Spatial and temporal changes in compliance following implantation of bioresorbable vascular grafts Howard P. GreisIer* and Kathleen A. Joyce Loyola University Medical Center, Department of Surgery, Maywood, Illinois 60153 Dae Un Kim St. Barnabas Medical Center, Department of Pathology, Livingsfon, New Jersey 07039 Si M. Pham, Scott A. Berceli, and Harvey S. Borovetz University of Pittsburgh, Department of Surgery, Pittsburgh, Pennsylvania Compliance matching between the host vessel and vascular grafts used for smalldiameter arterial replacements is thought to be important for long-term patency. However, currently available grafts elicit fibroplastic reactions, resulting i n decreasing compliance with time after implantation. Bioresorbable prostheses elicit ingrowth of myofibroblasts containing abundant contractile elements. This led us to investigate whether compliance of implanted bioresorbable prostheses decreased as a function of time and if the kinetics of change correlated with the progession of tissue ingrowth. Woven polyglactin 910 prostheses (10 mm X 4 mm i.d.) were implanted into adult NZW rabbit infrarenal aortas and replicates were harvested serially through 8 months. Control grafts were implanted, and immediately resected. Dynamic compliance was measured a t 1-mm axial increments along each explant using a pulse duplicator apparatus which exposed the harvested samples to realistic pulsatile hemodynamics. Compliance was calculated for proximal, mid, and distal seg-

ments of each graft and averaged at each time point by grouping into control (zero time, n = 3), early (1-4 weeks, n = 13), and late (6-36 weeks, n = 9) explant periods. At late explant periods both proximal and distal compliance were significantly greater than mid graft compliance (p < .02 and p < .03, respectively). There was a significant increase in proximal compliance between early and late explant times (p < .01). Measured increases in mid and distal segment compliance over time did not reach statistical significance. Myofibroblast laden tissue i n g r o w t h i n t o the inner capsule followed macrophage phagocytosis and was nearly complete prior to the time that an increase in compliance was demonstrated. Thus since the major histologic episodes precede the change in compliance, these are not likely initiated by this biomechanical change. We hypothesize the graft resorption coupled with the ingrowth of more compliant tissue likely leads to the increased compliance of the graft material. 0 1992 John Wiley & Sons, Inc.

Supported by grants from the National Institutes of Health, RO1 HL41272 (H.P. Greisler) and RO1 HL34739 (H.S. Borovetz). *To whom correspondence should be addressed. Journal of Biomedical Materials Research, Vol. 26, 1449-1461 (1992) 0 1992 John Wiley & Sons, Inc. CCC 0021-9304/92/111449-13

GREISLER ET AL.

1450 hhhhhhINTRODUCTION

Bioresorbable lactide/glycolide copolymeric vascular grafts implanted into several animal models elicit an extensive transinterstitial ingrowth of capillaries with endothelial cells, and of myofibroblasts resulting in the replacement of the resorbable material by a tissue conduit. The rate of tissue ingrowth parallels the kinetics of macrophage mediated resorption of the material.’” Following the observation of macrophage phagocytosis, inner capsule thickening develops rapidly and stabilizes at the time of total prosthetic resorption. These tissues are comprised primarily of smooth musclelike myofibroblasts beneath the endothelial blood-contacting surface. The mitotic index of the myofibroblast population similarly parallels the kinetics of prosthetic resorption. This cell proliferation contributes heavily to the tissue ingrowth. Ultrastructurally the great majority of these myofibroblasts are contractile in phenotype with abundant myofilaments with dense bodies throughout their cytoplasm. This is in contradistinction to the more fibroplastic reaction elicited by Dacron prostheses when implanted into the same animal models. Polyglactin 910 (PG910) is totally resorbed over a 2- to 3-month period primarily by macrophage mediated hydrolysis. This material has been shown previously to elicit the tissue reactions described above.6 In related cell culture studies, PG910 was added to cultures of rabbit peritoneal macrophages. The conditioned media of these flasks contained significantly more mitogenic activity capable of stimulating thymidine incorporation into DNA in previously quiescent rabbit aortic smooth muscle cells, BALB/c 3T3 fibroblasts, and murine capillary lung endothelial cells as compared to cultures of macrophages exposed either to Dacron or to neither graft material.7f8Among the substances released by these bioresorbable polymer-stimulated macrophages was basic fibroblast growth factor (basic FGF). The purpose of this current study was twofold. First, all currently available small diameter vascular grafts have shown a decrease in dynamic compliance as a function of time following implantation, due presumably to the fibroplastic reactions they elicit. It was hypothesized that the bioresorbable graft might not share this fate in view of the resorption of the foreign material as well as the more contractile phenotype of the inner capsular myofibroblasts. The current study evaluated the change in compliance of implanted bioresorbable vascular grafts as a function of time. Second, the extensive transinterstitial ingrowth of tissue could theoretically be initiated either by a macrophage derived bioactive substance such as a chemoattractant and/or growth factor, or conversely by a prior change in prosthetic biomechanical characteristics following implantation. In previous work we documented that at the time of implantation PG910 prostheses had compliance characteristics resembling similarly woven Dacron prostheses. We wished to determine in the present study whether the more extensive tissue ingrowth elicited by PG910 prostheses preceded or followed any demonstrable change in dynamic compliance. Such information would provide further insight as to whether initiation of this tissue ingrowth was mechanically or biochemically mediated.

COMPLIANCE CHANGES IN BIORESORBABLE GRAFTS

1451

MATERIALS A N D METHODS

Twenty-five adult female New Zealand White (NZW) rabbits weighing 3-4 kg were anesthetized with intravenous 2% methohexital sodium. Woven polyglactin 910 (Vicryl, Ethicon, Somerville, NJ) prostheses (10 mm X 4 mm i.d.) were implanted into infrarenal aortas using a double end-to-end technique. Prostheses were implanted following preclotting and using continuous 7-0 polypropylene sutures for the anastomoses. Postoperatively, the animals were provided a standard diet and routine care by staff veterinarians until they were sacrificed at specific time intervals between 1 and 36 weeks following surgery. Prior to sacrifice, the prostheses and/or regenerated blood vessels were then resected from the still-living anesthetized animals. Control grafts were implanted following preclotting, and the grafts immediately resected. The explants were grouped into three different time periods: (a) Control time (n = 3), grafts that were implanted and immediately explanted; (b) early explants (n = 13), those grafts which were explanted between 1 and 4 weeks following surgery; and (c) late explants (n = 9), those grafts which were explanted beyond 1 month (6 to 36 weeks) following surgery. These time periods were chosen based on histologic results of PG910 implantation from our previous studies. The control time was used to determine the compliance of the graft material at the time of implantation. The early period of 1to 4 weeks corresponds to macrophage recruitment and myofibroblast proliferation. The late period encompasses the period of complete graft resorption, when an arterial-like conduit has been established. In addition, the normal infrarenal aortas from three anesthetized and previously unoperated NZW rabbits was resected for similar compliance determinations of the normal rabbit aorta. Pulse duplicator apparatus and calculation of graft compliance An in vitra pulsatile perfusion apparatus was utilized in these experiments to simulate realistic arterial hemodynamic~.~ Graft segments and proximal and distal segments of native artery were dissected free of surrounding fibrotic tissue and transferred to the perfusion apparatus utilizing a specially designed set of parallel vascular clamps (American V. Mueller, Chicago, IL; DeBakey Coarctation Clamps, #CH 7110). These clamps served to maintain physiologic longitudinal strain upon excision of the vessel and placement in the perfusion system. The perfusion apparatus provided pulsatile flow of physiologic saline through the grafts at pressures of 67 +- 6 mm Hg systolic and 41 ? 5 mm Hg diastolic, mean flow of 70 mL/min and frequency of 1 Hz. Selected values for perfusion pressure and flow rate were chosen to match measurements made in v i m on anesthetized NZW rabbits. For convenience all experiments were performed at room temperature (25OC). Measurement and calculation of the dynamic mechanical and rheologic properties of mammalian vessels depend upon a number of biologic and physical factors." A review of the literature indicates that Bergel" adhered to

GREISLER ET AL.

1452

the format of classical elasticity theory in calculating the elastic moduli of excised vessel segments. Another convenient way of expressing vessel wall elasticity utilizes the pressure-strain elastic modulus, E which is expressed as1’

EP = (ps - pd)/(Ds - Dd) x Dd (1) Here D represents diameter, P pressure, and the subscripts s and d correspond to end-systole and end-diastole, respectively. The expression (Ps - Pd)/ ( D , - Dd) is the ratio of the increment in pressure to its associated increment in radius. Equation (1) is particularly useful if the Poisson ratio and vessel wall thickness are unknown, since both quantities are required for calculation of the classic elasticity modulus according to the formulation of Bergel. In the present work the inverse of Ep, which can be thought of the dynamic vessel compliance, Cdynis used to quantitate the pressure-radius relationship of the explanted PG910 material, e.g., (Ds

Dd)/{Dd x (ps -

(2) Cdynwhich can also be viewed as a measure of the overall circumferential rigidity of the PG910 material,13 is a commonly used clinical parameter for assessing vascular compliance. Its use in this study permits direct comparison with published reports for natural and artificial conduits. Figure 1 depicts the experimental protocol for determining Cdynin this study. Intraluminal pulsatile pressure was followed both proximal and distal to the PG910 segment using microtip pressure catheters (Millar Model PC360, Millar Corp., Houston, TX). The external diameter of the PG910 segment was continuously monitored using a helium-neon scanning laser micrometer (Model 50-03, Laser Mike Inc., Dayton, OH) whose scanning range (diameter between 0.254 and 50 mm) and overall system accuracy (0.0124 mm) Diameter meamake it particularly well suited for use in this app1i~ation.l~ surements were made (in duplicate) at l mm (20.01 mm) longitudinal increments as determined with a rapid advance positioner (Uni Slide Series A2500, Velmex Inc., E. Bloomfield Holcomb, NY) connected to a digital readout Cdyn

=

-

Prox

0

Mid

3

pd))

Dist

7

10mm

Figure 1. Spatial distribution of measurement sites along proximal, middle, and distal segments of PG910 explants. Compliance measurements were performed at 1-mm axial increments (not within 1 mm of either anastomosis). During perfusion ex vivo the graft-host segment was bathed externally in lactated Ringer’s solution.

COMPLIANCE CHANGES IN BIORESORBABLE GRAFTS

1453

(ACU-RITE Quik Count 11, Bausch & Lomb, Rochester, NY). Analog pressure and laser voltages were digitized through a 16-channel A/D converter interfaced with an IBM AT computer. Five seconds of analog data were acquired (at a sampling frequency of 64 Hz) for each of the longitudinal sites depicted in Figure 1. Dynamic measurements of external diameter were converted to internal diameter following previously described protoc01s.’~Each graft segment was sectioned into 1-mm increments and internal and external diameters determined in the unstrained (e.g., undistended) position using the helium-neon laser micrometer. These measurements, in combination with assumption of constant tissue density, served as input for calculation of internal diameter in the strained configuration. Accordingly the data for D, and Dd in Eq. (2) correspond to the calculated internal diameters at end-systole and end-diastole, respectively. In our opinion internal diameter is the more appropriate parameter to utilize in calculations of Cdynsince it is independent of the variable thickness of the PG910 tissue capsule which remains on explant. Values for Ps and p d in Eq. (2) represent the arithmetic mean of the data from the proximal and distal measurement sites. Histologic correlation

Because of histologic alterations potentially encountered following ex vivo perfusion of explanted grafts, histologic correlation was made using identical historical Data analysis For purposes of data analysis the calculation of Cdynfor each explant were divided into three spatial elements as displayed in Figure 1, namely proximal, middle, and distal segments of the graft. Any measurement which was made within 3 mm of the suture line (not including the anastomoses) was considered to be either proximal or distal. All other sites of compliance measurement were considered as the middle portion of the graft. The individual data points for the proximal section were then averaged as were those for the mid and distal section to obtain a total of three compliance values per each explanted PG910 graft, e.g., one proximal compliance value, one middle and one distal graft dynamic compliance. A two-way Analysis of Variance was then run following logarithmic transformation of the data to compare differences in Cdynamong the proximal, mid, and distal sections at each time period as well as to compare differences between early and late explant times for each of the three locations on the explanted specimens. A Bonferroni corrected Student’s t test was then used to determine differences between specific groups. Statistical significance was assumed for P values of less than 0.05. All analyses were performed using programs 3D and 7D of the BMDP statistical software (BMDP Statistical Software, Los Angeles, CA).

GREISLER ET AL.

1454 RESULTS

All 25 explanted prostheses were patent and without aneurysmal dilatation or significant stenosis. The mean values of compliance for the proximal, mid and distal segments in the control early and late groups are listed in Table 1, and shown in Figure 2. Analysis of variance demonstrated statistical significance in dynamic compliance as a function of both time and position. For the control grafts, there were no significant differences between proximal, mid and distal sections. The value of Cdynfor native PG910 is similar to that for Dacron (2.0) which has been measured in other studie~.'~ Grafts explanted between 1 and 4 weeks following surgery, which is the period of macrophage recruitment and infiltration and cellular proliferation, did not demonstrate differences in Cdynfrom the controls. However, in the late explant period (the period of complete graft resorption) compliance in the proximal and distal TABLE I The Compliance of PG910 Grafts (% Radial Change/mm Hg Control 4.0

Proximal Mid Distal

r;'

? 1.3 2.2 +- 0.8 3.6 +- 1.4

X

lo-')

Early

Late

3.4 rt 0.7 2.7 f 0.2 5.1 ? 1.7

7.4 ? 1.3" * * 3.4 f 0.4 10.3 -C 2.7""

0Proximal Middle

X Heavy Tissue lngrowth (Weeks 2 4

I

Macrophage Infiltration

Native

1

2

3-4

6-9

13-16

30-36

Implant Duration (Weeks) Figure 2. Bar graph showing radial compliance as a function of time for proximal, middle, and distal segments of PG910 explants (mean t standard error). The implant durations between 1 week and 4 weeks correspond to the "early" period in Table I. Implants extending beyond one month (e.g., 6 weeks to 36 weeks) are included in the "late" period in Table I.

COMPLIANCE CHANGES IN BIORESORBABLE GRAFTS

1455

graft segments increased significantly compared to controls ( p < 0.02). Interestingly, the compliance in the graft midportion did not significantly vary as a function of time. However, at the late explant times, the values for Cdynin the proximal and distal segments were elevated versus the graft mid segment ( p < .02 and p < .03, respectively). For comparison compliance measurements of the normal rabbit aortas averaged 47 ? 3 which far exceed all of the data presented in Table I and Figure 2 for the PG910 explants. Figure 3 shows representative histologic sections of a PG910 prosthesis explanted after 1 month from a rabbit aorta. The prosthetic interstices are infiltrated with macrophages and giant cells and the relatively thick inner capsule consists predominantly of longitudinally and circumferentially oriented myo-

Figure 3(a). Photomicrograph of midportion of a 1-month PG910 explant showing inner capsule containing numerous myofibroblasts, collagen, and extracellular matrix beneath an endothelialized blood-contacting surface. PG910 graft material is seen below this inner capsule (hematoxylin & eosin, original magnification X340).

GREISLER ET AL

1456

Figure 3(b). Transmission electron micrograph of a 3-month PG910 explant showing a myofibroblast containing abundant myofilaments with dense bodies (arrows) and rough endoplasmic reticulum (R) within the cytoplasm (original magnification X11,250).

fibroblasts amidst a collagen-rich matrix. Immunohistochemical staining for smooth muscle specific alpha-actin is positive in the cytoplasm of the inner capsular myofibroblasts. DISCUSSION

Before considering the physiologic implications of the present results, certain questions relating to procedural accuracy should be addressed. As described above, C d y n is derived from measurements of pulsating outside diameter made in a noncontacting fashion using a scanning laser micrometer. The accuracy of this instrument is 0.0125 mm. The measurement accuracy for transmural pressure in our perfusion apparatus is k0.87 mm Hg. If these values of experimental uncertainty are input into the statistical expression for error pr~pagation,'~ the accuracies of calculated parameters, e.g., C dyn, are readily obtained.16In particular the percent standard deviation for C dyn associated with instrumentation error is 9.4%. Other sources of variability include biologic variability in animal response and differences in specimen preparation and harvest. While we cannot accurately quantify the percent contribution of these factors, we note that the variability in Table I and Figure 2 is consistent with previous vascular biomechanics measurements by our group for ePTFE arterial conduits in canine model^.'^ Currently available small-diameter vascular prostheses have generally been found to be less efficacious than autogenous tissues when used in distal

COMPLIANCE CHANGES 1N BIORESORBABLE GRAFTS

1457

bypass procedures. Among the most common failure modes for implanted small-diameter grafts is anastomotic pseudointimal hyperplasia. Several investigators have postulated that the mismatch in compliance between native artery and graft contributes to this hyperplastic resp~nse.’*~~~ Clinical studies have shown a direct correlation between graft compliance and long-term paten~y,’~,’’those grafts with lower compliance (and thus a larger compliance mismatch with the native artery) having the higher failure rate. Moreover, the mismatch becomes even more pronounced over time, since the compliance of Dacron and ePTFE implants decreases as a function of time following implantation while a para-anastomotic hypercompliant zone develops on the native artery side of the anastomosis?’ Dacron and ePTFE prostheses elicit a highly fibroplastic response when implanted in man consisting largely of fibroblast deposition primarily in the outer capsules along with collagen within both inner and outer capsules. The compliance of the implanted prosthesis tends to decrease following the development of this reactive fibroplasia. By contrast the compliance of the midsegment of implanted lactide/glycolide bioresorbable PG910 prostheses in the current study showed no such decrease. Measured dynamic compliance actually rose from 2.2 -t 0.8 to 3.4 It_ 0.7 (p > 0.05). This avoidance of a decrease in radial compliance as a function of time is likely a result of both prosthetic resorption and the nature of the regenerating tissues replacing the resorbed material. PG910 prostheses are infiltrated and phagocytized by macrophages and giant cells as early as 2 weeks following their implantation with a subsequent ingrowth of myofibroblasts heavily laden with abundant myofilaments with dense bodies throughout their cytoplasm, immunoreacting with an antismooth-muscle specific alpha-actin antibody, and ultrastructurally resembling smooth muscle cells (Fig. 3). By contrast, Dacron prostheses implanted into the same rabbit model elicit a significantly attenuated myofibroblast response with those myofibroblasts appearing more synthetic in nature, ultrastructurally resembling the phenotype of fibroblasts. In addition the retained Dacron serves functionally to limit potential increases in dynamic radial compliance. Myofibroblast infiltration into the inner capsule within implanted PG910 prostheses occurs primarily between weeks 2 and 4 as evidenced by both inner capsule thickness measurements and myofibroblast mitotic index determination~.’~ Of note is the time course of this histologic change vis-a-vis the kinetics of compliance alteration, both in para-anastomotic (e.g., proximal/ distal) and central zones. While no significant change in Cdynwas measured as a function of time in the middle segment, the proximal and distal segments yielded significant increases over time. However, these compliance changes followed rather than preceded the period of maximal tissue ingrowth, i.e., weeks 2 to 4. Thus it is highly unlikely that biomechanical factors per se were solely responsible for the initiation of the extensive tissue ingrowth elicited by bioresorbable prostheses. It is more likely that other mechanisms such as macrophage activation with subsequent release of growth factors induced by macrophage/biomaterial interactions were respon-

1458

GREISLER ET AL.

sible for the initiation of the tissue ingrowth which may then be modulated over time by biomechanic alterations. Activated macrophages are capable of producing growth factors which may regulate mesenchymal cell proliferation including PDGF, TGF-P, IL-1, IL-6, basic FGF, TNF-a, and y-interfer~n.~’-~~ In related cell culture studies, we have shown the greater ability of lactide/glycolide copolymers to induce macrophages to release growth factors, including basic-FGF, capable of stimulating DNA synthesis in cultured murine capillary lung endothelial cells, human umbilical vein endothelial cells, BALB/c 3T3 mouse fibroblasts, and rabbit aortic smooth muscle cells.7~x~28~2y By contrast, exposure of the same macrophage population to Dacron resulted in an inhibition of endothelial cell proliferation and a diminution in the stimulation of fibroblast and smooth muscle cell proliferation. Analysis of conditioned media groups utilizing both immunoprecipitation with neutralizing antibody techniques as well as SDS PAGE and Western blotting have shown the presence of basic FGF in the media from the macrophages in the presence of biomaterials but not in their absence. It thus appears that biomaterials may differentially activate macrophages to release growth promoting substances including basic FGF which may induce the high mitotic index within the myofibroblast population demonstrated in z l i ~ o . ~ A compliance mismatch between artery and graft evolves when a stiff segment is interposed into a more elastic, compliant segment. The compliance of a vessel contributes to the impedance of the vascular tree and ultimately to ventricular work and cardiac output. When a rigid segment is interposed into a compliant system, there is a partial loss of the pulsatile energy component of left ventricular output, and an increase in wave refle~tion.~’ Complex flow patterns result, concomitant with an increase in wall vibration and oscillations of the adjacent artery, and shearing at the an as tor nose^.^^^^' These factors may result in endothelial damage and/or para-anastomotic smooth muscle cell pr~liferation.~’-~~ Injured endothelium may release PDGF or may be the site of platelet deposition and/or macrophage infiltration either of which may result in activation with subsequent mitogen release stimulating smooth muscle cell migration into the intima and proliferation with the formation of pseudointimal hyperplasia. Previous work by the authors utilized bicomponent prostheses containing yarns composed of both PG910 and polypropy1ene.36Two groups of prostheses were utilized with the polypropylene component differing in elasticity. Following resorption of the PG910, the polypropylene then became the major influence on graft compliance. We found that the cellularity and thickness of the inner capsules were significantly greater (and the patency rate significantly lower) in the group with the more elastic polypropylene c ~ m p o n e n t In .~~ light of our current results these complementary studies suggest further that during the period of prosthetic resorption, any measurable increase in compliance may further stimulate myofibroblast proliferation. In the proximal and distal segments of the explanted PG910 prostheses there is a large increase in compliance compared to the middle segment and

COMPLIANCE CHANGES IN BIORESORBABLE GRAFTS

1459

compared to the proximal and distal segment compliance at the earlier time periods. This resembles the formation of a para-anastomotic hypercompliant zone similar to that described by Abbott et al.’9,2’on the native artery side of the anastomosis. Thus for PG910, a para-anastomotic hypercompliant zone likely occurs at both sides of the anastomosis. The mechanisms behind the para-anastomotic hypercompliant zone formation for PG910 implants are unclear. In the area of the anastomosis, flow patterns and intramural tensile stress development are known to be disturbed leading to increased deformation of the tissues, increased plasma insudation and stimulated monocyte recruitment and presumably all the concomitant effects of both mitogen release and release of proteolytic enzymes including macrophage derived collagenase and elastase. Bioresorbable lactide/glycolide copolymeric vascular prostheses appear to be unique among biomaterials in their lack of decrease in wall compliance as a function of time following implantation. The increase in compliance in the proximal and distal regions of the PG910 segment corresponds with the previously described para-anastomotic hypercompliant zone and likely serves to lessen the compliance mismatch found at the anastomosis when nonresorbable vascular prostheses are implanted. The extensive infiltration of myofibroblasts into the inner capsules within lactide/glycolide copolymeric prostheses and the high mitotic index (20.1 2 16.6 at 3 weeks‘) in this myofibroblast population precedes any demonstrable change in radial compliance and thus is unlikely initiated by such biomechanical alterations. The authors wish to acknowledge the assistance of Jacqueline D. Garfield in the preparation of electron microscopic material and Delores Breen in the preparation of the photomicrographs.

References 1. H. P. Greisler, ‘Arterial regeneration over absorbable prostheses,” Arch. Surg., 117, 1425-1431 (1982). 2. H. P. Greisler, D.U. Kim, C. Fenoglio, J. B. Price, and A. B. Voorhees,

3.

4.

5.

6.

’Arterial regenerative activity after prosthetic implantation,” Arch. Surg., 120, 315-323 (1985). H. P. Greisler, J. Ellinger, T. H. Schwarcz, J. Golan, R. M. Raymond, and D. U. Kim, ‘Arterial regeneration over polydioxanone prostheses in the rabbit,” Arch. Surg., 122, 715-721 (1987). H. P. Greisler, D.U. Kim, J.W. Dennis, J. J. Klosak, K. A. Widerborg, E. D. Endean, R. M. Raymond, and J. Ellinger, ”Compound polyglactin 9lO/polypropyIene small vessel prostheses,” J. Vasc. Surg., 5, 572-583 (1987). H. P. Greisler, “Macrophage activation in bioresorbable vascular grafts,” in Vascular endothelium and physiological basis of clinical problems, J.D. Catravas, A.D. Callow, C.N. Gillis, and U. Ryan (eds.), Plenum Publishing, New York, 1991, pp. 253-254. H. P. Greisler, J.W. Dennis, E. D. Endean, and D.U. Kim, ”Derivation of neointima of vascular grafts,” Circulation, Suppl. I, 78, 16-112 (1988).

1460

GREISLER ET AL. 7. 8.

9. 10. 13. 12. 13.

14. 15.

16.

17.

18. 19.

20. 21. 22. 23.

24. 25.

26. 27. 28.

H.P. Greisler, J.W. Dennis, E.D. Endean, J. Ellinger, R. Friesel, and W. Burgess. “Macrophage/biomaterial interactions: The stimulation of endothelialization,” J. Vasc. Surg., 9, 588-593 (1989). H. P.Greisler, J. Ellinger, S.C. Henderson, A.M. Shaheen, W. H. Burgess, D.U. Kim, and T. M. Lam, “The effects of an atherogenic diet on macrophage/biomaterial interactions,” 1. Vasc. Surg., 14, 10-23 (1991). A.M. Brant, J. F. Chmielewski, T.-K. Hung, and H.S. Borovetz, “Simulation in vitro of pulsatile vascular hemodynamics using a CAD/CAM designed cam disc and roller follower,” Artif. Organs, 10,419-421 (1986). C . M. Buntin and F. H. Silver, “Noninvasive assessment of mechanical properties of peripheral arteries,” Ann. Biomed. Eng., 18,549-566 (1990). D.H. Bergel, “The static elastic properties of the arterial wall,” J. Physiol. London, 156, 445-457 (1961). B. S. Cow and M.G. Taylor, “Measurement of viscoelastic properties of arteries in the living dog,” Circ. Res., 23, 111-122 (1968). R. Walden, G. J. L‘Italien, J. Megerman, and W. M. Abbott, ”Matched elastic properties and successful arterial grafting,” Arch. Surg., 115, 1166-1169 (1980). A. M. Brant, V.G. J. Rodgers, and H. S. Borovetz, ”Measurement in vitro of pulsatile arterial diameter using a helium-neon laser,” 1. Appl. Physiol., 62, 679-683 (1987). A. C.Melissinos, Experiments in modern physics. Academic Press, New York, 1966, pp. 469-473. V.G. J. Rodgers, M. F. Teodori, and H. S. Borovetz, ”Experimental determination of mechanical shear stress about a n anastomotic junction,” J. Biomechanics, 8, 795-803 (1987). V.G. J. Rodgers, M. F. Teodori, A. M. Brant, and H. S. Borovetz, ”Characterization in vitro of the biomechanics properties of anastomosed host artery-graft combinations,” J. Vasc. Surg., 4, 396-402 (1986). R. N. Baird and W. M. Abbott, “Pulsatile blood-flow in arterial grafts,” Lancet, 30, 948-950 (1976). J. E. Hasson, J. Megerman, and W. M. Abbott, ”Increased compliance near vascular anastomoses,” 1. Vasc. Surg., 2, 419-423 (1985). I.G. Kidson and W. M. Abbott, “Low compliance and arterial graft occlusion,” Circulation Suppl. I, 11-14 (1978). W.M. Abbott, J. Megerman, J.E. Hasson, G. J. L‘Italien, and D. J. Warnock, “Effect of compliance mismatch on vascular graft patency,” J. Vasc. Surg., 5(2), 376-382 (1987). R. Shimokado, E.W. Raines, D.K. Madtes, T.B. Barrett, E.P. Benditt, and R. Ross, “A significant part of macrophage-derived growth factor consists of at least two forms of PDGF,” Cell, 43, 277 (1985). A. Baird, I? Mormede, and P. Bohlen, ”Immunoreactive fibroblast growth factor in cells of peritoneal exudate suggests its identity with macrophage-derived growth factor,” Biochem. Biophys. Res. Comm., 126, 358 (1985). J. A . Schmidt, S. B. Mizel, D. Cohen, and I. Green, ”Interleukin 1, a potential regulator of fibroblast proliferation,” 1. lmmunol., 128, 2177 (1982). G. Tosato, K. B. Seamon, N. D. Goldman, P.B. Sehgal, L.T. May, G.C. Washington, K. D. Jones, and S. A. Pike, “Monocyte-derived human Bcell growth factor identified as interferon-p2 (BSF-2, IL-6),” Science, 239, 502 (1988). R.K. Assoian, A. Komoriya, C.A. Meyers, D.M. Miller, and M.B. Sporn, “Transforming growth factor-beta in human platelets,” J. Biol. Chem., 24(58), 7155-7160 (1983). L. J. Old, “Tumor necrosis factor (TNF),” Science, 230, 630-632 (1985). H. P. Greisler, “Growth factor mediation of healing of vascular grafts,” in Cardiovascular Science and Technology: Basic &?Applied, I1 Precised Proceedings, 1990, pp. 55-57.

COMPLIANCE CHANGES IN BIORESORBABLE GRAFTS 29.

30. 31. 32. 33. 34. 35. 36.

1461

H. P. Greisler, J. Ellinger, S.C. Henderson, A. M. Shaheen, W. H. Burgess, and T.M. Lam, ”The effects of an atherogenic diet on macrophage/ biomaterial interactions,” I. Vasc. Surg., 14, 10-23 (1991). M. F. O’Rourke, “Steady and pulsatile energy losses in the systemic circulation under normal conditions and in simulated arterial disease,” Cardiovasc. Res., 1, 313-326 (1967). V.G. J. Rodgers, M. F. Teodori, A.M. Brant, and H.S. Borovetz, “Characterization in vitro of the biomechanical properties of anastomosed host artery graft combinations,” J. Vasc. Surg., 4, 396-402 (1986). C. E. Kinley and A. E. Marble, ”Compliance: A continuing problem with vascular grafts,” J. Cardiovasc. Surg., 21, 163-170 (1980). R.M. Zwolak, M.C. Adams, and A.W. Clowes, “Kinetics of vein graft hyperplasia: Association with tangential stress,“ I. Vasc. Surg., 5, 126136 (1987). A.W. Clowes, M. A. Reidy, and M. M. Clowes, “Kinetics of cellular proliferation after arterial injury. I. Smooth muscle growth in the absence of endothelium,” Lab. Invest., 49(3), 327-333 (1983). D. L. Fry, “Certain histological and chemical responses of the vascular interface to acutely induced mechanical stress in the aorta of the dog,” Circ. Res., 24, 93-108 (1969). E. D. Endean, D.U. Kim, J. Ellinger, S. Henderson, and H. P. Greisler, ”Effects of polypropylene’s mechanical properties on histological and functional reactions to polyglactin 910/polypropylene vascular prostheses,’’ Surg. Forum, 38, 323-325 (1987).

Received August 1,1991 Accepted March 17, 1992