- Email: [email protected]
An Elastic Dacron Arterial Substitute
An Elastic Dacron Arterial Substitute D. EMERICK SZILAGYI, M.D., F.A.C.S.*
UNTIL a few months ago the surgeon had small cause for concern about the choice of the type of arterial substitute he was to use for a grafting operation. The view was generally accepted that a well-processed homologous arterial graft not only looked and behaved like the patient's own undiseased artery but that it would continue indefiqitely to function in the manner of a normal blood vessel. Except for the problems of obtaining and processing them, which were considerable, arterial homografts appeared to be the ideal substitute. Recently, however, as the result of continued postoperative observations of the fate of homologous arterial implants, a radical change has taken place in one's estimate of the value of grafts of this type. It has been learned that the lasting quality of these homotransplants is unsatisfactory.! Histological studies have shown that arterial homografts are not actively accepted and incorporated but only passively tolerated and at times even rejected by the host, and that the implanted grafts must maintain their structural integrity through the mechanical strength inherent in their histological elements. When these wear out, as they inevitably do in time, the wall of the graft weakens, becomes dilated and is eventually occluded by thrombosis, or rarely it ruptures. Since the histologic elements essential for mechanical integrity are the elastic laminae, in aortic grafts-extremely rich in elastic components-the wearing-out process may take many years; indeed, in most cases the graft outlives the patient. Femoral grafts, on the contrary, being very poor in elastic tissue and consisting mostly of smooth muscle and collagen fibers, show structural failure relatively early and in a high incidence. These observations largely invalidated the clinical usefulness of arterial homografts. In spite of their versatility, ease of technical handling and excellent early success rate, the use of these substitutes has been gradually given up. Aside from the recognition of their intrinsic deficiency, which made
The experimental work described was in part supported by a grant/rom the Michigan Heart Association.
* Surgeon-in-Charge, Second General Surgical Division, Henry Ford Hospital, Detroit, Michigan. 1523
1524
D. Emerick Szilagyi
their obsolescence inevitable, the passing of arterial homografts was hastened by the emergence of arterial substitutes constructed of woven yarn of plastic materials of diverse types. Investigation of these prostheses had in fact begun, mostly for reasons of convenience, even before the inadequacy of homografts was established, following the pioneering observation of Voorhees and his associates reported in 1952 that the animal body builds a connective tissue channel around a porous textile fabric placed in the path of the flowing blood. The awareness of the lack of permanence in the structural strength of arterial homografts lent further urgency to this search, and as a result a number of plastic substitutes have been put to clinical use. In what follows I wish to describe the experimental qualities and clinical use of an arterial prosthesis made of Dacron* textile fabric developed in our laboratory. EXPERIMENTAL OBSERVATIONS
In our experimental search for a plastic arterial substitute during the past five years we have kept in sight certain desirable qualities that such a prosthesis must possess in order to be suitable for wide-range clinical use. The arterial prosthesis must be seamless, smooth-walled, easy to handle, durable and relatively nonreactive; its porosity and bulk must meet certain critical requirements; and finally it must possess the ability to elongate and recoil. (Further important but not truly essential requirements for such a prosthesis would be economical cost, ease of storage and simplicity of sterilization.) The reasons for the significance of these attributes are cogent. The presence of a seam and of ridges on the luminal surface of the prosthesis is undesirable since surface irregularities would increase friction and, therefore, resistance to flow;2 moreover they would promote excessive clotting along the areas of contact with the blood, a circumstance that, under many marginal conditions, may mean the difference between patency andthrombotic occlusion. The need for ease of handling (as well as for the less essential factors of low cost and simplicity of storage and sterilization) is self-evident. Durability is also an obvious requirement. It is less obvious that durability is largely dependent on lack of tissue reactivity. Regardless of the magnitude of tensile strength that the original fabric possesses, if after implantation the fabric provokes active cellular reaction in the host tissues, it will be attacked by the exudative process, and in the end, partly or completely destroyed. The optimal degree of porosity in the fabric is the sum of two requirements: the mesh size must be large enough to allow ingress for the ingrowth of connective tissue but it must also be small enough not to interfere with hemostasis. The restrictions on the bulk (or wall thickness) of the prosthesis are less easy to justify. Undoubtedly thick-walled, heavy 'prostheses have·served
* "Dacron" is the registered trade name of E. I. du Pont de Nemours & Company for their polyester fiber.
A n Elastic Dacron Arterial Substitute
1525
Fig. 416. Dacron yarn unstretched (above) and under tension (below). The ability to elongate comes from the uncurling of the fibers when put on stretch.
well in many situations during short-term observations. However, a plastic prosthesis is a foreign body, and sound surgical principles would suggest that its mass should be kept to a minimum as long as this restriction does not result in a disadvantage such as lessening its initial tensile strength or its durability. While conclusive evidence condemning thickwalled prostheses cannot at this time be quoted, it is an easily confirmed observation that the connective tissue reaction around bulky prostheses is excessive, and this is undesirable. The quality of elasticity must be present owing to the necessity for the prosthesis to accommodate itself to the anatomical features of its surroundings. A rigid, nonextensile tube when curved or bent will buckle, and this in turn leads to clotting. It may be also mentioned that the accomplishment of elasticity through specially treated fibers appears to have advantages over methods involving the treatment of the finished tube, namely crimping, which increases bulk and luminal friction. The prosthetic tube we selected for investigation seemed to meet these requirements to a satisfactory degree. This elastic Dacron prosthesis is woven with a taffeta weave and is seamless (Fig. 416). As disposed in the tubular fabric, the longitudinal threads forming the warp are made of fibers that have been specially processed (Fig. 417); the transverse threads, or woof or filling, consist of untreated Dacron fibers. The special process to which the fibers of the warp yarn have been subjected is the same as that used for the manufacture of "Helanca" nylon fiber. In this process the fibers are twisted into tight spirals and then heat-set at 270 0 F. Fibers of opposite directions of twist are combined into the yarn in
D. Emerick Szilagyi
1526
Fig. 417. Simple taffeta weave of the Dacron fabric. It is the fluffy-appearing strands forming the warp (horizontal in this photograph) that have the quality of elasticity.
about equal amounts in order to assure balance. The fibers are stretchable since on pull the spirals uncurl; when the pulling force is released the fibers shorten owing to a tendency to resume their heat-set spiral shape. When yarn composed of such fibers is woven into fabric, this quality of elasticity is largely retained (Fig. 418). It is necessary, however, to boil the fabric in liquid soap to restore complete resiliency. (The contact with soap leaves no deleterious effects provided the fabric is thoroughly rinsed.) The finished tube has an ability to elongate 15 to 20 per cent without deformity. The textile characteristics of the prosthesis used experimentally are given in Table 1. The mesh size and the weight of the yarn were arrived at after many experimental trials and appeared to fulfill optimally the Table 1 TEXTILE CHARACTERISTICS OF ELASTIC DACRON PROSTHESIS
Diameter: 10 mm. Ends (warp): 80 to 160 per centimeter Picks (woof): 47 per centimeter The yarn in the warp: 70-denier, 2-ply, 70-14 Bright Type 51 Dacron The yarn in the woof: 220-denier, 220-50 Bright Type 51 Dacron
An Elastic Dacron Arterial Substitute
1527
Fig. 418. Double exposure of an elastic bifurcation prosthesis showing the range of stretch of the fabric.
basic criterion of minimum bulk with maximum strength. The porosity is such that preclotting is necessary (see below). The canine preparation employed for the experimental investigation of the prosthesis was that first described by McCune and slightly modified by us. The details of the operative technique were given in full in earlier publications. 3 Briefly, the procedure consists in implanting a bypass between the upper thoracic and lower abdominal canine aorta, the proximal anastomosis being an end-to-end one, thus assuring a complete shunting of the aortic blood into the prosthesis. The method permits the insertion of implants 18 to 26 cm. in length and 7 to 10 mm. in diameter (depending on the size of the animal). We believe that this preparation affords a much more realistic test of a prosthesis than the customary short-segment replacements in many of which the bridging effect of the healing process of the stumps of the host aorta is difficult or impossible to distinguish from the histologic changes peculiar to arteriogenesis. Twenty-one such preparations were made, in ten of which the prosthesis was implanted with enough slack to allow a tortuous course, thus)esting the ability of the prosthesis to mold. The experimental animals were investigated by observation of their general condition and state of health; by periodic aortograms; and by scheduled sacrifice. The recovered grafts were inspected for gross appearance, tested for tensile strength and studied for histologic characteristics. All the implants remained patent until the death of the animals. In 12 animals that were sacrificed on schedule after periods of time up to 24 months after implantation, and in five animals that died of intercurrent causes not related to the function of the implants, the recovered prosthesis showed excellent arteriogenesis (Figs. 419 and 420), and a remarkably slight tendency to elicit inflammatory reaction in the host tissues
D. Emerick Szilagyi
1528
Fig. 419. Gross appearance of a canine implant 6 months after insertion. The fabric is completely covered by a smooth, flawless pseudo-intima about 1 mm. in thickness. The connective tissue shell and ingrowth are firmly adherent to the fabric and thin but sturdy. The mean total thickness of the new artery is 1 to 1.5 mm.
r.
Fig. 420. Microscopic section of the midportion of implant shown in Figure 419. Very satisfactory process of arteriogenesis. Cellular exudative reaction is almost completely absent. The growth of connective tissue appears to serve only one purpose: the integration of the plastic fibers into a firm structure. (X70, H. & E. Stain.) (Reproduced by permission of A.M.A. Arch. Surg. 3 )
An Elastic Dacron Arterial Substitute
1529
DURABILITY OF PLASTIC ARTERIAL SUBSTITUTES
He/anca Dacran .c
.~ 50 L!)
ci
"-
,,
c: o=>
\
.;g 40 a.
-
.;:
\
30
C>
c:
Q)
~ 20
\
,, , ,
\
\
\
,, \
\
e-----
--------... ... ~ ...
He/anca Ny/on
---
--.-- ----- - - - ---~--
--... ...........
~-.---r--.--.---r--.---.-~---r--.--.---.-./-,/-.-!~ 2 3 4 5 6 7 8 9 10 II 12 18 24 36
Duration of Implantation (Months)
Fig. 421
(Fig. 420). During 24 months the tensile strength of the implants remained essentially unchanged in marked contrast to the findings with "Helanca" nylon prostheses in which the tensile strength rapidly declined after two months' implantation (Fig. 421). The excellent durability and the remarkably complete angiogenesis observed about the implanted tubes justified the clinical application of the prosthesis. CLINICAL EXPERIENCE
Technical Details in the Use of the Prosthesis
In external appearance, the prosthetic tube of stretch Dacron as used by us is white (or, rather, colorless), soft, pliable, smooth, light and moderately porous. On handling it, one is impressed by the lack of bulkiness that is peculiar to knitted or crimped tubes and the absence of stiffness that is characteristic of Teflon textiles. The prosthesis does not require any special precautions for storage. It can be sterilized chemically, or by boiling and autoclaving. As a sterilizing agent Zephiran Chloride in a 1: 1000 solution for 18 hours of contact has been effective. It should be noted that repeated autoclaving reduces elasticity slightly. Without some appropriate pretreatment, a fabric of the porosity of the elastic Dacron prosthesis will leak blood rather vigorously. We have found that by rinsing the inner and outer surfaces of the prosthesis with 15 to 20 ml. of blood obtained from an unheparinized segment of the exposed arterial tree, and by allowing the prosthesis to soak in this blood for three or four minutes one can bring about just enough clotting to seal off the meshes of the fabric. A moderate amount of oozing may still occur after the prosthesis has been inserted and blood allowed to flow through
1530
D. Emerick Szilagyi
it, but this can be readily controlled by short periods (20 to 30 seconds) of occulsion of the inflow while the blood is allowed to remain in the prosthesis. It is of the utmost importance to use great care in the process of clot-sealing. After the in-vitro clotting, the prosthesis must be checked for the presence of visible or palpable thrombi, and if any are found they must be removed by gentle milking or, if necessary, by rinsing with a heparin-saline solution (0.5 mg. of heparin per 1.0 ml. of saline). The same precaution must be observed during the performance of the insertion of the implant. Obviously, stagnant blood is never allowed to remain in the prosthesis once the clot-sealing has been accomplished. If owing to some technical mishap it becomes necessary to stop the flow of blood through a prosthesis already in place, before the distal clamp is applied the blood is milked out of the implant and the implant is filled with heparin-saline solution until the flow may be re-established. Leaving a layer of thrombus in the lumen of the prosthesis that is thicker than about 0.1 to 0.2 mm. or allowing the presence of lumps of cloth may have disastrous consequences. It may not only delay or prevent the orderly formation of the pseudo-intima, thus resulting in early occlusion, but it may lead to a rapid, almost immediate, progression of the clotting process to a point where the prosthesis cannot function at all. The free cut end of a woven fabric tends to unravel, particularly if the component yam contains smooth-surfaced fibers as is the case with the elastic Dacron prosthesis. This tendency can be easily corrected, however, by melting and fusing the ends of the cut fibers through the application of heat. We have found it convenient to heat-seal the prosthesis before beginning the implantation. The length of the implant is judged by placing a heavy black thread in the exact position the prosthesis is to occupy and by using this measure as the base for the simple calculation. Ten to 15 per cent of the length of the thread (depending on the slightly variable elasticity of the plastic tube) is subtracted to obtain the desired dimension. In case of doubt, care is exercised to make the implant slightly longer rather than shorter than is ideal. The ends of the prosthesis of the desired length are cut with scissors at the angle needed for the proposed anastomosis. The cut ends are lightly touched with a flat-tipped electric cautery which has been heated just short of being red. The ideal heat-seal is a narrow (about 1 mm.) even and smooth band of fused plastic which lies exactly in line with the scissors cut. A light touch and attention to technique are necessary for the proficient use of any plastic prosthesis, and they are of special import when dealing with light woven fabrics. The manual handling and the suturing of the elastic Dacron prosthesis demand great gentleness. The prosthesis must not be picked up with instruments having a rough surface, and clamps applied to it must be rubber-shod. If, nevertheless, separation of some of the threads occurs, the flaw can usually be corrected by gentle
An Elastic Dacron Arterial Substitute
1531
rubbing between fingers. In placing sutures only the slightest strain is allowed on the fabric. Experimental observation reported elsewhere2 have shown that in by-pass grafting the optimal ratio of the size of the prosthesis to the size of the host artery is about 1.6: 1.0, or, roughly, the prosthesis should be half again as large in diameter as the host artery. The technical details enumerated are valid for the straight tubular as well as for the bifurcated prostheses. Because of problems inherent in weaving techniques, the undersurface of bifurcations (the so-called "crotch") is not truly seamless. Although we have found that the seal of the weave at the crotch has been consistently strong, it is advisable to inspect this part of the bifurcation prosthesis with particular attention before use. Clinical Results
In assessing the results of the operations in which elastic Dacron prostheses were used as graft material, certain principles of operative case selection must be kept in mind. During the period in question (June 1957 to May 1959), in dealing with arterial lesions requiring angioplastic correction, we used both grafting procedures (in about 80 per cent of the cases) and thromboendarteriectomy. We reserved thromboendarteriectomy for instances in which the occlusive process did not extend beyond the bifurcation of the common iliac artery or to those cases of femoral lesions in which the disease process was sharply limited to a segment no longer than 5 cm:. It can be seen, therefore, that a factor of selectivity was operative and that it was a factor prejudicial to the procedures utilizing plastic prostheses. In general, it was the more advanced and more diffuse lesions that were treated with the last-named methods. In the aorto-iliac area three types of procedure involving the use of plastic prostheses were employed. Direct replacement of the aorto-iliac segment was resorted to in aneurysmal disease or in cases in which the occlusive process extended for some distance into the aorta. In instances of aortic aneurysm, the entire disease segment was resected and replaced; in occlusive lesions of the aorta, on the other hand, the iliac arteries were not, as a rule, dissected out and replaced. In all operations of direct replacement the proximal anastomosis (prosthesis to aorta) was end-toend, while the distal anastomoses were end-to-side, either at the iliac, the common femoral or popliteal level, according to need. In cases of unilateral iliac occlusive disease, when the occlusive process extended beyond the iliac bifurcation, and thus endarteriectomy was deemed unsuitable, aortofemoral by-passes were used, with the proximal anastomosis at the distal aortic level (distal to the origin of the inferior mesenteric artery), and the distal anastomosis at the common femoral or popliteal level, and with both anastomoses of the end-to-side variety. In extensive bilateral iliac occlusive disease without significant involvement of the aorta, the
1532
D. Emerick Szilagyi
Fig. 422. Over-all view of the operative field in a bilateral aortofemoral by-pass. (Reproduced by permission of A.M.A. Arch. f3urg.3)
bilateral aortofemoral by-pass was the operation of choice (Figs. 422 and 423). The prosthesis for this procedure was a plastic bifurcation of a size slightly smaller than the respective host vessels; the levels and types of anastomosis were the same as in placement of the unilateral aortofemoral by-pass. In the femoropopliteal area direct replacement was employed for aneurysmal resection. In occlusive disease the prostheses used were almost exclusively by-passes, virtually all of which extended from the common femoral to the distal popliteal artery, both anastomoses being end-to-side (Fig. 424). The frequent use of bilateral aortofemoral and of the long femoropopliteal by-passes is another circumstance that emphasizes the severity of the test to which the prosthesis was put in this clinical series. Both these operations are rather complex, and they demand a tortuous placement of the implant; moreover, in the femoropopliteal by-passes the length of the prosthesis is great (without exception in excess of 31 em.),
An Elastic Dacron Arterial Substitute
1588
Fig. 423. Angiographic appearance of the bilateral aortofemoral by-pass shown in Figure 422, six weeks postoperatively. (Reproduced by permission of A.M.A. Arch. Surg. 8)
a circumstance that-other factors being equal-is inimical to patency in an arterial replacement. The types of operative procedures involving the use of elastic Dacron prostheses, the length of the prostheses, and the immediate and early Table 2 RESULTS OF GRAFTING PROCEDURES WITH ELASTIC DACRON PROSTHESES IMMEDIATE
Open Aorto-iliac. . . . . . . . . . . . .. 110 (88%) Femoropopliteal. . . . . . . .. 43 (71 %)
EARLY
Closed
Open
Closed
15 (12%) 18 (29%)
110 (88%) 40 (66%)
15 (12%) 21 (34%)
Notes: The term "immediate" signifies the results at discharge; the term "early" the cumulative results up to 24 months. The mean longitudinal dimensions of the aorto-iliac prostheses were 9 cm. and 17 cm. and 18· cm.; the mean length of the femoropopliteal prostheses was 48 cm.
1534
D. Emerick Szilagyi results are listed in Table 2. The postoperative periods of observation in these cases ranged from one to 24 months. Patency of the graft was always verified by angiography. Clinical examinations were made monthly. The results tabulated are cumulative. "Immediate results" denotes the state of the implant at the time of discharge from the hospital; "early results" refers to the condition of the implant at the time of the last angiographic examination up to 24 months postoperatively. The clinical results so far obtained with the use of the elastic Dacron prosthesis are slightly inferior to those we formerly achieved with homografting. A brief reference has been made to some of the reasons for this difference. Perhaps the most important unfavorable factor was the prejudicial effect of case selection, the more advanced and diffuse cases having been assigned for treatment with angioplastic procedures utilizing the prosthesis. It must be borne in mind, however, that insofar as immediate success is concerned, a good homograft is so nearly the ideal arterial substitute that probably no plastic prosthesis will match its performance. The advantage of plastic prostheses consists in the promise that their rate of deterioration after longer follow-up periods will be much less than that of homografts. It this is true, the late patency rate of plastic implants will be better than that of homografts, since many more of the early good results will persist. The assumption that this is what will happen is supported by excellent experimental evidence but will not be proved until the prostheses will have been observed for much longer periods of time. Fig. 424. Angiographic appearance of a femoropoplial by-pass with a 57 cm. prosthesis 6 weeks postoperatively. (The operation was performed after the failure of a homograft that had been in place for 3Y2 years. Note markers at the previous distal anastomosis.) (Reproduced by permission of A.M.A. Arch. Surg. a)
An Elastic Dacron Arterial Substitute
1535
A Comparison of Plastic Arterial Prostheses
It would be of obvious practical importance if one could give an objective and definitive assessment of the comparative merits of the various arterial prostheses now available for clinical use, including the one described in this article. There are, however, serious difficulties in the way of making such an evaluation. In the clinical study of arterial prostheses the common practice currently is to use the early patency rate as the criterion of success. The value of the early patency rate as the standard of performance of a prosthesis has, however, severe restrictions. Obviously, the fact that an arterial implant remains or does not remain open after a grafting operation depends on a number of factors other than the intrinsic properties of the prosthesis, that are highly variable and in clinical practice cannot be controlled. Technical skill and the state of advancement of the existing arterial disease, for instance, have great influence in this regard, and -assuming that the given substitute has reasonably good prosthetic qualities-have probably more to do with early patency than the behavior of the implant proper. Late patency rates-that is, the rate of functioning grafts two years or more after implantation-are much more relevant but still require great care in evaluation, since even at this period in the life of the prosthesis the role of the occlusive process and that of technical factors in causing loss of function are significant. The most troublesome fact, however, is that clinical series of cases using different types of prostheses suitable for meaningful comparison have not been reported. For one thing, the time is too early for the building up of clinical series of sufficient size and sufficiently long follow-up (two years or more); for another, in the available reports, the criteria of case selection are variable and the standards of follow-up investigation vague. All that can be said on the grounds of the reported clinical experience is that at least the four most commonly used prostheses now available (crimped knitted Dacron, crimped woven Teflon, crimped woven Dacron, and smooth elastic woven Dacron) have given satisfaction to their users, and that the early patency rates have ranged from 70 to 90 per cent, depending largely on the anatomical location of the surgical procedures in question and showing slight, if any, correlation with the type of prosthesis. It is thus evident that a firm judgment of the comparative value of these arterial substitutes based on clinical results cannot as yet be made, and that for any such appraisal a different approach must be used. It would seem that such an approach might well consist in a scrutiny of the various arterial prostheses with respect to the degree to which they possess the qualities that by fairly universal agreement are regarded as essential for their long-term function. These qualities have been exhaustively investigated in the laboratory and many experimental facts have been established with regard to them that allow reliable, albeit speculative, conclusions concerning their clinical importance.
1536
D. Emerick Szilagyi
The three qualities that are indispensable for the continued function of a prosthetic arterial implant are the ability to serve as a framework for the building of a new artery, lasting tensile strength, and flexibility. Without the first, the implant will not function at all; without the second, it will function for a short time only; and without the third, its clinical use will be severely limited. Durability is by far the simplest of the three requirements to deal with. The lasting quality of the prosthesis is almost entirely determined by the chemical and physical (particularly hygroscopic) properties of the plastic fiber of which it is made. As briefly stated before, whether the initial tensile strength of the plastic (which in almost all instances is far above the minimum need) will persist depends on its chemical stability as well as lack of tissue reactivity or physicochemical inertness. Among the three plastic materials most widely used, nylon has been found to deteriorate rapidly in tensile strength while both Dacron and Teflon have proved extremely durable. Our own experimental findings have shown Dacron to be practically unchanged in strength after two years of implantation. Teflon, owing to its extremely stable chemical composition, is believed by some to be even more durable. On the grounds of experimental observations, however, there is little to choose between these two from the point of view of the persistence of their tensile strength. The arteriogenetic quality is a much more complex problem. All commonly used arterial prostheses permit the formation of a new connective tissue tube on their framework, but the details and the degree of perfection of this process vary from plastic to plastic. The two important variables that determine the finer details of the process of arteriogenesis are the nature of the fiber in the fabric and the type of textile construction. The chemical nature of the fiber is important since on it depends the nature and intensity of tissue reaction around the implant. This reaction must represent a happy medium between too much and too little. Nylon is not too reactive for good arteriogenesis, but the exudative response it elicits gravely impairs its durability after implantation, as mentioned before. Teflon is almost completely nonreactive. It is this writer's impression that this high degree of nonreactivity is not entirely desirable and that the cohesion of the new intima and adventitia in Teflon prostheses is less than ideal. Dacron appears to possess just the right degree of reactivity to stimulate a good fibroblastic growth without excessive exudative reaction. As regards textile construction, both the type of weave and the size (or denier) of the component yarn seem to have relevance. A taffeta weave (such as ordinary shirting materials have) affords a simple geometric pattern for the trellis upon which the new connective tissue can be laid down; this has the advantage of achieving strength with least bulk. A certain lightness of the yarn is also desirable since this facilitates the selection of the most satisfactory porosity. The use of heavier yarn leads to a fabric that is too highly porous or else to a
An Elastic Dacron Arterial Substitute
.1537
fabric that is too bulky. Among the available Dacron prostheses, the crimped woven prosthesis has the lightest construction, the crimped knitted one is the heaviest and the elastic woven one is intermediate. In our laboratory studies, the most persistently satisfactory arteriogenesis was observed with the use of the last-named prosthesis. In these studies, particularly in regard to the development of the neo-intima, the heavier fabrics often showed imperfections. Flexibility, or stretchability, or expansibility can be achieved by various means: by using a knitted fabric, by crimping the tube, or by incorporating longitudinal yarns with elastic qualities. All these variations in construction produce good flexibility. However, crimping-with or without a knitted pattern-creates an irregular surface that impedes blood flow. Although this decrease of flow is not great, under marginal conditions it may be critical. Crimping also interferes with the growth of neointima, although only to a slight degree. In addition to the three essential requirements enumerated, one may also look into certain minor qualities that have a good deal of practical significance. All presently used plastic prostheses are reasonably inexpensive, easy to store and sterilize. From the point of view of ease of handling, Dacron fabrics can be sutured with somewhat greater facility than Teflon fabrics. For safe use, all porous fabrics must be preclotted, that is, before insertion blood must be rinsed through the tube to allow the formation of fibrin plugs in the pores. Knitted fabrics have the advantage of being able to be trimmed without the need of heat-sealing the cut edges; in woven fabrics the edges must be melted with a cautery blade to prevent unraveling of the yarn in course of suturing. In this writer's opinion both preclotting and heat-sealing are negligible inconveniences. At the moment the preference for one or another of these substitutes is often decided on grounds of minor advantages. In the long-range view, however, the future belongs to that prosthesis which will combine the most nearly perfect arteriogenesis with the greatest durability. SUMMARY
An account is given of the experimental and clinical findings in the use of an arterial substitute, woven of Dacron yarn, which is seamless, smooth-walled, finely porous, light, and has elastic qualities. In animal experiments, durin@; 24 months of observation, the prosthesis showed low tissue reactivity, good arteriogenesis and excellent durability. Used as direct replacement and as by-pass graft in the aorto-iliac and femoropopliteal areas, in 186 clinical cases (the earliest of which has been followed for 24 months) the prosthesis has yielded patency rates only slightly lower than those obtained by us with homografts. The assumption that the late patency rate of these prostheses will be better than that
D. Emerick Szilagyi
1538
of homografts is supported by both theoretical considerations and by experimental evidence, but it will not be proved or disproved until the results will have been followed for much longer periods of time. REFERENCES 1. Szilagyi, D. E., McDonald, R. T., Smith, R. F. and Whitcomb, J. G.: Biologic
Fate of Human Arterial Homografts. A. M. A. Arch. Surg. 75: 506,1957. 2. Szilagyi, D. E., Whitcomb, J. G., Waibel, P. and Schenker, W.: Hemodynamic Factors in Arterial Grafting: An Expermiental Study of Anastomotic Types, Graft Size and Graft Surface Characteristics. Surgical Forum 9: 319, 1959. 3. Szilagyi, D. E., France, L. C., Smith, R. F. and Whitcomb, J. G.: Clinical Use of an Elastic Dacron Prosthesis. A. M. A. Arch. Surg. 77: 538,1958. Henry Ford Hospital Detroit 2, Michigan
UNTIL a few months ago the surgeon had small cause for concern about the choice of the type of arterial substitute he was to use for a grafting operation. The view was generally accepted that a well-processed homologous arterial graft not only looked and behaved like the patient's own undiseased artery but that it would continue indefiqitely to function in the manner of a normal blood vessel. Except for the problems of obtaining and processing them, which were considerable, arterial homografts appeared to be the ideal substitute. Recently, however, as the result of continued postoperative observations of the fate of homologous arterial implants, a radical change has taken place in one's estimate of the value of grafts of this type. It has been learned that the lasting quality of these homotransplants is unsatisfactory.! Histological studies have shown that arterial homografts are not actively accepted and incorporated but only passively tolerated and at times even rejected by the host, and that the implanted grafts must maintain their structural integrity through the mechanical strength inherent in their histological elements. When these wear out, as they inevitably do in time, the wall of the graft weakens, becomes dilated and is eventually occluded by thrombosis, or rarely it ruptures. Since the histologic elements essential for mechanical integrity are the elastic laminae, in aortic grafts-extremely rich in elastic components-the wearing-out process may take many years; indeed, in most cases the graft outlives the patient. Femoral grafts, on the contrary, being very poor in elastic tissue and consisting mostly of smooth muscle and collagen fibers, show structural failure relatively early and in a high incidence. These observations largely invalidated the clinical usefulness of arterial homografts. In spite of their versatility, ease of technical handling and excellent early success rate, the use of these substitutes has been gradually given up. Aside from the recognition of their intrinsic deficiency, which made
The experimental work described was in part supported by a grant/rom the Michigan Heart Association.
* Surgeon-in-Charge, Second General Surgical Division, Henry Ford Hospital, Detroit, Michigan. 1523
1524
D. Emerick Szilagyi
their obsolescence inevitable, the passing of arterial homografts was hastened by the emergence of arterial substitutes constructed of woven yarn of plastic materials of diverse types. Investigation of these prostheses had in fact begun, mostly for reasons of convenience, even before the inadequacy of homografts was established, following the pioneering observation of Voorhees and his associates reported in 1952 that the animal body builds a connective tissue channel around a porous textile fabric placed in the path of the flowing blood. The awareness of the lack of permanence in the structural strength of arterial homografts lent further urgency to this search, and as a result a number of plastic substitutes have been put to clinical use. In what follows I wish to describe the experimental qualities and clinical use of an arterial prosthesis made of Dacron* textile fabric developed in our laboratory. EXPERIMENTAL OBSERVATIONS
In our experimental search for a plastic arterial substitute during the past five years we have kept in sight certain desirable qualities that such a prosthesis must possess in order to be suitable for wide-range clinical use. The arterial prosthesis must be seamless, smooth-walled, easy to handle, durable and relatively nonreactive; its porosity and bulk must meet certain critical requirements; and finally it must possess the ability to elongate and recoil. (Further important but not truly essential requirements for such a prosthesis would be economical cost, ease of storage and simplicity of sterilization.) The reasons for the significance of these attributes are cogent. The presence of a seam and of ridges on the luminal surface of the prosthesis is undesirable since surface irregularities would increase friction and, therefore, resistance to flow;2 moreover they would promote excessive clotting along the areas of contact with the blood, a circumstance that, under many marginal conditions, may mean the difference between patency andthrombotic occlusion. The need for ease of handling (as well as for the less essential factors of low cost and simplicity of storage and sterilization) is self-evident. Durability is also an obvious requirement. It is less obvious that durability is largely dependent on lack of tissue reactivity. Regardless of the magnitude of tensile strength that the original fabric possesses, if after implantation the fabric provokes active cellular reaction in the host tissues, it will be attacked by the exudative process, and in the end, partly or completely destroyed. The optimal degree of porosity in the fabric is the sum of two requirements: the mesh size must be large enough to allow ingress for the ingrowth of connective tissue but it must also be small enough not to interfere with hemostasis. The restrictions on the bulk (or wall thickness) of the prosthesis are less easy to justify. Undoubtedly thick-walled, heavy 'prostheses have·served
* "Dacron" is the registered trade name of E. I. du Pont de Nemours & Company for their polyester fiber.
A n Elastic Dacron Arterial Substitute
1525
Fig. 416. Dacron yarn unstretched (above) and under tension (below). The ability to elongate comes from the uncurling of the fibers when put on stretch.
well in many situations during short-term observations. However, a plastic prosthesis is a foreign body, and sound surgical principles would suggest that its mass should be kept to a minimum as long as this restriction does not result in a disadvantage such as lessening its initial tensile strength or its durability. While conclusive evidence condemning thickwalled prostheses cannot at this time be quoted, it is an easily confirmed observation that the connective tissue reaction around bulky prostheses is excessive, and this is undesirable. The quality of elasticity must be present owing to the necessity for the prosthesis to accommodate itself to the anatomical features of its surroundings. A rigid, nonextensile tube when curved or bent will buckle, and this in turn leads to clotting. It may be also mentioned that the accomplishment of elasticity through specially treated fibers appears to have advantages over methods involving the treatment of the finished tube, namely crimping, which increases bulk and luminal friction. The prosthetic tube we selected for investigation seemed to meet these requirements to a satisfactory degree. This elastic Dacron prosthesis is woven with a taffeta weave and is seamless (Fig. 416). As disposed in the tubular fabric, the longitudinal threads forming the warp are made of fibers that have been specially processed (Fig. 417); the transverse threads, or woof or filling, consist of untreated Dacron fibers. The special process to which the fibers of the warp yarn have been subjected is the same as that used for the manufacture of "Helanca" nylon fiber. In this process the fibers are twisted into tight spirals and then heat-set at 270 0 F. Fibers of opposite directions of twist are combined into the yarn in
D. Emerick Szilagyi
1526
Fig. 417. Simple taffeta weave of the Dacron fabric. It is the fluffy-appearing strands forming the warp (horizontal in this photograph) that have the quality of elasticity.
about equal amounts in order to assure balance. The fibers are stretchable since on pull the spirals uncurl; when the pulling force is released the fibers shorten owing to a tendency to resume their heat-set spiral shape. When yarn composed of such fibers is woven into fabric, this quality of elasticity is largely retained (Fig. 418). It is necessary, however, to boil the fabric in liquid soap to restore complete resiliency. (The contact with soap leaves no deleterious effects provided the fabric is thoroughly rinsed.) The finished tube has an ability to elongate 15 to 20 per cent without deformity. The textile characteristics of the prosthesis used experimentally are given in Table 1. The mesh size and the weight of the yarn were arrived at after many experimental trials and appeared to fulfill optimally the Table 1 TEXTILE CHARACTERISTICS OF ELASTIC DACRON PROSTHESIS
Diameter: 10 mm. Ends (warp): 80 to 160 per centimeter Picks (woof): 47 per centimeter The yarn in the warp: 70-denier, 2-ply, 70-14 Bright Type 51 Dacron The yarn in the woof: 220-denier, 220-50 Bright Type 51 Dacron
An Elastic Dacron Arterial Substitute
1527
Fig. 418. Double exposure of an elastic bifurcation prosthesis showing the range of stretch of the fabric.
basic criterion of minimum bulk with maximum strength. The porosity is such that preclotting is necessary (see below). The canine preparation employed for the experimental investigation of the prosthesis was that first described by McCune and slightly modified by us. The details of the operative technique were given in full in earlier publications. 3 Briefly, the procedure consists in implanting a bypass between the upper thoracic and lower abdominal canine aorta, the proximal anastomosis being an end-to-end one, thus assuring a complete shunting of the aortic blood into the prosthesis. The method permits the insertion of implants 18 to 26 cm. in length and 7 to 10 mm. in diameter (depending on the size of the animal). We believe that this preparation affords a much more realistic test of a prosthesis than the customary short-segment replacements in many of which the bridging effect of the healing process of the stumps of the host aorta is difficult or impossible to distinguish from the histologic changes peculiar to arteriogenesis. Twenty-one such preparations were made, in ten of which the prosthesis was implanted with enough slack to allow a tortuous course, thus)esting the ability of the prosthesis to mold. The experimental animals were investigated by observation of their general condition and state of health; by periodic aortograms; and by scheduled sacrifice. The recovered grafts were inspected for gross appearance, tested for tensile strength and studied for histologic characteristics. All the implants remained patent until the death of the animals. In 12 animals that were sacrificed on schedule after periods of time up to 24 months after implantation, and in five animals that died of intercurrent causes not related to the function of the implants, the recovered prosthesis showed excellent arteriogenesis (Figs. 419 and 420), and a remarkably slight tendency to elicit inflammatory reaction in the host tissues
D. Emerick Szilagyi
1528
Fig. 419. Gross appearance of a canine implant 6 months after insertion. The fabric is completely covered by a smooth, flawless pseudo-intima about 1 mm. in thickness. The connective tissue shell and ingrowth are firmly adherent to the fabric and thin but sturdy. The mean total thickness of the new artery is 1 to 1.5 mm.
r.
Fig. 420. Microscopic section of the midportion of implant shown in Figure 419. Very satisfactory process of arteriogenesis. Cellular exudative reaction is almost completely absent. The growth of connective tissue appears to serve only one purpose: the integration of the plastic fibers into a firm structure. (X70, H. & E. Stain.) (Reproduced by permission of A.M.A. Arch. Surg. 3 )
An Elastic Dacron Arterial Substitute
1529
DURABILITY OF PLASTIC ARTERIAL SUBSTITUTES
He/anca Dacran .c
.~ 50 L!)
ci
"-
,,
c: o=>
\
.;g 40 a.
-
.;:
\
30
C>
c:
Q)
~ 20
\
,, , ,
\
\
\
,, \
\
e-----
--------... ... ~ ...
He/anca Ny/on
---
--.-- ----- - - - ---~--
--... ...........
~-.---r--.--.---r--.---.-~---r--.--.---.-./-,/-.-!~ 2 3 4 5 6 7 8 9 10 II 12 18 24 36
Duration of Implantation (Months)
Fig. 421
(Fig. 420). During 24 months the tensile strength of the implants remained essentially unchanged in marked contrast to the findings with "Helanca" nylon prostheses in which the tensile strength rapidly declined after two months' implantation (Fig. 421). The excellent durability and the remarkably complete angiogenesis observed about the implanted tubes justified the clinical application of the prosthesis. CLINICAL EXPERIENCE
Technical Details in the Use of the Prosthesis
In external appearance, the prosthetic tube of stretch Dacron as used by us is white (or, rather, colorless), soft, pliable, smooth, light and moderately porous. On handling it, one is impressed by the lack of bulkiness that is peculiar to knitted or crimped tubes and the absence of stiffness that is characteristic of Teflon textiles. The prosthesis does not require any special precautions for storage. It can be sterilized chemically, or by boiling and autoclaving. As a sterilizing agent Zephiran Chloride in a 1: 1000 solution for 18 hours of contact has been effective. It should be noted that repeated autoclaving reduces elasticity slightly. Without some appropriate pretreatment, a fabric of the porosity of the elastic Dacron prosthesis will leak blood rather vigorously. We have found that by rinsing the inner and outer surfaces of the prosthesis with 15 to 20 ml. of blood obtained from an unheparinized segment of the exposed arterial tree, and by allowing the prosthesis to soak in this blood for three or four minutes one can bring about just enough clotting to seal off the meshes of the fabric. A moderate amount of oozing may still occur after the prosthesis has been inserted and blood allowed to flow through
1530
D. Emerick Szilagyi
it, but this can be readily controlled by short periods (20 to 30 seconds) of occulsion of the inflow while the blood is allowed to remain in the prosthesis. It is of the utmost importance to use great care in the process of clot-sealing. After the in-vitro clotting, the prosthesis must be checked for the presence of visible or palpable thrombi, and if any are found they must be removed by gentle milking or, if necessary, by rinsing with a heparin-saline solution (0.5 mg. of heparin per 1.0 ml. of saline). The same precaution must be observed during the performance of the insertion of the implant. Obviously, stagnant blood is never allowed to remain in the prosthesis once the clot-sealing has been accomplished. If owing to some technical mishap it becomes necessary to stop the flow of blood through a prosthesis already in place, before the distal clamp is applied the blood is milked out of the implant and the implant is filled with heparin-saline solution until the flow may be re-established. Leaving a layer of thrombus in the lumen of the prosthesis that is thicker than about 0.1 to 0.2 mm. or allowing the presence of lumps of cloth may have disastrous consequences. It may not only delay or prevent the orderly formation of the pseudo-intima, thus resulting in early occlusion, but it may lead to a rapid, almost immediate, progression of the clotting process to a point where the prosthesis cannot function at all. The free cut end of a woven fabric tends to unravel, particularly if the component yam contains smooth-surfaced fibers as is the case with the elastic Dacron prosthesis. This tendency can be easily corrected, however, by melting and fusing the ends of the cut fibers through the application of heat. We have found it convenient to heat-seal the prosthesis before beginning the implantation. The length of the implant is judged by placing a heavy black thread in the exact position the prosthesis is to occupy and by using this measure as the base for the simple calculation. Ten to 15 per cent of the length of the thread (depending on the slightly variable elasticity of the plastic tube) is subtracted to obtain the desired dimension. In case of doubt, care is exercised to make the implant slightly longer rather than shorter than is ideal. The ends of the prosthesis of the desired length are cut with scissors at the angle needed for the proposed anastomosis. The cut ends are lightly touched with a flat-tipped electric cautery which has been heated just short of being red. The ideal heat-seal is a narrow (about 1 mm.) even and smooth band of fused plastic which lies exactly in line with the scissors cut. A light touch and attention to technique are necessary for the proficient use of any plastic prosthesis, and they are of special import when dealing with light woven fabrics. The manual handling and the suturing of the elastic Dacron prosthesis demand great gentleness. The prosthesis must not be picked up with instruments having a rough surface, and clamps applied to it must be rubber-shod. If, nevertheless, separation of some of the threads occurs, the flaw can usually be corrected by gentle
An Elastic Dacron Arterial Substitute
1531
rubbing between fingers. In placing sutures only the slightest strain is allowed on the fabric. Experimental observation reported elsewhere2 have shown that in by-pass grafting the optimal ratio of the size of the prosthesis to the size of the host artery is about 1.6: 1.0, or, roughly, the prosthesis should be half again as large in diameter as the host artery. The technical details enumerated are valid for the straight tubular as well as for the bifurcated prostheses. Because of problems inherent in weaving techniques, the undersurface of bifurcations (the so-called "crotch") is not truly seamless. Although we have found that the seal of the weave at the crotch has been consistently strong, it is advisable to inspect this part of the bifurcation prosthesis with particular attention before use. Clinical Results
In assessing the results of the operations in which elastic Dacron prostheses were used as graft material, certain principles of operative case selection must be kept in mind. During the period in question (June 1957 to May 1959), in dealing with arterial lesions requiring angioplastic correction, we used both grafting procedures (in about 80 per cent of the cases) and thromboendarteriectomy. We reserved thromboendarteriectomy for instances in which the occlusive process did not extend beyond the bifurcation of the common iliac artery or to those cases of femoral lesions in which the disease process was sharply limited to a segment no longer than 5 cm:. It can be seen, therefore, that a factor of selectivity was operative and that it was a factor prejudicial to the procedures utilizing plastic prostheses. In general, it was the more advanced and more diffuse lesions that were treated with the last-named methods. In the aorto-iliac area three types of procedure involving the use of plastic prostheses were employed. Direct replacement of the aorto-iliac segment was resorted to in aneurysmal disease or in cases in which the occlusive process extended for some distance into the aorta. In instances of aortic aneurysm, the entire disease segment was resected and replaced; in occlusive lesions of the aorta, on the other hand, the iliac arteries were not, as a rule, dissected out and replaced. In all operations of direct replacement the proximal anastomosis (prosthesis to aorta) was end-toend, while the distal anastomoses were end-to-side, either at the iliac, the common femoral or popliteal level, according to need. In cases of unilateral iliac occlusive disease, when the occlusive process extended beyond the iliac bifurcation, and thus endarteriectomy was deemed unsuitable, aortofemoral by-passes were used, with the proximal anastomosis at the distal aortic level (distal to the origin of the inferior mesenteric artery), and the distal anastomosis at the common femoral or popliteal level, and with both anastomoses of the end-to-side variety. In extensive bilateral iliac occlusive disease without significant involvement of the aorta, the
1532
D. Emerick Szilagyi
Fig. 422. Over-all view of the operative field in a bilateral aortofemoral by-pass. (Reproduced by permission of A.M.A. Arch. f3urg.3)
bilateral aortofemoral by-pass was the operation of choice (Figs. 422 and 423). The prosthesis for this procedure was a plastic bifurcation of a size slightly smaller than the respective host vessels; the levels and types of anastomosis were the same as in placement of the unilateral aortofemoral by-pass. In the femoropopliteal area direct replacement was employed for aneurysmal resection. In occlusive disease the prostheses used were almost exclusively by-passes, virtually all of which extended from the common femoral to the distal popliteal artery, both anastomoses being end-to-side (Fig. 424). The frequent use of bilateral aortofemoral and of the long femoropopliteal by-passes is another circumstance that emphasizes the severity of the test to which the prosthesis was put in this clinical series. Both these operations are rather complex, and they demand a tortuous placement of the implant; moreover, in the femoropopliteal by-passes the length of the prosthesis is great (without exception in excess of 31 em.),
An Elastic Dacron Arterial Substitute
1588
Fig. 423. Angiographic appearance of the bilateral aortofemoral by-pass shown in Figure 422, six weeks postoperatively. (Reproduced by permission of A.M.A. Arch. Surg. 8)
a circumstance that-other factors being equal-is inimical to patency in an arterial replacement. The types of operative procedures involving the use of elastic Dacron prostheses, the length of the prostheses, and the immediate and early Table 2 RESULTS OF GRAFTING PROCEDURES WITH ELASTIC DACRON PROSTHESES IMMEDIATE
Open Aorto-iliac. . . . . . . . . . . . .. 110 (88%) Femoropopliteal. . . . . . . .. 43 (71 %)
EARLY
Closed
Open
Closed
15 (12%) 18 (29%)
110 (88%) 40 (66%)
15 (12%) 21 (34%)
Notes: The term "immediate" signifies the results at discharge; the term "early" the cumulative results up to 24 months. The mean longitudinal dimensions of the aorto-iliac prostheses were 9 cm. and 17 cm. and 18· cm.; the mean length of the femoropopliteal prostheses was 48 cm.
1534
D. Emerick Szilagyi results are listed in Table 2. The postoperative periods of observation in these cases ranged from one to 24 months. Patency of the graft was always verified by angiography. Clinical examinations were made monthly. The results tabulated are cumulative. "Immediate results" denotes the state of the implant at the time of discharge from the hospital; "early results" refers to the condition of the implant at the time of the last angiographic examination up to 24 months postoperatively. The clinical results so far obtained with the use of the elastic Dacron prosthesis are slightly inferior to those we formerly achieved with homografting. A brief reference has been made to some of the reasons for this difference. Perhaps the most important unfavorable factor was the prejudicial effect of case selection, the more advanced and diffuse cases having been assigned for treatment with angioplastic procedures utilizing the prosthesis. It must be borne in mind, however, that insofar as immediate success is concerned, a good homograft is so nearly the ideal arterial substitute that probably no plastic prosthesis will match its performance. The advantage of plastic prostheses consists in the promise that their rate of deterioration after longer follow-up periods will be much less than that of homografts. It this is true, the late patency rate of plastic implants will be better than that of homografts, since many more of the early good results will persist. The assumption that this is what will happen is supported by excellent experimental evidence but will not be proved until the prostheses will have been observed for much longer periods of time. Fig. 424. Angiographic appearance of a femoropoplial by-pass with a 57 cm. prosthesis 6 weeks postoperatively. (The operation was performed after the failure of a homograft that had been in place for 3Y2 years. Note markers at the previous distal anastomosis.) (Reproduced by permission of A.M.A. Arch. Surg. a)
An Elastic Dacron Arterial Substitute
1535
A Comparison of Plastic Arterial Prostheses
It would be of obvious practical importance if one could give an objective and definitive assessment of the comparative merits of the various arterial prostheses now available for clinical use, including the one described in this article. There are, however, serious difficulties in the way of making such an evaluation. In the clinical study of arterial prostheses the common practice currently is to use the early patency rate as the criterion of success. The value of the early patency rate as the standard of performance of a prosthesis has, however, severe restrictions. Obviously, the fact that an arterial implant remains or does not remain open after a grafting operation depends on a number of factors other than the intrinsic properties of the prosthesis, that are highly variable and in clinical practice cannot be controlled. Technical skill and the state of advancement of the existing arterial disease, for instance, have great influence in this regard, and -assuming that the given substitute has reasonably good prosthetic qualities-have probably more to do with early patency than the behavior of the implant proper. Late patency rates-that is, the rate of functioning grafts two years or more after implantation-are much more relevant but still require great care in evaluation, since even at this period in the life of the prosthesis the role of the occlusive process and that of technical factors in causing loss of function are significant. The most troublesome fact, however, is that clinical series of cases using different types of prostheses suitable for meaningful comparison have not been reported. For one thing, the time is too early for the building up of clinical series of sufficient size and sufficiently long follow-up (two years or more); for another, in the available reports, the criteria of case selection are variable and the standards of follow-up investigation vague. All that can be said on the grounds of the reported clinical experience is that at least the four most commonly used prostheses now available (crimped knitted Dacron, crimped woven Teflon, crimped woven Dacron, and smooth elastic woven Dacron) have given satisfaction to their users, and that the early patency rates have ranged from 70 to 90 per cent, depending largely on the anatomical location of the surgical procedures in question and showing slight, if any, correlation with the type of prosthesis. It is thus evident that a firm judgment of the comparative value of these arterial substitutes based on clinical results cannot as yet be made, and that for any such appraisal a different approach must be used. It would seem that such an approach might well consist in a scrutiny of the various arterial prostheses with respect to the degree to which they possess the qualities that by fairly universal agreement are regarded as essential for their long-term function. These qualities have been exhaustively investigated in the laboratory and many experimental facts have been established with regard to them that allow reliable, albeit speculative, conclusions concerning their clinical importance.
1536
D. Emerick Szilagyi
The three qualities that are indispensable for the continued function of a prosthetic arterial implant are the ability to serve as a framework for the building of a new artery, lasting tensile strength, and flexibility. Without the first, the implant will not function at all; without the second, it will function for a short time only; and without the third, its clinical use will be severely limited. Durability is by far the simplest of the three requirements to deal with. The lasting quality of the prosthesis is almost entirely determined by the chemical and physical (particularly hygroscopic) properties of the plastic fiber of which it is made. As briefly stated before, whether the initial tensile strength of the plastic (which in almost all instances is far above the minimum need) will persist depends on its chemical stability as well as lack of tissue reactivity or physicochemical inertness. Among the three plastic materials most widely used, nylon has been found to deteriorate rapidly in tensile strength while both Dacron and Teflon have proved extremely durable. Our own experimental findings have shown Dacron to be practically unchanged in strength after two years of implantation. Teflon, owing to its extremely stable chemical composition, is believed by some to be even more durable. On the grounds of experimental observations, however, there is little to choose between these two from the point of view of the persistence of their tensile strength. The arteriogenetic quality is a much more complex problem. All commonly used arterial prostheses permit the formation of a new connective tissue tube on their framework, but the details and the degree of perfection of this process vary from plastic to plastic. The two important variables that determine the finer details of the process of arteriogenesis are the nature of the fiber in the fabric and the type of textile construction. The chemical nature of the fiber is important since on it depends the nature and intensity of tissue reaction around the implant. This reaction must represent a happy medium between too much and too little. Nylon is not too reactive for good arteriogenesis, but the exudative response it elicits gravely impairs its durability after implantation, as mentioned before. Teflon is almost completely nonreactive. It is this writer's impression that this high degree of nonreactivity is not entirely desirable and that the cohesion of the new intima and adventitia in Teflon prostheses is less than ideal. Dacron appears to possess just the right degree of reactivity to stimulate a good fibroblastic growth without excessive exudative reaction. As regards textile construction, both the type of weave and the size (or denier) of the component yarn seem to have relevance. A taffeta weave (such as ordinary shirting materials have) affords a simple geometric pattern for the trellis upon which the new connective tissue can be laid down; this has the advantage of achieving strength with least bulk. A certain lightness of the yarn is also desirable since this facilitates the selection of the most satisfactory porosity. The use of heavier yarn leads to a fabric that is too highly porous or else to a
An Elastic Dacron Arterial Substitute
.1537
fabric that is too bulky. Among the available Dacron prostheses, the crimped woven prosthesis has the lightest construction, the crimped knitted one is the heaviest and the elastic woven one is intermediate. In our laboratory studies, the most persistently satisfactory arteriogenesis was observed with the use of the last-named prosthesis. In these studies, particularly in regard to the development of the neo-intima, the heavier fabrics often showed imperfections. Flexibility, or stretchability, or expansibility can be achieved by various means: by using a knitted fabric, by crimping the tube, or by incorporating longitudinal yarns with elastic qualities. All these variations in construction produce good flexibility. However, crimping-with or without a knitted pattern-creates an irregular surface that impedes blood flow. Although this decrease of flow is not great, under marginal conditions it may be critical. Crimping also interferes with the growth of neointima, although only to a slight degree. In addition to the three essential requirements enumerated, one may also look into certain minor qualities that have a good deal of practical significance. All presently used plastic prostheses are reasonably inexpensive, easy to store and sterilize. From the point of view of ease of handling, Dacron fabrics can be sutured with somewhat greater facility than Teflon fabrics. For safe use, all porous fabrics must be preclotted, that is, before insertion blood must be rinsed through the tube to allow the formation of fibrin plugs in the pores. Knitted fabrics have the advantage of being able to be trimmed without the need of heat-sealing the cut edges; in woven fabrics the edges must be melted with a cautery blade to prevent unraveling of the yarn in course of suturing. In this writer's opinion both preclotting and heat-sealing are negligible inconveniences. At the moment the preference for one or another of these substitutes is often decided on grounds of minor advantages. In the long-range view, however, the future belongs to that prosthesis which will combine the most nearly perfect arteriogenesis with the greatest durability. SUMMARY
An account is given of the experimental and clinical findings in the use of an arterial substitute, woven of Dacron yarn, which is seamless, smooth-walled, finely porous, light, and has elastic qualities. In animal experiments, durin@; 24 months of observation, the prosthesis showed low tissue reactivity, good arteriogenesis and excellent durability. Used as direct replacement and as by-pass graft in the aorto-iliac and femoropopliteal areas, in 186 clinical cases (the earliest of which has been followed for 24 months) the prosthesis has yielded patency rates only slightly lower than those obtained by us with homografts. The assumption that the late patency rate of these prostheses will be better than that
D. Emerick Szilagyi
1538
of homografts is supported by both theoretical considerations and by experimental evidence, but it will not be proved or disproved until the results will have been followed for much longer periods of time. REFERENCES 1. Szilagyi, D. E., McDonald, R. T., Smith, R. F. and Whitcomb, J. G.: Biologic
Fate of Human Arterial Homografts. A. M. A. Arch. Surg. 75: 506,1957. 2. Szilagyi, D. E., Whitcomb, J. G., Waibel, P. and Schenker, W.: Hemodynamic Factors in Arterial Grafting: An Expermiental Study of Anastomotic Types, Graft Size and Graft Surface Characteristics. Surgical Forum 9: 319, 1959. 3. Szilagyi, D. E., France, L. C., Smith, R. F. and Whitcomb, J. G.: Clinical Use of an Elastic Dacron Prosthesis. A. M. A. Arch. Surg. 77: 538,1958. Henry Ford Hospital Detroit 2, Michigan