vessel and secured in place with two ligatures of 3-0 silk. The two anastomoses thus prepared, the connecting piece of blood vessel was severed thus permitting the graft to lie flat in its "fascial bed." The Kelly clamps were then removed relieving tension from the "booties" over the umbilical tapes (distal to blood flow removed first) and allowing blood to flow through the vascular prosthesis. The vascular grafts were left in place for the specified periods of time with the wound covered with saline-soaked gauze pads.

The internal diameters of the blood vessels and the prostheses were calculated by measuring the external diameters of the vessels and implants with blood flowing through the system with a micrometer and substracting two wall thicknesses from these diameters. The discrepancies between the internal diameters of recipient and prosthetic blood vessels are found in Table 1.

TABLE 1 __________________________________________________________________________ DISCREPANCY BETWEEN INTERNAL DIAMETER OF RECIPIENT AND PROSTHETIC BLOOD VESSELS *I.D. of Recipient *I.D. of Prosthetic Change in Vessel Vessel (mm) Vessel (mm) Internal Diameter __________________________________________________________________________ Carotid 4 10 6 Artery 5 9 4 5 10 5 5 8 3 4 10 6 x = 4.6 x = 9.4 x = 4.8 Jugular 8 10 2 Vein 7 9 2 8 10 2 8 10 2 8 10 2 x = 7.8 x = 9.8 x = 2.0 Femoral 4 10 6 Artery 5 9 4 4 10 6 5 8 3 4 10 6 x = 4.4 x = 9.4 x = 5.0 Femoral 7 10 3 Vein 7 8 1 7 9 2 8 10 2 7 10 3 x = 7.2 x = 9.4 x = 2.2 __________________________________________________________________________ *I.D. = Internal Diameter

For femoral artery and femoral vein implantation, an incision overlying the femoral triangle was extended caudad over the medial aspect of the thigh to the level of the knee joint. The superficial femoral fascia was dissected free from its loose attachment to the underlying femoral sheath revealing the femoral artery and vein lying within the sheath at the fossa ovalis. The femoral vessels were dissected free of their sheath and the branches of the femoral artery (superficial circumflex iliac artery, external pudendal artery, medial femoral circumflex artery and lateral femoral circumflex artery) were identified, ligated with 3-0 silk and divided close to their origin. A loose ligature of umbilical tape was encircled about the origin of the femoral artery. The rostral branches of the femoral vein were likewise identified, ligated with 3-0 silk and divided near their entrance and the femoral vein was also encircled with a loose ligature of umbilical tape. The sartorius muscle was retracted laterally and the adductor longus muscle retracted medially with a pair of self-retaining retractors thus exposing the descending genicular artery and vein. Blunt dissection was carefully continued deep to the descending genicular vessels to expose the caudad continuation of the femoral artery and vein. Small perforating vessels and muscular branches were ligated with 3-0 silk and divided close to their origin or entrance. Loose ligatures of umbilical tapes were secured around the respective femoral vessel and its descending genicular branch. Implantation of the respective modified collagen vascular prosthesis was carried out as previously described..

For streaming potential measurements, a pair of silversilver chloride electrodes 22 and 24 (FIGS. 2 and 3) were placed into the lumen 26 of the vascular prosthesis 27 through the center of two purse-strings 28 and 30 of 5-0 Dacron sewn into its superficial surface and separated by a distance equivalent to at least 10 radii. Electrodes were attached via conventional shielded leads to a Kiethley electrometer 32 (Sawyer, P. N., Himmelfarb, E., Lustrin, I., and Ziskind, H.: Measurement of streaming potentials of mammalian blood vessels, aorta, and vena cava, in vivo. Biophys. J. 6: 641 (1966)).

For removal, fixation and macroscopic observation, at the conclusion of the respective time periods, the blood vessels upstream and downstream from both the arterial and venous implants were doubly cross-clamped with straight Kelly clamps and divided between the two clamps. Both prostheses (arterial and venous) filled with blood and occluded at both ends with the Kelly clamps were immediately removed and placed into a 10% formalin solution. While remaining under the fluid level of the formalin, the Kelly clamps were removed along with the stainless steel cannulas and the 3-0 silk ligatures and the vascular implants were placed in appropriately labeled tubes containing 10% formalin. The degree of thrombotic occlusion was graded on a scale of 0 to 4+ by three independent observers in a single-blind manner. The averages of the three graded values for each vessel and time period appear in Table 2.

TABLE 2 __________________________________________________________________________ MACROSCOPIC OBSERVATION OF MODIFIED COLLAGEN VASCULAR PROSTHESES IMPLANTATION* Blood Dialdehyde Time Vessel Unmodified Positive Negative Neutral Starch Tanned __________________________________________________________________________ 0 min LFA 0 2+ 0 0 0 0 min LFV 0 4+ 0 0 0 15 min LCA 1+ 4+ 0 1+ 0 15 min LJV 2+ 4+ 0 1+ 1+ 30 min RFA 1+ 4+ 1+ 2+ 1+ 30 min RFV 4+ 4+ 1+ 2+ 2+ 2 hrs RCA 3+ 4+ 1+ 3+ 2+ 2 hrs RJV 3+ 4+ 1+ 3+ 3+ __________________________________________________________________________ *The degree of thrombotic occlusion graded on a scale of 0 to 4+ by three independent observers. 0 = macroscopically clean prosthetic intimal surface; 1+ = junctional thrombi at the stainless steel cannula-prosthesi interface; 2+ = junctional thrombi plus macroscopically evident mural thrombi adherent to the prosthetic wall; 3+ = junctional thrombi plus macroscopically evident mural thrombi significant enough to cause marked stenosis of the graft lumen; 4+ = total thrombotic occlusion of the prosthetic vascular lumen. All scaled observations were carried out "single-blind. (LFA = Left Femoral Artery; LFV = Left Femoral Vein; LCA = Left Carotid Artery; LJV = Left Jugular Vein; RFA = Right Femoral Artery; RFV = Right Femoral Vein; RCA = Right Carotid Artery; RJV = Right Jugular Vein)

The degree of thrombotic occlusion of the variously modified ficin digested bovine vascular heterografts is seen in Table 2 above. The unmodified arterial and venous heterografts did not initially thrombose when exposed to the blood elements. As the implantation time increased, however, it can be noted that both the arterial and to a greater extent the venous vascular prostheses became thrombotically occluded. A similar sequence of events was found to occur with the neutrally modified collagen prostheses and the dialdehyde starch tanned vessels. In the latter case, the thrombotic events proceeded at a slower pace.

The chemical procedure carried out in order to cause an increase in the net positive surface charge density on the intimal surface of the vascular heterografts resulted in an acceleration of the thrombotic events. It can be noted from Table 2 that the mere contact of the blood elements with the prosthetic vascular surface resulted in significant occlusion of the vessel lumen (more marked in the slower flowing venous implant). Total thrombotic occlusion occurred in the venous implant immediately and was noted within 15 minutes in the arterial grafts.

The negative modification reaction carried out on the collagen tubes resulted in a most favorable performance of these implants. It can be seen from Table 2 that the thrombotic events occurring on the intimal surface of the vascular prostheses were significantly prolonged. A macroscopically clean intimal surface was achieved on both the arterial and venous grafts for up to 15 minutes of blood flow. After 2 hours of contact with the flowing blood elements only junctional thrombi at the stainless steel cannula-prosthesis interface was noted.

The results of the in vivo arterial and venous streaming potential measurements of modified collagen vascular prostheses implantations appear in Table 3.

TABLE 3 ______________________________________ IN VIVO STREAMING POTENTIAL MEASUREMENTS OF MODIFIED COLLAGEN VASCULAR PROSTHESES VASCULAR ARTERIAL STREAMING VENOUS STREAMING PROSTHESIS POTENTIAL (mv) POTENTIAL (mv) ______________________________________ Dialdehyde +0.1 -0.1 Starch Tanned +0.2 -0.1 +0.2 +0.1 x = +0.13 x = -0.03 Positive Ficin -0.8 -0.4 Digested -0.5 -0.4 -0.8 -0.4 x = -0.7 x = -0.4 Negative Ficin +0.7 +0.3 Digested +0.4 +0.3 +0.7 +0.6 x = +0.6 x = +0.4 Neutral Ficin +0.3 +0.1 Digested +0.2 +0.2 +0.2 +0.1 x = +0.23 x = +0.13 Unmodified -0.2 -0.1 Ficin Digested ______________________________________

The polarity of the streaming potential reflects the interfacial potential between the vascular prosthesis and the blood flowing through the graft with a positive streaming potential recorded at the downstream electrode indicating a net negative surface charge density on the intimal surface of the graft with respect to the blood elements.

The polarity of the in vivo arterial and venous streaming potential measurements obtained from the implantation of the unmodified ficin digested bovine vascular heterografts was on the order of -0.2 and -0.1 mv, respectively. These values indicate a slightly positive interfacial potential exists between the intimal surface of the vascular graft and the "streaming" blood elements. An even greater magnitude of negative streaming potential (positive interfacial potential) was noted with the arterial and venous heterografts chemically modified to increase the positive surface charge density (-0.7 mv-arterial; -0.4 mv-venous).

Collagen vascular heterografts chemically modified to neutralize the electrical surface charges on the prosthetic intimal wall produced arterial and venous streaming potentials on the order of +0.23 and +0.13 mv, respectively. These collagen vascular implants chemically altered so as to increase the negative surface charge density gave positive streaming potential measurements of greater magnitude (+0.6 mv-arterial; +0.4 mv-venous). These latter values indicate that a significant negative interfacial potential was in fact produced on the intimal surface of the vascular grafts by the chemical modification procedure.

An ideal vascular prosthesis would possess the recoil characteristics of an intact elastica plus the strength of a collagen backbone. The strength of this collagen backbone to withstand many months of repetitive pulsatile arterial pressures and function as a prosthetic vascular heterograft has been documented (Rosenberg, N., Henderson, J., Lord, G. H., and Bothwell, J. W.: Collagen arterial prosthesis of heterologous vascular origin: physical properties and behavior as an arterial graft. In: Biophysical Mechanisms in Vascular Homeostasis and Intravascular Thrombosis, P. N. Sawyer, Ed. Appleton-Century-Crofts, New York 1965, pp. 314-321). Furthermore, many of the desirable features of a vascular prosthesis (Wesolowski, S. A., Fries, C. C., and Sawyer, P. N.: Some desirable physical characteristics of prosthetic vascular grafts. In: Biophysical Mechanisms in Vascular Homeostasis and Intravascular Thrombosis, P. N. Sawyer, Ed. Appleton-Century-Crofts, New York 1965, pp. 322-336) are present in ficin digested bovine carotid artery preparations. Enzyme treatment removes the parenchymal elements subject to necrosis and thus prevents the associated inflammatory reaction which impairs the integrity of the graft wall and predisposes to intimal thrombosis. It further produces a degree of porosity of the prosthetic vascular wall which favors the orderly invasion by host connective tissue and due to the relative absence of inflammatory reaction, favors early endothelialization as well.

Early thrombosis after insertion of these grafts can be inhibited by increasing the negative surface charge density on their intimal surfaces. The above results substantiate the importance of the negative intimal surface charge density in preventing early thrombotic occlusion of collagen heterograft vascular prostheses. Chemical modification designed to increase the negative surface charge density on the inner wall of ficin treated bovine arterial heterografts was found to prevent luminal occlusion by mural thrombosis significantly. Chemical modification by a carbodiimide-promoted amide formation reaction designed to increase the positive surface charge density on the intimal surface was found to greatly accelerate the thrombotic events (Table 2).

The polarity of the arterial and venous streaming potentials measured through the lumens of these vascular prostheses served to check the chemical modification procedures and assure that the electrical alteration of surface charge density occurred with the predicted polarity. It can be seen from Table 3 that the arterial and venous streaming potentials measured through the lumens of those grafts modified to increase the intimal surface electronegativity were uniformly positive in polarity, while those grafts designed to present a positively charged surface to the flowing blood elements produced streaming potentials with a negative polarity. Thus the chemical modification procedures were effective in modulating the polarity of the interfacial potential between the luminal surface of the vascular prosthesis and the blood elements flowing within.

Ficin digested bovine carotid artery vascular heterografts were also implanted into the vascular system of mongrel dogs in their unmodified state and in a chemically altered condition designed to neutralize the electrical surface charges on the graft. It can be seen from Table 2 that both of these implants were inferior in their performance when compared with the negatively modified surface, but superior to the more positively charged graft.

The streaming potential measurements obtained from the unmodified ficin digested collagen grafts indicate that their interfacial polarity is slightly positive while that of the neutrally modified implants is slightly negative, (Table 3). It can be seen from these results that the chemical procedure designed to neutralize the slightly positive surface charge on the ficin digested surfaces (unmodified) actually overshot electrical neutrality and produced a surface with a slight excess negative charge. Even though the excess charges on the surfaces of these two vascular prostheses were opposite in polarity, their funtioning as vascular grafting material as judged macroscopically was comparable. This apparent lack of difference in performance between the two can probably be accounted for on the basis of their relatively small surface charge density (as compared with the positively and negatively modified vascular prostheses -- see Table 3).

A further type of ficin digested bovine vascular heterograft was also evaluated. This dialdehyde starch tanned prosthesis has previously been reported to perform well in long term (3 years) studies (Bothwell, J. W., Lord, G. H., Rosenberg, N., Burrowes, C. B., Wesolowski, S. A. and Sawyer, P. N.: Modified arterial heterografts: relationship of processing techniques to interface characteristics. In: Biophysical Mechanisms in Vascular Homeostasis and Intravascular Thrombosis, P. N. Sawyer, Ed. Appleton-Century-Crofts, New York, 1965 pp. 306-313). It can be seen from Table 2 that this grafting material was markedly inferior to the negatively modified collagen prostheses but comparable to the unmodified and neutrally altered collagens. The streaming potential measurements obtained from these implants were quite ambiguous. The arterial streaming potentials were uniformly positive in polarity, indicating that the prosthesis presents a slightly negative surface to the flowing blood. The venous streaming potentials, however, were not consistent, two of the three measurements indicating that the graft actually presents a slightly positive surface to the bloodstream.

The critical role of platelets in intravascular thrombosis has been well documented in the following:

Mitchell, J. R. A.: Chapter XVI Platelets and Thrombosis, Sci. Basis Med. Ann. Rev. 266-288 (1968);

Hampton, J. R.: The study of platelet behavior and its relation to thrombosis. J. Atheroscler. Res. 7:729 (1967);

Mustard, J. F., Packham, M. A. Rowsell, H. C., and Jorgensen, L.: The role of platelets in thrombosis and aterosclerosis. Thromb. Diath. Haemorrh. Suppl. 23:261 (1967).

It is believed by some that the initial event leading to the formation of an intravascular mural thrombus is the unmasking of subintimal collagen and its subsequent recognition by circulating platelets with the latter's resulting adhesion and aggregation (this view is not universally accepted). It is known, however, that collagen is capable of initiating platelet adhesion to itself and that this adhesion can result in platelet aggregation. The precise biochemical mechanism by which platelets recognize collagen and subsequently adhere to it has been shown to involve the formation of a complex between incomplete carbohydrate side chains in collagen and glucosyltransferase present on the outer surface of platelets (see FIG. 1). The reaction involves the enzymatic coupling of glucose (supplied by the platelets as uridinediphosphoglucose) to galactosyl residues attached to hydroxylysine side chains incorporated into the collagen peptides.

The reaction employed to increase the net negative surface charge density on the intimal surface of the collagen vascular prostheses was a succinylation reaction designed to cover the free amino groups of the protein. Succinylation of solubilized collagen has been demonstrated to result in an approximately 95% conversion of epsilon amino groups to free carboxyl groups (Gustavson, K. H.: Akiv. for Kemi 17:541 (1961)). This reaction was observed to result in the production of a vascular prosthesis which when implanted into the vascular system of mongrel dogs resulted in the least amount of platelet deposition and subsequent thrombotic occlusion. It can be seen from the above biochemical data that this succinylation reaction which covered the epsilon amino groups of the protein sufficiently altered the substrate of platelet glucosyltransferase such that the specificity of this enzyme no longer recognized the heterograft collagen as such. Without platelet recognition there could be no adhesion to the prosthetic intimal surface and thus no thrombotic occlusion.

Furthermore, it has been shown that blocking the free carboxyl groups of collagen results in increased platelet aggregation activity in vitro. This has been postulated to result from the potentiation of the effects of the free epsilon amino groups (Wilner, G. D., Nossel, H. L., and LeRoy, E. C.: Aggregation of platelets by collagen. J. Clin. Invest. 47:2616 (1968 )). The results obtained in this study with the reaction designed to increase the net positive surface charge density on the intimal surface of the collagen vascular prosthesis by amidation of the free carboxyl groups is entirely consistent with these in vitro studies. The positive modification reaction resulted in the rapid and total thrombotic occlusion of these vascular heterografts when implanted into the vascular system of mongrel dogs.

A negative intimal surface charge density results in a favorable vascular heterograft performance by preventing early thrombotic occlusion of the prosthesis' lumen. A simple chemical modification procedure is given by which the intimal surface of ficin digested bovine carotid arteries can be prepared with this added electronegativity.

In addition to the above, the invention provides a new kind of prosthesis consisting of a ficin digested artery modified with anionic dialdehyde starch. This prosthesis will be tanned and will have had negative charges incorporated into it in one step. Thus the resulting product will be strengthened as by dialdehyde-starch tanning and will have an excess of negative charge so as to make a better non-thrombogenic surface. The dialdehyde-starch derivative needed is the 6-carboxy-methyl derivative I ##SPC3##

This material is either available or can be made. It is available as a product called "Water Soluble Anionic Polymeric Dialdehyde" known as DASOL A available from Miles Laboratories, Inc., which is of interest as a tanner and suitable charge carrier.

As a dification of an existing prosthesis to introduce negative charge, a ficin digested dialdehyde starch modified artery can be further modified with succinic anhydride to introduce negative charges. The reaction of dialdehyde starch may occur mainly with the .epsilon.-amino groups of lysine under conditions of tanning (pH 8.8). Some reaction occurs with the guanidino groups of arginine as it is known that this group reacts with neighboring dicarboxyl groups which is one way of looking at dealdehyde-starch. Be that as it may, there may be sites which are still available for attack by succinic anhydride. This will result in a tanned, negatively charged prosthesis.

Further, there is the exhaustive decationization of a succinic anhydride modified prosthesis. Succinic anhydride reacts mainly with the .epsilon.-amino groups of lysine introducing a negatively charged carboxy propeonoyl moeity. Assuming all the lysine groups are blocked, positive charge sites exist on the arginine residues and on the histidine residues (some of the imidazole groups of histidine may have the positive charge removed at pH 7.4). Modification of the arginine residues can be effected by phenyl glyoxal which reacts under mild conditions with the positively charged guanidinum group and results in a neutral derivative. Histidine can be converted to a neutral derivative by reaction with diethyl pyrocarbonate to given an ethoxyformyl derivative on the imidazole ring.

The above techniques involve the negatively charged tanning of collagen for protection against intravascular thrombosis. It is appropriate to apply these techniques to other areas in the body when collagen prostheses and, in fact, heterograft materials are used which are taken from various species of animals and used in humans. The most significant of these prosthetic devices are homograph valves taken from one human and placed in another human and heterograft valves which conventionally are taken from pigs and inserted into the aortic valve area. These valves are known to have certain advantages in that their leaflets permit central flow. The basic problem with various aspects of their application has always been that, while they work satisfactorily in the aortic position, there has always been some question about their utility in the mitral valve area. The negatively charged tanning of these valves enhances their utility.

By way of further embodiment of the invention, a heterograft valve is removed under semi-sterile conditions from the porcine heart. It is tanned using the negatively charged tanning techniques described above. The tanning procedure is effected using either succinic anhydride or dasol to produce a highly negatively charged collagen surface on the valve which makes the prosthesis more anticoagulent in character than it would be under conventional circumstances. The charge characteristic provides not only additional antithrombogenesis but also increases the life span of the implanted heterograft valve because of its resistance to fibrin deposition and destruction. The same also applies to homograph valves.

There will now be obvious to those skilled in the art many modifications and variations based on the above. These modifications and variations will not depart from the scope of the invention as defined by the following claims.

Claims

What is claimed is:

1. A method of improving the performance of a graft prosthesis comprising making the prosthesis of collagen and increasing the negative surface charge of the intimal surface thereof by a succinylation reaction.

2. A method as claimed in claim 1 wherein the collagen is prepared from ficin digested bovine carotid artery material.

3. A method as claimed in claim 1 wherein the prosthesis is formed as a vascular prosthesis.

4. A method as claimed in claim 1 wherein the prosthesis is formed as a valve.

5. A prosthesis made as claimed in claim 1.

6. A method of improving the performance of a graft prosthesis comprising making the prosthesis of collagen, the prosthesis including protein with free amino groups, and increasing the net negative surface charge of the intimal surface thereof by covering the free amino group.

7. A method of improving the performance of a graft prosthesis comprising making the prosthesis of collagen and increasing the negative surface charge of the intimal surface thereof by the following chemical reaction ##SPC4##

8. A method of improving the performance of a graft prosthesis comprising making the prosthesis of collagen and increasing the negative surface charge of the intimal surface thereof, said collagen being dialdehyde starch tanned collagen which is treated with succinic anhydride.

9. A method as claimed in claim 8 comprising further modifying said surface with phenylglyoxal.

10. A method as claimed in claim 8 comprising further modifying said surface with diethyl pyrocarbonate.

11. A method of improving the performance of a graft prosthesis comprising making the prosthesis of collagen, said graft being in the form of a collagen tube, comprising closing off an end of said tube, inserting said end into a fluid and inserting a liquid chemical succinylation reactant into the lumen of the tube to increase the negative surface charge of the intimal surface thereof.

12. A method as claimed in claim 11 wherein the chemical reactant is succinic anhydride in a basic solution.

13. A method as claimed in claim 12 wherein a NaHCO.sub.3 buffer solution is inserted into the lumen in sequential additions of approximately 10 ml. aliquots to which about 0.1 gm. of crystals of succinic anhydride were respectively added.