Artificial Organs 21(5):391-395, Blackwell Science, Inc., Boston 0 1997 International Society for Artificial Organs

Is the Dog a Useful Model for Accelerated Calcification Study of Cardiovascular Bioprostheses?

Jyotirmay Chanda, Ryosei Kuribayashi, Tadaaki Abe, Satoshi Sekine, Yoshiki Shibata, and Itsuro Yamagishi

Department of Cardiovascular Surgery, Akita University School of Medicine, Akita, Japan

Abstract: Chitosan posttreatment has been shown to be effective in prevention of calcification of the glutaral- dehyde treated bovine pericardium when implanted sub- dermally in rats for 12 weeks. The efficacy of chitosan posttreatment in complete calcium mitigation of the glu- taraldehyde treated porcine aortic valves implanted in the right side of the heart in dogs was well-documented in our previous study. In this study, an attempt has been made to evaluate the merit of the chitosan posttreatment in pre- vention of calcification of the glutaraldehyde (GA) treated porcine aortic valved conduits in the systemic circulation in dogs for a period of 5 months. Eleven mongrel dogs underwent left thoracotomy. Porcine aortic valved con- duits treated with 0.625% GA (n = 5) and GA-chitosan (n

= 6) were implanted in the descending thoracic aortas of the dogs for 5 months. Gross histological observations showed no calcification in either the 0.625% GA treated or in the GA-chitosan treated valved conduits at 5 months. This was confirmed by results of quantitative analyses for calcium in each explant. There was no significant differ- ence in calcium content between the GA only (Ca, 0.43 * 0.26 mg/g) and GA-chitosan treated (Ca, 0.51 f 0.19 mg/g; p = 0.5959) valved conduits. This study suggests that the dog is not a suitable model for evaluating the efficacy of a calcium mitigating agent in bioprostheses implanted in sys- temic circulation. Key Words: Aortic surgery-Bio- prosthetic valved conduit-Glutaraldehyde-Chitosan- Apicoaortic shunt.

The quality of life after mechanical heart valve replacement is not ideal, mainly because of the ne- cessity of lifelong anticoagulation and the ever pres- ent risk of thromboembolism (1). Valve repair in- stead of valve replacement is thus increasingly attempted, but the procedure is not always feasible (2,3), and the hemodynamic results can be unsatis- factory (4). The alternative to mechanical valve re- placement is the use of biological prostheses: there is no need for lifelong anticoagulation even in the pres- ence of chronic atrial fibrillation, and the risk of sys- temic thromboembolism is very low. Furthermore, the quality of life can be considered very close to normal, allowing heavy work, sports, normal preg- nancy and delivery, and life in remote areas (5,6). The primary concern with bioprostheses, well- documented over the previous years, is skepticism regarding the durability and the potential morbidity

Received February 1996; revised June 1996. Address correspondence and reprint requests to Dr. Jyotirmay

Chanda, Department of Cardiovascular Surgery, Akita University School of Medicine, Akita 010, Japan. E-mail: jchandaamed. akita-u.ac.jp

and mortality related to degenerative prostheses (7- 10). The major concern with porcine bioprostheses has been their long-term durability because of fibro- calcification and fatigue-induced leaflet disruption. A new technique for prevention of calcification of bioprostheses was devised that takes advantage of the conventional glutaraldehyde fixation technique for these devices. Chanda and associates (11-14) demonstrated that glutaraldehyde treated chitosan or other amino compound posttreated bovine peri- cardium did not calcify when implanted subcutane- ously in rats for 12 weeks, and the posttreatment with chitosan did not have any detrimental effect on the mechanical strength of the glutaraldehyde treated pericardium. Later, efficacy of the chitosan posttreatment in complete calcium mitigation of the glutaraldehyde treated porcine aortic valves im- planted in the right side of the heart in dogs was well documented (15). The current study was designed to determine the efficacy of chitosan posttreatment in prevention of calcification of glutaraldehyde treated porcine aortic valves in the systemic circulation in dogs.

391

Exh. 2002 Page 1

University of Michigan Exhibit 2002St. Jude Medical v. University of Michigan IPR 2013-0041

University of Michigan Exhibit 2002St. Jude Medical v. University of Michigan IPR 2013-0041

392 J. CHANDA ET AL.

MATERIALS AND METHODS

Partially degraded heparin Commercially available heparin (100 ml) (Novo,

Nordisk A& Bollerup, Denmark; 1,000 U/ml) was mixed with 400 ml normal saline and chilled to 2°C. To this mixture, 10 mg NaNOz (sodium nitrite) was added, and the pH was adjusted to 2.0 by adding 1 M HCl (hydrochloric acid). The reaction was allowed to run for 1 h at 2°C. The pH was adjusted to 7.0 with 1 M NaOH (sodium hydroxide). The solution was millipore filtered and preserved at 4°C in sterile con- dition.

Covalently bonded heparin coated tube A tube covalently bonded with heparin was pre-

pared according to the modified method described by Larm and colleagues (16). A 5 mm polyvinyl tube was filled with concentrated sulfuric acid containing 0.2% potassium permanganate and incubated for 2 min and carefully rinsed with deionized water. Poly- ethyleneimine (Nacalai Tesque., Kyoto, Japan) was dissolved in borate buffer (pH 9) to yield a final concentration of 0.1 mglml. The sulfuric acid/ potassium permanganate treated polyvinyl tube was exposed to the polyethyleneimine for 10 min at room temperature, rinsed with borate buffer for 5 min, and exposed to a solution of partially degraded heparin (100 U/ml) for 15 min at room temperature and fi- nally rinsed in deionized water. These heparin coated tubes were sterilized with ethylene oxide gas.

Valved conduits Porcine aortic valves were obtained from a slaugh-

ter house and under zero pressure were crosslinked with glutaraldehyde (Glutaraldehyde EM 25%, TAAB Laboratories Equipment Limited, Reading, United Kingdom) in 0.067 M phosphate buffer solu- tion (pH 8.2) at 4-10°C with a gradual increase in concentration from 0.1% to 0.25% over a period of more than 8 weeks. After the onset of glutaralde- hyde treatment, subannular tissues of the valves were trimmed down to approximately 2 mm from the base of the leaflets, and knitted Dacron grafts (USCI knitted Dacron vascular grafts, Bard Im- plants Division, C.R. Bard, Inc., Billerica, MA, U.S.A.) were sewn to the ventricular and aortic ends of the valves with 5-0 and 4-0 polypropylene, respec- tively. To neutralize the free aldehyde groups and slow the release of residual glutaraldehyde, the glu- taraldehyde treated valved conduits were post- treated with 0.125% chitosan (Sigma Chemical Com- pany, St. Louis, MO, U.S.A.) in deionized water (pH 6.0) for 3 days and were stored in normal saline until implantation. Because most of the commercially

available bioprosthetic valves are 0.625% glutaralde- hyde treated, 0.625% glutaraldehyde treated valved conduits were prepared and stored in 0.625% glutar- aldehyde solution as controls. Both the glutaralde- hyde only and glutaraldehyde-chitosan treated valved conduits were washed in copious amounts of normal saline (Otsuka Pharmaceutical Co., LTD., Tokyo, Japan) before surgery.

Surgical technique Six glutaraldehyde-chitosan treated and five

0.625 YO glutaraldehyde treated porcine aortic valved conduits were implanted in 11 mongrel dogs (age, 3-7 years) weighing 15 to 20 kg. All animals received humane care in compliance with the Principles of Laboratory Animal Care formulated by the National Society for Medical Research and the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health (NIH Publica- tion No. 80-23, revised in 1985). Recipient dogs were preanesthesized with intramuscular ketamine (15 mg/kg). After additional anesthesia with intravenous pentobarbital sodium (30 mgkg), dogs were endo- tracheally intubated, and 0.1 mg/kg pancuronium was administered. Ventilation was maintained with a mixture of oxygen (3 Llmin) and room air. Under aseptic conditions, a left thoracotomy incision was performed through the fifth intercostal space. The descending thoracic aorta, distal to the subclavian artery 4 to 5 cm in length, was mobilized and proxi- mally cross-clamped and cut at the middle of the mobilized part after heparinization (100 U/kg). To avoid the peripheral ischemia and to minimize the blood loss, a temporary apicoaortic (through the dis- tal portion of the descending thoracic aorta) shunt was established with a piece of polyvinyl tube cova- lently bonded with partially degraded heparin (de- scribed previously). Proximal and distal ends of the valved conduit (Dacron grafts preclotted with blood before heparinization) were anastomosed to the cut ends of the descending thoracic aorta in end-to-end fashion with continuous 5-0 polypropylene, and the shunt was discontinued. The chest tube was removed immediately upon completion of the skin suture. In- travenous cefazolin (2 g) (Cefamezin, Fujisawa Phar- maceuticals, Tokyo, Japan) and amikacin (100 mg) (Banyo Pharmaceuticals, Tokyo, Japan) were rou- tinely administered before each operation. Amikacin (100 mg) was repeated twice daily for 5 days after surgery. All dogs were sacrificed 5 months after im- plantation. The porcine valved conduit was removed from the dog’s descending thoracic aorta, grossly ex- amined, and incised longitudinally. Cusps were care- fully excised and individually lyophilized for calcium

Artif Organs, Vol. 21, No. 5, 1997

Exh. 2002 Page 2

VALVED CONDUIT IN DOG 393

estimation. A longitudinal section of each cusp was preserved in 10% formalin. Histologic sections were stained with hematoxylin eosin, Masson’s trichrome. elastin van Gieson, Azan Mallory, and von Kossa for general morphology and calcification study.

Calcium estimation The retrieved samples were dissected free of host

tissue, rinsed with copious amounts of deionized wa- ter, freeze-dried to a constant weight, and weighed. The amount of calcium was determined by atomic absorption spectroscopy (Hitachi 2 6100, Tokyo, Ja- pan) at a wave length of 422.7 nm on aliquots of 60% H2N0, (Kanto Chemicals, Tokyo, Japan) hydroly- sates of dried tissue, which were diluted with 1% lanthanum chloride solution (Wako Chemicals, To- kyo, Japan). A standard curve was obtained using the calcium standard solution (Wako Chemicals) in lanthanum solution. The amount of calcium was ex- pressed as milligrams per gram of dry tissue.

RESULTS

All dogs of both groups survived the surgery, and they did well through the planned 5 month implant period. Both glutaraldehyde only and glutaralde- hyde-chitosan treated Dacron vascular grafts had ad- herent inner capsules and were incorporated on the external surface by a dense fibrous tissue. The pseu- dointima in the Dacron vascular graft varied widely in thickness and was thicker at the anastomosis with the porcine aortic valve. Because of marked tissue ingrowth, especially from the proximal end, and ac- cumulation of fibrin in the sinuses of the Valsalva, dysfunction of the valves (limitation of leaflet move-

FIG. 1. This photograph, a gross cross-section of a valve, shows the inflow surface of a porcine aortic valve of a 0.625% glutaral- dehyde treated valved conduit implanted in a dog for 5 months. Note that the marked tissue ingrowth in the valved conduit caused the leaflet to become immobile and the valve to dysfunc- tion.

FIG. 2. This photograph shows the inflow surface of a porcine aortic valve of a 0.625% glutaraldehyde treated valved conduit implanted in the descending thoracic aorta of a dog for 5 months. Good coaptation of the leaflets can be noticed when compared with the leaflets in Fig. 1.

ment) occurred in 2 cases of each group (Fig. 1). Mild to moderate fibrous ingrowth did not cause val- vular dysfunction in the remaining implants of both groups (Fig. 2). Severe conduit stenosis with fibrous peel or peel dissection was not noticed in any case. Leaflets of porcine aortic valves were glossy and pli- able without any susceptible change. Light micro- graph showed no tissue degradation, cuspal hema- toma, tear, or cellular infiltration in the explanted valves of both groups. Histological and chemical analyses did not reveal any calcification of valves of either group (Figs. 3 and 4). The calcium contents of the 0.625% glutaraldehyde treated and glutaralde- hyde-chitosan treated valves were 0.43 * 0.26 mg/g and 0.51 f 0.19 mg/g dry wt, respectively. These values did not significantly differ from each other (p = 0.5959).

DISCUSSION The specific mechanisms by which glutaraldehyde

fixation facilitates mineralization are not understood (17). It is known that small but cytotoxic levels of glutaraldehyde are released from glutaraldehyde treated bioprostheses over a period of time (18). One may hypothesize that in clinical situations after implantation of glutaraldehyde treated bioprosthe- ses, free aldehyde moieties on the surface of the im- plant undergo oxidation followed by acid formation, and this acid probably traps the host plasma calcium. The slow release of residual (unbound) glutaralde- hyde from the prosthesis over a period of time after implantation reinforces the host plasma calcium acid bound complex. This acid plasma calcium complex promotes further mineralization when glutaralde- hyde plays no more role as the primary factor for nucleation of calcification (13). Perhaps calcium im- pregnation starts immediately on exposure of the

Artif Organ.r, Vol. 21. N o . 5. I997

Exh. 2002 Page 3

394 .I. CHANDA ET AL.

FIG. 3. This histologic section shows no calcification of a cusp of a glutaraldehyde treated porcine aortic valved conduit in a dog at 5 months (von Kossa stain) (original magnification x200).

glutaraldehyde treated bioprostheses to the host or- ganism (13). To prevent the slow release of residual glutaraldehyde and to neutralize the free aldehyde groups on the surface of the glutaraldehyde treated bioprostheses, chitosan (a biopolymer with a large number of amino termini), posttreatment was intro- duced to covalently crosslink with the aldehyde moi- eties (11-13).

The effectiveness of anticalcification treatment could be overestimated in subcutaneous implants be- cause specimens in this model are not subjected to mechanical and dynamic stresses, and neither circu- latory washout of beneficial factors nor reaccumula- tion of previously extracted calcifiable material can be easily assessed with this model. Moreover, there is no blood surface contact, thereby preventing as- sessment of interactions with blood-borne sub- stances, particularly platelets, coagulation factors, li- poproteins, and lipids. The most important areas of safety evaluation are blood surface interactions, lo- cal or systemic toxicities, and inflammatory effects. The subsequent step in assessing anticalcification ef- ficacy and possible clinical relevance and utility is large animal testing of intact treated valves (19). Okamura and co-workers (20) reported that there is an increase in calcification with the increase in con- centration of glutaraldehyde of glutaraldehyde treated heterografts implanted in descending tho- racic aortas in mongrel dogs. Perhaps because glu- taraldehyde of any concentration is equally effective in induction of calcium deposition on bioprostheses, we did not find any correlation of calcification with the concentration of glutaraldehyde in glutaralde- hyde only treated implants in rats (13). To judge the efficacy of our proposed anticalcification treatment in systemic circulation, we implanted 0.625% glutar- aldehyde and glutaraldehyde-chitosan treated por-

cine aortic valved conduits in the descending tho- racic aortas in dogs for 5 months. Neither 0.625% glutaraldehyde treated nor glutaraldehyde-chitosan treated porcine aortic valved conduits showed any sign of calcification during our observation period. However, there was mild to moderate calcification of 0.625% glutaraldehyde treated porcine aortic non- coronary cusps implanted in pulmonary circulation in dogs for 5 months (15). Conceivably, discrepan- cies in calcification might have occurred because of differences in age among mongrel dogs used in dif- ferent experiments.

The dog is not generally considered to be a useful model for accelerated calcification (21). Kethar- nathan and Christie (22) implanted glutaraldehyde treated ovine collagen conduits as vascular xeno- grafts in dogs. No calcification of the conduits was noted at 3 years. It is difficult to explain why even glutaraldehyde only treated valves do not calcify af- ter implantation in systemic circulation and do cal- cify in pulmonary circulation in dogs. Because the age of mongrel dogs is uncertain, conceivably the calcification feature of glutaraldehyde treated grafts would be quite different if implanted in young pup- pies. Presumably, because of discrepancies in fea- tures of calcification of bioprostheses in dogs, the juvenile sheep model is the standard required by the U.S. Food and Drug Administration for such chronic bioprosthetic valve implant studies.

CONCLUSION

Because of uncertainty of age, mongrel dogs are not a reliable model for evaluating the efficacy of any anticalcification agent in cardiovascular biopros- theses.

FIG. 4. On this photomicrograph of a leaflet of a 0.625% glutar- aldehyde treated porcine aortic valved conduit implanted in the descending thoracic aorta of a dog for 5 months, no calcific de- posits can be seen (von Kossa stain) (original magnification x200).

Artif Organs3 Vol. 21, No. 5, 1997

Exh. 2002 Page 4

VALVED CONDUIT IN DOG 395

REFERENCES 1. Horstkotte D. Korfer R, Seipel L, Looge NF. Late complica-

tions in patients with Bjork-Shiley and St. Jude medical heart valve replacement. Circulation 1983:68:(Suppl 2):175-85.

2. Angel1 WW, Oury JH, Shah P. A comparison of replacement and reconstruction in patients with mitral regurgitation. J Thorac Cardiovasc Surg 1987;93:665-74.

3. Duran C, Kumar N, Gometza B, Alhalees A. Indications and limitations of aortic valve reconstruction. Ann Thorac Surg 1991 :52:447-54.

4. Duran CG, Revuelta JM. Gaite L. Stability of mitral recon- structive surgery at 10-12 years for predominantly rheumatic valvular disease. Circularion 1988;(Suppl 1):914.

5 . Salazar E, Zajaris A. Gutierezz N. Iturbe I. The problem of cardiac valve prostheses, anticoagulants, and pregnancy. Cir- culation 1984:7O(Suppl 1):169-70.

6. Jamieson WRE, Miller DC. Akins CW, Munro AI, Glower DD, Moore KA, Henderson C. Pregnancy and bioprostheses: Influence on the structural valve deterioration. Ann Thorac Surg 1996;60:S282-7.

7. Gallo I. Ruiz B, Nistal F, Duran CMG. Degeneration in por- cine bioprosthetic cardiac valves: Incidence of primary tissue failure among 938 bioprostheses at risk. A m J Cardiol 1984; 53:1061-5.

8. Pomar JL, Bosch X, Chaitman BR, Pelletier C, Grondin CM. Late tears in leaflets of porcine bioprostheses in adults. Ann Thorac Surg 1983:36:246-52.

9. Gallo I, Nistal F, Artinano E. Six- to ten-year follow-up of patients with the Hancock cardiac bioprostheses. J Thorac Cardiovasc Surg 1986;92:14-20.

10. Jamieson WRE, Munro I. Miyagishima RT, Burr LH, Gerein AN, Janusz MT, Tyres GFO, Allen P. The Carpentier- Edwards supraannular porcine bioprosthesis. A new genera- tion tissue valve with excellent intermediate clinical perfor- mance. J Thorac Cardiovasc Surg 1988;96:65246.

11. Chanda J. Anticalcification treatment of pericardial prosthe- ses. Biomaterials 1994:15:465-9.

12. Chanda J, Rao SB, Mohanty M, La1 AV. Muraleedharan CV. Bhuvaneshwar GS. Valiathan MS. Prevention of calcification of tissue valves. Arfif Organs 1994;18:752-7.

13. Chanda J . Prevention of calcification of heart valve biopros- theses: An experimental study in rat. Ann Thorac Surg 1995: 60:S33942.

14. Chanda J. Posttreatment with amino compounds effective in prevention of calcification of glutaraldehyde treated pericar- dium. Artif Organs 1994;18:408-10.

15. Kuribayashi R, Chanda J, Abe T. Efficacy of the chitosan posttreatment in calcification prevention of the glutaralde- hyde treated porcine aortic noncoronary cusp implanted in the right ventricular outflow tract in dogs. Artif Organs 19%:20:7614.

6. Larm 0, Larsson R, Olsson P. A new non-thrombogenic sur- face prepared by selective covalent binding of heparin via modified reducing terminal residue. Biotnater Med Devices Arrif Organs 19833 1:161-73.

7. Schoen FJ, Tsao JW, Levy RJ. Calcification of bovine peri- cardium used in cardiac valve bioprostheses. Implications for the mechanisms of bioprosthetic tissue mineralization. A m J Parhol 1986;123:13445.

18. Lynn LH, Lee H, Cheung DT. Nimni ME. Biochemical changes and cytotoxicity associated with the dehydration of polymeric glutaraldehyde derived crosslinks. J Biomed Mater Res 1990;24:1185-91.

19. Schoen FJ. Levy RJ. Hilbert SL. Bianco RW. Antimineral- ization treatments for bioprosthetic heart valves. Assessment of efficacy and safety. J Thorac Cardiovasc Surg 1992:104:

20. Okamura K. Chiba C. Iriyama T. Itoh T. Maeta H , Ijima H, Mitsui T. Hori M. Antigen depressant effect of glutaralde- hyde for heterografts with a valve. with special reference to a concentration right fit for the preservation of grafts. Surgery 1980;87:170-6.

21. Shemin RJ. Schoen FJ. Hein R. Austin J, Cohn LH. Hemo- dynamic and pathologic evaluation of a unileaflet pericardial bioprosthetic valve. J Thorac Cardiovasc Surg 1988;95:912-9.

22. Ketharnathan V. Christie BA. Glutaraldehyde-tanned ovine collagen conduits as vascular xenografts in dogs. Arch Surg 1980:115:967-9.

2285-8.

Artif Organs, Vol. 21. No. S. 1997

Exh. 2002 Page 5