Evaluation of soft tissue response to a poly[urethane urea]

651

Evaluation of soft tissue response to a polylurethane urea) Mira Mohqty Pathophysiology and Technology,

Division, Biomedical Technology Trivandrum, 695012 India

Wing, Sree Chitra

Tirunal Institute

of Medical

Sciences

J.A. Hunt, P.J. Doherty, D. Annis and D.F. Williams Department

of Clinical Engineering,

University

of Liverpool,

PO Box 147, Liverpool

L69 36X UK

Variations in the performance of vascular prostheses constructed of polyurethanes, and some evidence which suggested that these variations could be due not to the properties of the polymer itself, but to differences in the cellular response to the various microstructures of porous polyurethanes require investigation. Experiments were performed to evaluate quantitatively the extent of the cell behaviour adjacent to a series of polyurethane samples. It was shown that, with BiomeP, a polyurethane urea, the profile of cell behaviour as a function of distance from the implant surface and of time following implantation, the response of cells in general and macrophages in particular, varied considerably with different internal microstructure. This supports the suggestion that the cellular response to different structures and susceptibility to degradation are related. Keywords:

Vascular prostheses,

poly(urethane

urea), tissue response

Received 25 August 1991; accepted 15 January 1992

Many synthetic materials have been used in attempts to replace diseased small diameter blood vessels including polymers such as poly(ethylene terephthalate) (Dacron@), poly(tetrafluoroethylene) and polyetherurethanes. Polyurethanes have been found to have excellent physical properties which have been put to good use in the artificial heart, pacemaker leads and catheters. One of these, Biomer@, a segmented poly(etherurethane urea), is known to have excellent compatibility with blood’. However, there is conflicting evidence concerning the performance of polyurethanes such as Biomer. In animal experiments, vascular grafts made from polyurethanes and having a diameter < 6 mm have generally failed to remain patent following long-term implantation?, although Annis has reported excellent 2 yr patency rates in dogs3. The advantages of using a material with a mesh structure were first reported by Voorhees et d4. Clowes et al.’ showed that only if the wall of a graft had spaces large enough to allow the passage of capillaries would tissue penetrate the wall of the prosthesis. Annis and co-workers’ designed a small-diameter arterial prosthesis using the poly(etherurethane urea) (PEULJ) Biomer. It was electrostatically spun to produce cylindrical tubes of randomly-orientated PEUU fibrils of about 1.0,em diameter. The size of the pores generated within this structure could be varied. The present study was undertaken to assess, both qualitatively and Correspondence

to Professor

@ 1992 Butterworth-Heinemann 0142-9612/92/100651-06

D.F. Williams.

Ltd

quantitatively, the soft tissue response in a rat model to samples of these vascular prostheses, having pores of varying sizes, to assess the role of microarchitecture in the determination of this response.

MATERIALS

AND METHODS

Samples of the wall of three variations of the electrostatically spun Biomer prosthesis which had different internal structures were selected for evaluation. The microstructure of all three was essentially a lattice of connected fibres, approximately l-2 pm in diameter. The size of the spaces between the fibres ranged from 5 to 10j~rn in sample A. Sample B resembled sample A but had some voids of up to 20 pm. Sample C, also similar to sample A, also had voids of up to 300pm. Scanning electron micrographs of cross-sections of all three grafts clearly show the different internal structures (Figure 2). The open structure of all samples extended through their entire thickness. The prostheses from which the samples were taken were made by the electrostatic spinning process described by Annis et al6 Twenty-four black and white hooded Lister rats were used in this study. Flat specimens removed from each prosthesis, approximately 5 X 5 mm, were implanted bilaterally into the dorsolumbar musculature of each animal [four of each graft for each time period]. All specimens were cleaned in a sonicator before implantaBiomaterials

1992. Vol. 13 No. 10

652

Soft tissue response to a PEUU: M. Mo~anfy et al.

Figure1 Scanning electron micrographs of cross-section of graft material: a, sample A (original magnification X100); b, sample A (original magnification X1000); c, sample B (original magnification X100); d, sample B (original magnification X1000); e, sample C (original magnification X36); f, sample C (original magnification X1000).

tion. Animals

were anaesthetized using small animal immobilon (0.5 ml/kg). The implants were inserted deep into the muscle, secured by a single Dexon@ suture. The wound site was closed with silk. Periods of implantation were 15 d, 1 month, 2 months and 3 months. The rats were killed at these time periods using a standard technique and the implant with surrounding tissue removed. The tissue with the implant in sifu was immediately frozen using isopentane and dry ice, fixed on to metal chucks and 7pm thick sections cut on a cryostat. Selected sections were stained with haematoxylin and eosin and examined by light microscopy. Further sections were evaluated using an immunoBiomaterials

1992, Vol. 13 No. 10

histochemical technique7. A series of commercially obtained monoclonal antibodies were employed to permit specific cell staining. Macrophages were identified with ED2 which recognizes a membrane antigen. T-lymphocytes were labelled using CDB-type antibody and activated T-lymphocytes were stained with Intedeukin 2 (ILZ) receptor antibody. Subsequent histomorphometry was carried out using an image analysis technique involving a Joyce Loebl Mini-Magiscan. The stained sections were viewed under a Zeiss Jenoval photomicroscope and the image captured with a Hitachi KP140 CCD mono~h~me video camera. The parameters measured included cell number, cell area, distance from implant and circularity’.

Soft tissue response to a PEUU: M.

~oban~ et al.

.~

653

RESULTS All 48 implants were recovered at all time intervals, and all had retained their original form, with no gross evidence of material degradation.

Light microscopy Microscopic examination of haematoxylin-stained crosssections of implant with surrounding tissue revealed a marked cellular response in the vicinity of all three implants at 15 d. The local cell population comprised numerous macrophages and a few neutrophils. The chronic inflammation observed around samples A and B appeared to have subsided by 1 and 2 months. There was evidence of repair with a few fibroblasts and fibrocytes. However, implant C continued to elicit a marked inflammatory reaction even at this time period, with numerous macrophages and a few giant cells close to the implant. At 3 months, implants A and B were surrounded by tissue showing a moderate cellular response with numerous macrophages still present (Figure 2). Marked chronic inflammation persisted in response to implant C. Though samples A and B retained their form in all sections, implant C fragmented at all time periods during sectioning, no doubt in part due to its more fragile structure.

The identification and quantitative assessment of the specifically stained cells corresponded to the visual observations made by light microscopy. Cell numbers were counted and plotted with respect to distance from the tissue implant interface. Counts were prepared from representative sections, with 20 captured frames providing the data from each section. Sections were examined for each cell type and at each time period. Photomicrographs of sections selectively stained for macrophages are shown in Figure 3 and the macrophage distribution is shown in Figure 4. Sections stained with haematoxylin only were used to count total number of cells (Figure 5). The inflammatory cell response to sample A was predominantly composed of macrophages (101 cells per section unit at 15 df. Although the number of macrophages increased to 181 cells at 1 month, followed by a sharp drop to 71 cells at 2 months, they were found to have increased again at the end of 3 months (114 cells per section unit). Most of these cells were evenly distributed at 15 d, but were concentrated within a distance of 150pm from the interface at the later time periods. The cellular response to sample B was similar in quantity and distribution to that seen with graft A. The number of macrophages at each time period followed a similar pattern, dropping gradually from 185 cells at 15 d to 178 at 1 month and 93 cells at 2 months, but increasing to 131 cells at the end of 3 months. Sample C elicited a larger macrophage response at all time periods, with 328 cells per section unit at 15 d, which although declining to 111 at 1 month, rose again to 206 at 2 months and 260 at 3 months. At 15 d, the cells were spread out over a large distance from the tissueimplant interface, but were all concentrated close to the interface at 1 month. At 3 months, however, the

Figure2 Cellular response adjacent to implant surface; a, sample A; b, sample B; c, sample C.

distribution had broadened again with the maximum number of cells around 312pm from the implant interface and a few still present 600 pm away from it. No other cell type could be identified and the macrophage response corresponded very closely with the total cellular response.

DISCUSSION During the development of a new synthetic arterial prosthesis, it has often been observed that materials which have excellent physical properties do not necessarily prove successful in their place of intended use. The assessment of the biocompatibility of implant materials is an important aspect of device development and entails a detailedin vivo study of the tissue response, Biomaterials

1992, Vol. 13 No. 10

654

Soft tissue response

to a PEUU: M. Mohanty et al.

15 Days

Material

1 Month

2 Months

Figure 3 Specifically stained macrophages around implant site: a, sample A; b, sample B; c, sample C.

not only to the material in solid form and the precise structure of the prosthesis. This response is initially one of acute inflammation dominated by the polymorphonuclear leucocytes, followed by a chronic phase with macrophages and l~phoc~es, The dete~ination of the type of cellular response, its duration and magnitude, is important in ascertaining the biocompatibility of a material. Recently, greater importance has been given to the quantification of the cellular response surrounding implants and the activity of specific cell types, particularly macrophages. This study is a quantitative and qualitative assessment of the cellular response to in tivo implantation of samples of vascular prosthesis, each prepared from Biomer but having different internal structures. Histomorphometry has enabled an accurate quantification of the cells present around the implant at each time interval. Biomaterials

1992, Vol. 13 No. 10

3 Months

Figure4 Plot of macrophage distribution against time: a, 15 d; b, t month; c, 2 months; d, 3 months post-implantation.

Soft tissue resoonse

655

to a PEUU: M. Mohantv et al. 15 Days P

1 Month

125

2 Months t

Material

3 Months

Figure 5 Plot of total cellular distribution b, 1 month; c, 2 months: d, 3 months.

against time: a, 15 d;

Biomer has been used increasingly in the development of a variety of implant devices, including arterial prostheses. Although it has been found to have the desired physical properties essential for some applications, frequent failures have been encountered in others. Attempts have been made to alter the various fabrication techniques to overcome these failures. Porous surfaces have been found to encourage tissue ingrowth, the size of the pores being an important determinant of ingrowth of tissue and maintenance of the inner lining of a vascular prosthesis’s lo. The vascular prosthesis in this study had a porous internal structure varying from small pores in sample A to much larger irregular ones in sample C. The pores were interconnected and open to both the internal and external surface of the prosthesis. The inflammatory response around sample A showed a gradual decline over 3 months, though some macrophages were still seen at the tissue material interface. Sample C, however, was found to elicit a large macrophage response even at 3 months over a wider area. Arterial prostheses having an internal structure of A have shown good long-term patency at 2 yr following implantation3. The small pores allowed only superficial capillary ingrowth and a virtual absence of inflammatory cells in the adjacent tissue. Prostheses having a structure of C were grossly patent when removed after 6 months. However, a histological study of these explanted grafts revealed a large number of inflammatory cells, including macrophages, in the outer half of the graft wall and adjacent tissues. Some of the macrophages contained fragments of the graft material, indicating degeneration of the polyurethane fibres. This suggested that the tissue response and the susceptibility to degradation were closely related and that both phenomena are intimately related to the microstructure of the prosthesis. Large cavities within the wall permit easy access from outside of an inflammatory process dominated by macrophage activity. It was observed, in another context, that an increase in porosity leads to an increase in the rate of degradation of polyurethane with external surfaces being the most susceptible to degradation”. The long-term in vivo stability of polyurethanes has been the subject of many recent studies, some showing conclusive evidence of definite degradation while others do not”. Many hypotheses have been postulated regarding the causes of the observed degradation. Pits on the surface of Biomer film 120 d post-implantation’3 prompted Marchant et al. to carry out anin vivo study on leucocyte interactions with Biomer14. The results of this were interesting in that they showed a preferential adherence of macrophages to the Biomer surface even at day 4 following implantation, There is abundant evidence to implicate the macrophage as one of the major effector cells in ‘wound healingI and activated macrophages have been found to produce many growth factors and enzymes which play an important role in inflammation and wound healing. Interleukin 1 (IL 1) is an important product which induces proliferation of fibroblasts and monocytes, thus inducing fibrous tissue formation’6. A suppressed IL 1 production has been found when monocyteslmacrophages interact with Biomer17. The absence of significant fibroblast invasion accompanying degradation of the external surface of the graft was earlier observed’, ‘a*lg. Biomaterials

1992. Vol. 13 No. 10

656

Soft tissue response

More recently Brothers et al. have suggested that inhibition of fibroblast p~liferation by pol~rethane may account for the failure of autologous tissue to provide adequate structural integrity as polyurethane degenerates”‘. The results obtained in this study confirm that the magnitude of the macrophage response is influenced by the internal structure of the polyetherurethane prosthesis, and in particular by the size of the pores that penetrate the wall of the prosthesis. Although in this study there was no gross evidence of material degradation in any group at any time period, prolonging the period of implantation may be expected to result in degradation, as found in the vascular replacement study with prosthesis having large pores, sample C in this study. The moderate total cell number and macrophage content in response to material B is encouraging.

4

5 0

7

8

9

10

CONCLUSION The extent of the cellular response in general, and especially of the macrophage response, is clearly dependent on the pore microstructure of an arterial prosthesis of PEUU Biomer. A pore size in the region 5-lOpurn results in little infiltration and generally low number of cells in the surrounding tissue, Large pore size in some regions of the wall of the prosthesis being ZOO300pm, results in a considerably greater cell number in the tissues surrounding the sample. In all cases, however, the number of cells decreased after 15-30 d, but increased again at 3 months. These variations in cell numbers, especially macrophages, is consistent with the hypothesis that the biodegradation of pol~rethanes is related to cell activity on the surface.

11

12

13

14

15

16

ACKNOWLEDGEMENTS Mira Mohanty acknowledges the support of the British Council in providing a training scholarship. REFERENCES 1

2

3

Boretos, J.W., Pierce, W.S., Baier, R.E., Leroy, A.F. and Donachy, H.J., Surface and bulk characteristics of a polyether urethane for artificial hearts,l. Biomed. Mater. Res. 1975, 9, 327-340 Martz, H., Paytner, R., Slimane, S.B., Beaudoin, G. and Guiloin, R., Hydrophilic microporous polyu~thane versus expanded PTFE grafts as substitutes in the carotid arteries of dogs. A limited study, f. Biomed. Mater. Res. 1988, 22, 63-69 Annis, D., The development and the testing of a

Biomaterials

1992, Vol. 13 No. 10

17

18

19

20

to a PEUU: M. Mohanty

et a/.

polyurethane arterial prosthesis, Bull. Mater. Sci. 1989, 12, 33-34 Voorhees, A.B., Jaretzki, A. and Blakemore, A.H., The use of tubes of Vinyon N cloth in bridging arterial defects, Ann. Surg. 1952, 135, 332 Clowes, A.W., Kirkman, T.R. and Reidy, M.A., Mechanisms of arterial graft, Am. J. Pathol. 1986, 123, 220-230 Annis, D., Bornat, A., Edwards, R.O., Higham, A., Loveday, 8. and Wilson, J., An elastomeric vascular prosthesis, Trans. Am. Sot. Artif, Int. Organs 1978, 24, 204-214 Vince, D.G., Hunt, J.A. and Williams, D-F., Quantitative assessment of the tissue response to implanted biomaterials Hunt, J.A., Vince, D.J. and Williams, D.F., Histomorphometry of biomaterial evaluation, I. Riomed. Eng. [in press] Fry, W.J., Deweese, M.S., Craft, R.O. and Ernst, C.B., Importance of porosity in arterial prosthesis, Arch. Surg. 1964, 88, 838-842 Campbell, CD., Goldfarb, D. and Rose, R., A small arterial substitute: expanded microporous poiytetrafluoroethylene: patency versus porosity, Ann. Surg. 1975, 182, 138-143 Marchant, R.E., Zhao, Q., Anderson, J.M. and Hiltner, A., Degradation of a poly(ether urethane urea)elastomer: infra-red and XPS studies, Polymer 1987, 28, 2032 Ratner, B.D., Gladhill, W.D. and Horbett, T.A., Analysis of in vitro enzymatic and oxidative deg~dation of polyurethanes,]. Biomed. Mater. Res. 1988,22,509-527 Marchant, R.E., Anderson, J.M,, Phua, K. and Hiltner, A., In vivo biocompatibility studies: II. Biomer: preliminary cell adhesion and surface characterization studies, I. Biomed. Mater. Res. 1984, 18, 309-315 Marchant, R.E., Miller, K.M. and Anderson, J.M., In vivo biocompatibility studies. V. In vivo leucocyte interactions with Biomer,J. Biomed. Mater. Res. 1984,18,1169-1190 Clark, R.A.F. and Henson, P.M., The Molecular and Cefluiar Biology of Wound Repair Plenum Press, New York, USA, 1988 Dinarello, CA., An update on human Interleukin 1 molecular biology to clinical relevance, 1. Clin. Immunol. 1985, 5, 287-297 Bonfield, T.L., Colton, C.E. and Anderson, J.M., Plasma protein adsorbed biomedical polymers: activation of human monocytes and induction of Interleukin 1, J. Biomed. Mater. Res. 1989, 23 (6) 535-548 Wilson, G.J., MacGregor, D.C., Kfement, P., Lee, J.M., de1 Nido, P.J., Wong, E.W.C. and Leidner, J., Anisotropic polyurethane non woven conduits: a new approach to a vascular prosthesis, Trans. Am. Sot. Artif. Inter. Organs 1983, 24, 280-268 Hess,F., Braun, B., Jerusalem, C., VanDet, R., Steeghs, S., Scotnicky, S. and Grande, P., Endothelization of polyurethane vascular prosthesis implanted in the dog carotid and femoral artery, J. Cam&was. Surg. 1988, 29,458 Brothers, T.E., Stanley, J.C., Burkell, W.E, and Graham, L.M., Small calibre polyurethane polytetrafluroethylene grafts: A comparative study in a canine aortic model, J. Biomed. Mater. Res. 1990, 24, 781-771