Journal of Orthopaedic Research E2128-137 The Journal of Bone and Joint Surgery, Inc. 0 1994 Orthopaedic Research Society

Intraosseous Incorporation of Composite Collagen Prostheses Designed for Ligament Reconstruction

Michael G. Dunn, Suzanne H. Maxian, and Joseph P. Zawadsky

Orthopaedic Research Laboratory, Division of Orthopaedic Surgery, University of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School, New Brunswick, New Jersey, U S A .

Summary: Composite collagen prostheses are potentially useful for reconstruction of the anterior cruciate ligament (ACL). We evaluated the intraosseous response to composite collagen prostheses to determine if “biological fixation” could be used to secure the pros- theses within surgical bone tunnels. The rate of degradation of the prosthesis and the re- sponse of the tissue were evaluated, as a function of collagen crosslinking agent and time, in nonloaded bone tunnels in rabbits. Prostheses were fabricated by the alignment of 200 reconstituted type-I collagen fibers (60 pm in diameter) and the embedding of the fibers within a collagen matrix. The prostheses degraded rapidly within the bone tunnels in comparison with soft-tissue implantation sites. Dehydrothermal-cyanamide crosslinked collagen fibers were completely degraded by 8 weeks. Only 10% of glutaraldehyde cross- linked collagen fibers remained intact at 12 weeks. Fibrous tissue and inflammatory cells rapidly infiltrated the prostheses, and new bone surrounded the circumference of the pros- theses, advancing toward the center at longer times. At the lateral cortex, where fibrous tis- sue emerged, the bone/soft-tissue interface was delineated by a tidemark, similar to that observed in a normal ligament insertion site. Preliminary pull-out testing of the soft tissue from the bone was discontinued because failure consistently occurred in the soft tissue; this suggests rapid incorporation of the prostheses within the bone tunnels. Composite col- lagen prostheses designed for ACL reconstruction degrade rapidly in bone and induce rapid ingrowth of fibrous tissue and bone. These results suggest that tissue ingrowth in the bone tunnels might provide biological fixation for collagen prostheses used for ACL reconstruction.

Biological grafts currently are used for reconstruc- tion of the anterior cruciate ligament (ACL), while the search for a so-called ideal ACL substitute con- tinues (18). An alternate approach to ligament re- construction involves the development of degradable biomaterials that provide a temporary scaffold to in- duce neoligament deposition (6). In our laboratories, evaluations are being conducted of degradable pros- theses consisting of reconstituted type-I collagen fi- bers aligned within a collagenous matrix (Fig. 1A).

Received February 24,1992; accepted March 31,1993. Address correspondence and reprint requests to Dr. M. G.

Dunn at UMDNJ-Robert Wood Johnson Medical School, Di- vision of Orthopaedic Surgery, MEB 424,l Robert Wood John- son Place-CN 19, New Brunswick, NJ 08903, U.S.A.

The reconstituted collagen fibers (13) have high ten- sile strengths (30-60 MPa) and small diameters (20- 60 pm) similar to fibers of the normal ACL. Results of previous implantation studies suggest that rapid induction of neotendon (9,19) and neoligament (6) tissues occurs in response to implantation of the com- posite collagen prostheses. Optimization of the me- chanical properties of the prostheses and further characterization of neoligament tissue are ongoing.

In most ACL reconstruction studies, only intra- articular healing is evaluated; intraosseous healing is equally critical (1,lO) but is largely ignored by investigators (21). In a previous study using com- posite collagen prostheses for ACL reconstruction, limited histological evaluation within the tibia1 bone tunnel showed new bone ingrowth advancing toward

128

INCORPORATION OF COLLAGEN PROSTHESES IN B O N E I 2 9

a central fibrous region (6). Biomechanically tested femur/neoligament/tibia specimens failed in the mid- substance and not at the boneheoligament inter- face; this suggests that the prostheses were rapidly incorporated within the bone tunnels. Rapid intra- osseous incorporation or biological fixation of the collagen prosthesis is one key to successful ACL reconstruction.

In the present study, therefore, we focused on the intraosseous response to a composite collagen pros- thesis implanted within a femoral metaphyseal bone tunnel. The intraosseous degradation rate of the prostheses and new tissue induction were evaluated, with the use of a simple, nonloaded bone implan- tation model in rabbits, as a function of the colla- gen crosslinking agent (dehydrothermal-cyanamide [DHTC] or glutaraldehyde [GLUT]) and the time after implantation (3,4,8, and 12 weeks).

MATERIALS AND METHODS Fabrication of Composite Collagen Prostheses

Insoluble collagen taken from bovine corium was washed with distilled water and isopropanol, lyoph- ilized, and stored at -3O"C, as previously described (13). The composition of the collagen was charac- terized (after reduction and heat denaturation) by SDS-polyacrylamide gel electrophoresis as alpha l(1) and alpha l(II), beta, gamma, and higher mo- lecular weight components of type-I collagen. The results of amino acid analyses were consistent with the composition of bovine type-I collagen.

A 1% (w/v) collagen dispersion was made by blending (at 10,000 rpm) of the lyophilized colla- gen with an HC1 solution (pH 2.0). The collagen dispersion was degassed in a vacuum of 100 mtorr and was stored in 30 ml disposable syringes at 4°C. Collagen fibers were produced by extrusion of the collagen dispersion through polyethylene tubing (in- ner diameter = 0.58 mm) into a 37°C bath of fiber formation buffer (pH 7 3 , as previously described (13). After 45 minutes, the fibers were transferred to an isopropanol bath (4 hours), washed in distilled water (20 minutes), and air dried under tension overnight.

DHTC crosslinking of fibers (20) took place in an oven at 110°C and a vacuum of 100 mtorr for 3 days, followed by cyanamide vapor crosslinking in a sealed desiccator containing 10 ml of a 400% (w/v) aqueous cyanamide solution for 24 hours. For GLUT vapor crosslinking of fibers, collagen fibers were placed on a shelf in a sealed desiccator containing 10 ml of a 25% (w/v) aqueous GLUT solution for 24 hours.

Composite collagen prostheses (Fig. 1A) were pre- pared by the alignment of 200 crosslinked collagen fibers (mean fiber diameter = 60 pm wet), coating with a 1% (w/v) collagen dispersion in pH 2.0 HC1, drying, and extensive washing in distilled water. The prostheses were cut into 4cm lengths and steri- lized in Exspor cold sterilant (Alcide, Norwalk, CT, U.S.A.). They then were rinsed for at least 60 min- utes in sterile saline solution prior to implantation.

Surgical Procedure Twelve skeletally mature male New Zealand White

rabbits weighing approximately 10 lb (4.5 kg) were used; a total of 24 prostheses were implanted. In each rabbit, a GLUT-crosslinked prosthesis was implanted in the left femur and a DHTC-crosslinked prosthesis was implanted in the right femur. Surgical procedures were performed with the animals in a sterile field, under general anesthesia induced by intramuscular injection of 7:5 solution of ketamine and xylazine (0.6 ml/kg body weight) and maintained by inhalation of a 2:l mixture of oxygen and nitrous oxide with 1% halothane.

The lateral aspect of the distal femur was exposed. With use of Kirschner wires and a mini-driver (3M; Minneapolis, MN, U.S.A.), a 2.8 mm diameter tunnel was created through the metaphyseal bone of the femur, distal to the epiphyseal scar and perpendicu- lar to the long axis of the femur (Fig. 1B). The sterile collagen prosthesis (4 cm in length) was placed with- in the bone tunnel, with great care being taken to ensure that the prosthesis filled the entire length of the drill defect. The end of the prosthesis exiting the lateral end of the drill defect was bent and was su- tured to the lateral femoral periosteum with use of 4-0 polypropylene (Prolene; Ethicon, Somerville, NJ, U.S.A.). This model allows for investigation of the osseous response to the implant in the absence of complicated biomechanical forces (1).

The skin was closed with 3-0 nylon (Ethilon; Ethi- con) and a running simple stitch. The incision site was sprayed with a topical antibacterial agent (4% furazolidone; Veterinary Products Labs, Phoenix, AZ, U.S.A.). The animals were returned to individual cages, allowed unrestricted activity, and given a diet of water and food ad libitum. Prophylactic oral an- tibiotics (tetracycline capsules, 500 mg/16 oz drink- ing water) were administered postoperatively for 10 days.

Three animals each were killed at 3,4,8, and 12 weeks after implantation; anesthesia was followed by an intracardiac injection of euthanasia solution

J Orthop Res. Vol. 12, No. I , 1994

130 M. G. D U N N ET AL.

200 COLLAGEN FIBERS EMBEDDED IN MATRIX

LENGTH 9 4 cm INDIVIDUAL FIBER DIAMETER = 50-70 urn

B MEDIAL CORTEX

CROSS-SECTION

0 = COLLAGEN FIBER = = COLLAGEN MATRIX

LATERAL CORTEX

COMPOSITE I 1 COLLAGENOUS PROSTHESIS

+ 2.8 mm DIAMETER

BONE TUNNEL

DISTAL FEMUR

FIG. 1. A: Schematic of composite collagen prosthesis designed for reconstruction of the anterior cruciate liga- ment (6). This fibrous prosthesis was fabricated by the alignment of 200 reconstituted type-I collagen fibers in parallel and embedding of the fibers in an uncrosslinked collagen matrix. B: Schematic of the nonloaded surgical model used to evaluate the intraosseous response to the collagen prostheses. Bone tunnels (2.8 mm in diameter) were made across the distal femoral metaphysis bilaterally in rabbits. One prosthesis was placed in each tunnel and was fixed to the lateral periosteum with use of nonresorb- able suture.

SUTURE

(Beuthanasia-D; Scheering, Kenilworth, NJ, U.S.A.).

Mechanical Testing In preliminary studies that used this bone implan-

tation model, pull-out testing of the soft tissue from the bone resulted in failure in the soft-tissue mid- substance; this finding suggests that the soft tis- sue was weaker than the bonelsoft-tissue interface. Pull-out testing therefore was discontinued, and the bonehoft-tissue interface was evaluated by histolog- ical analyses in this group of 12 animals.

Nondecalcified Histology The distal third of each femur was removed,

trimmed of excess soft tissue, and fixed in Carson’s buffered formalin for 1 day. Samples were dried in a series of graded alcohols for 1 week, cleared in xylene (8 hours, two changes), and infiltrated (2 weeks) and embedded (1 week) in Osteo-Bed bone embedding medium (Polysciences, Warrington, PA, U.S.A.). The implant site was sectioned with use of a low-speed diamond saw (Isomet; Buehler, Lake Bluff, IL, U.S.A.). Cross-sectional and longitudinal

cuts of the prostheses were made. All sections were surface-stained with toluidine blue and were exam- ined and photographed at magnifications of x40 and xl00 with use of a light microscope (American Optical Scientific Instruments, Buffalo, NY, U.S.A.).

Determination of Prosthesis Degradation Rate The rates of prosthesis degradation were deter-

mined by a count of the total number of fibers re- maining on three slides per implant per time period. As the bone specimen was cut lateral to medial on the saw, the first three sections with the implant with- in the bone tissue were used for quantitative analysis of this rate.

An analysis of variance (Statgraphics; Statistical Graphics, Rockville, MD, U.S.A.) was used to de- termine whether crosslinking and time significantly (p < 0.05) affected the rate of collagen fiber degra- dation. Multiple t tests (corrected for the total num- ber of groups) were used to determine significant differences (p < 0.05) in this rate for different exper- imental groups.

In addition, diameters of individual fibers were

J Orthop R q VoL 12, No. 1, 1994

INCORPORATION OF COLLAGEN PROSTHESES IN BONE 131

FIBER DIAMETER (urn) FIG. 2. Degradation rates of reconstituted collagen fiber in bone. A The number of intact fibers decreased with time. By 8 weeks, the dehydrothermal-cyanamide (DHTC) crosslinked fibers were completely degraded and only 10% of the glutaraldehyde (GLUT) crosslinked fibers were intact. *GLUT value significantly greater than DHTC value (p < 0.05). B: The collagen fiber diameter histo- gram resembled a normal distribution (fitted curve) at 0,3, and 4 weeks. The histogram shown is for GLUT crosslinked fiber diameters at 3 weeks: n = 300; mean = 60 pm; SD = 10 pm.

measured (as many as 50 fibers per slide) with the use of a calibrated eyepiece at a magnification of ~ 1 0 0 . Fiber diameter histograms were generated and were fitted to various distribution functions with the use of Statgraphics software. Kolmogorov-Smirnov tests were used to determine the distribution func- tion that most closely fit the experimental data.

RESULTS Gross Observations

The animals tolerated the operative procedures well; all recoveries were uneventful. At death, all of the skin wounds appeared to be well healed. The implant site was easily identifiable by the presence of the nonresorbable suture on the lateral femoral

J Orthop Res, Val. 12, No. 1, 1994

132 M . G. D U N N ET AL.

periosteum, proximal to the bone tunnel, with soft tissue emerging from the bone at the implantation site.

Collagen Prosthesis Degradation Rates Degradation of the implanted collagen fibers be-

gan on the periphery of the prosthesis and proceeded

centrally. The number of intact fibers decreased rap- idly with time from 0-4 weeks (Fig. 2A) for the DHTC and GLUT crosslinked prostheses. At 3 weeks, only 60% of the GLUT and 40% of the DHTC fibers remained intact. By 8 weeks, the DHTC crosslinked fibers were completly degraded. The GLUT cross- linked fibers also were rapidly degraded in the bone

FIG. 3. Light micrographs of cross sections of the prosthesis in the bone tunnel at 3 weeks after implantation. Bar = 200 km. A: Glu- taraldehyde crosslinked collagen fibers (CF) were surrounded by new bone (NB). B At higher magnification, new bone approached the peripheral collagen fibers, which appeared to be intact. C: New bone also approached the dehydrothermal-cyanamide crosslinked collagen fibers, which were in various stages of degradation.

J Orthop Res, Vol. 12, No. 1, 1994

INCORPORATION OF COLLAGEN PROSTHESES I N BONE 133

tunnel, with about 10% of the fibers intact at later time periods (there was no significant difference be- tween 8 and 12 week values). The mononuclear cells were most frequently found degrading the collagen fibers; multinucleated giant cells occasionally were found.

Although the total number of intact fibers de- creased with time, the mean fiber diameter and the shape of the distribution did not change significantly from 0-4 weeks after implantation, due to fibers re- maining intact in the center of the prosthesis. The measured fiber diameter distributions fitted most closely to a normal distribution function (Fig. 2B), with mean values nearly 60 pm. At 8 and 12 weeks, too few fibers remained for comparison of diameter distributions.

Bone Growth into Collagen Prostheses At 3 and 4 weeks, the prostheses were infiltrated

by fibrous tissue and inflammatory cells. At the pe- riphery, new bone surrounded the circumference of the prostheses, approaching individual collagen fi- bers (Fig. 3A). The GLUT crosslinked fibers ap- peared intact at their edges (Fig. 3B); the DHTC crosslinked fibers were degraded at their edges by mononuclear inflammatory cells (Fig. 3C). The diam- eter of the “hole” remaining in the bone (measured from histological slides) was approximately one-half of the initial diameter of the tunnel.

At 8 weeks, the DHTC crosslinked prostheses were completely degraded and the inflammatory re-

sponse had subsided. The limbs with a GLUT cross- linked prosthesis had a prolonged inflammatory re- sponse to the remaining fibers. New bone continued to advance toward the center of the defect (Fig. 4A). Polarized light microscopy revealed that the ingrow- ing trabecular bone formed a continuous network with the central fibrous tissue (Fig. 4B). At 12 weeks, the cancellous bone defects continued to close and in several samples the original defect site was com- pletely healed. Histological sections parallel to the longitudinal axis of the prosthesis showed fibrous tissue exiting through a funnel-shaped opening at the lateral cortex. In this region, the bonehoft-tissue in- terface was distinguished by a tidemark at 8 and 12 weeks (Fig. 5) .

DISCUSSION Intraosseous incorporation or biological fixation

of the collagen prosthesis is required for success- ful ACL reconstruction. The purpose of this study was to characterize intraosseous degradation rates and tissue growth into composite collagen prosthe- ses designed for ACL reconstruction. This bone im- plantation model was used by Arnoczky et al. (1) to compare biological fixation of various ligament pros- theses in bone. The surgical procedure is reproduci- ble and is appropriate for evaluation of prosthesis degradation and bone ingrowth, but it does not in- clude many variables associated with ACL recon- struction (1). This nonloaded model represents the early postoperative period of immobilization re-

J Orthop Res, Vol. 12, No. 1, 1994

134 M . G. D U N N ET A L .

quired for a degradable, tissue-inducing ACL re- placement device.

Collagen Prosthesis Degradation Rates The composite collagen prostheses contain very

thin (60 pm in diameter) collagen fibers, each with a high ratio of surface area to volume. Because the

degradation rate of collagen is dependent on the surface area, these prostheses containing many thin fibers degrade much more rapidly than does an equal mass of bulk collagen.

The degradation rate of reconstituted collagenous biomaterials also is greatly influenced by the cross- link density (20). GLUT treatment results in an in-

FIG, 4. Light micrographs of cross sections of the prosthesis in the bone tunnel at 8 weeks after implantation. Bar = 200 pm. A: The dehydrothermal-cyanamide crosslinked fibers were completely degraded and replaced by fibrous tissue (FT) centrally. New bone (NB) continued to advance toward the center of the prosthesis. B: Polarized light micrograph of the glutaraldehyde crosslinked pros- thesis at higher magnification showed a central network of fibrous tissue intimately attached to the new bone. The broken line delineates the bonelsoft-tissue interface.

J Orthop Res, Vol. 12, No. I , 1994

INCORPORATION OF COLLAGEN PROSTHESES IN BONE 135

crease in intermolecular crosslinking, an increase in tensile strength, and a decrease in the degradation rate, but cytotoxicity is a major concern (11). In this study, the GLUT crosslinked collagen fibers de- graded rapidly in the bone tunnels: only 10% of the GLUT fibers remained intact at 12 weeks. In con- trast, in an Achilles tendon implantation model in

rabbits, 75% of the GLUT crosslinked collagen fi- bers remained intact at 1 year after implantation (14). The rapid degradation rate of GLUT cross- linked implants in bone may be due to a more in- tense local inflammatory response. Bone has greater vascularity and more protease-producing cells than do soft-tissue implantation sites (3).

FIG. 5. Light micrographs of longitudinal sections of the prosthesis within cortical bone at 8 weeks after implantation. Bar = 200 pm. A The new bone (NB) was funnel-shaped near the lateral cortex, with fibrous tissue (FT) deposited in place of the completely degraded dehydrothermal-cyanamide-treated collagen prosthesis. B: At higher magnification, the bonelsoft-tissue interface had a tidemark (arrowheads) separating the fibrous tissue from the new bone.

J Orthop Res. Vol. 12, No. I, 1994

136 M. G. DUNN ET AL.

DHTC crosslinking of collagenous biomaterials produces synthetic ester and amide crosslinks with- out introducing cytotoxic aldehydes (20). DHTC treatment results in lower tensile strength and a more rapid rate of degradation than GLUT treat- ment. In this study, DHTC crosslinked collagen was completely degraded by 8 weeks, which is consistent with previous studies in soft-tissue sites (6,9). DHTC crosslinking is superior to GLUT crosslinking be- cause biocompatibility is improved and the resorp- tion rate is increased. Rapid resorption results in rapid tissue ingrowth, especially in the soft-tissue sites (6,9).

Bone Growth into Collagen Prostheses Even in the absence of an implanted biomaterial,

bone repair is a complex process that involves in- flammation; infiltration and proliferation of repair cells; matrix deposition and calcification; and remod- eling. Repair of cylindrical bone tunnels starts at the periphery and proceeds centripetally by intramem- branous or endochondral ossification, or both. The healing rate and the quality of healing are influenced by many factors, including the initial diameter of the defect (8). In this study, a 2.8 mm diameter defect was created, to be consistent with the tunnel size we used for earlier ACL reconstructions in rabbits (6).

The healing response of bone can be further com- plicated by the presence of an implanted biomaterial (3). Collagenous biomaterials (usually in a sponge form) can enhance bone healing by providing a scaf- fold to promote osteoconduction of new bone into the defect (5) . Our qualitative results suggest that bone heals and bone ingrowth is rapid in the pres- ence of degradable composite collagen prostheses and that the rate of bone ingrowth is influenced by the degradation rate of the prosthesis.

The Cortical Bone/Soft-Tissue Interface Cortical bone healing resulted in a funnel-shaped

indentation at the lateral cortex, with a tidemark found between the new bone and the fibrous tis- sue emerging from the bone. Similar tidemarks are found in certain ligament insertion sites, where fi- brocartilage provides a transition between soft tis- sue and bone (2). Several studies have reported that tendon (7) and ACL autografts (4) and allografts (12,16) were well incorporated in bone, and in the long term, their insertion sites matured to resemble fibrocartilage. In other studies, however, no direct contact has been found between implanted autoge- nous tendons and the bone perimeter of the drill

defect (17). Xenograft tendon, which was densely crosslinked with GLUT to reduce antigenicity, was poorly incorporated in bone (1,17), and this re- sulted in low interfacial strength (15). It is not clear whether fibrocartilage consistently redevelops dur- ing ACL graft incorporation (2,21) or whether its presence is advantageous in the long term.

CONCLUSIONS Previous studies suggested that degradable pros-

theses containing reconstituted collagen fibers (13) are potentially useful as tissue-inducing implants for reconstruction of the Achilles tendon (9,14,19) and the ACL (6). The results of this study suggest that tissue growth into composite collagen prostheses may provide biological fixation in surgically created bone tunnels. Healing is more rapid in rabbits than in humans, so prosthesis degradation and tissue in- growth in humans probably would require longer time periods than those reported here. In addition, the effects of mechanical loading on tissue ingrowth and the rate of prosthesis degradation in bone need to be investigated.

Acknowledgment: The authors wish t o thank Kevin S. Weadock, Ph.D., and Frederick H. Silver, Ph.D., for help- ful discussions. This study was supported by a General Research Support Grant from UMDNJ and by Grant 90-0195 from the Whitaker Foundation.

1.

2. 3.

4.

5.

6.

7.

8.

9.

REFERENCES Arnoczky SP, Torzilli PA, Warren RF, Allen AA: Biologic fixation of ligament prostheses and augmentations: an eval- uation of bone ingrowth in the dog. Am J Sports Med

Benjamin M, Evans EJ: Fibrocartilage.JAnar 171:l-15,1990 Boyde A, Jones SJ: Bone modelling in the implantation bed. J Biomed Muter Res 19:199-224,1985 Clancy WG Jr, Narechania RG, Rosenberg TD, Gmeiner JG, Wisnefske DD, Lange TA: Anterior and posterior cruci- ate ligament reconstruction in rhesus monkeys: a histologi- cal, microangiographic, and biomechanical analysis. J Bone Joint Surg [Am] 63:1270-1284,1981 Deporter DA, Komori N, Howley TP, Shiga A, Ghent A, Hansel P, Parisien K: Reconstituted bovine skin collagen enhances healing of bone wounds in the rat calvaria. Calcif Tissue Int 42:321-325,1988 Dunn MG, Tria AJ, Kato YP, Bechler JR, Ochner RS, Zawadsky JP, Silver FH: Anterior cruciate ligament re- construction using a composite collagenous prosthesis: a biomechanical and histologic study in rabbits. Am J Sports Med 20507-515,1992 Forward AD, Cowan RJ: Tendon suture to bone: an experi- mental investigation in rabbits. J Bone Joint Surg [Am]

Friedenberg ZB, Lawrence RR: The regeneration of bone in defects of varying size. Surg Gynecol Obstet 114:721-726, 1962 Goldstein JD, Tria AJ, Zawadsky JP, Kato YP, Christiansen D, Silver FH: Development of a reconstituted collagen ten-

16:106-112,1988

451807-823,1963

J Orthop Res, Val. 12, No. I , 1994

I N C O R P O R A T I O N OF C O L L A G E N PROSTHESES IN B O N E 137

don prosthesis. J Bone Joint Surg [Am] 71:1183-1191,1989 10. Holden JP, Grood ES, Butler DL, Noyes FR, Mendenhall

HV, Van Kampen CL, Neidich RL: Biomechanics of fascia lata ligament replacements: early postoperative changes in the goat. J Orthop Res 6639-647, 1988

11. Huang-Lee LLH, Cheung DT, Nimni ME: Biochemical changes and cytotoxicity associated with the degradation of polymeric glutaraldehyde derived crosslinks. J Biomed Mu- ter Res 24:1185-1201, 1990

12. Jackson DW. Grood ES, Arnoczky SP, Butler DL, Simon TM: Freeze dried anterior cruciate ligament allografts: pre- liminary studies in a goat model. Am J Sports Med 15:295- 303,1987

13. Kato YP, Christiansen DL, Hahn RA, Shieh S-J, Goldstein JD, Silver FH: Mechanical properties of collagen fibers: a comparison of reconstituted and rat tail tendon fibres. Bio- materials 10:38-42,1989

14. Kato YP, Dunn MG, Zawadsky JP, Tria AJ, Silver FH: Re- generation of Achilles tendon with a collagen tendon pros- thesis: results of a one-year implantation study. J Bone Joinf Surg [Am] 73561-574.1991

15. Montgomery RD, Milton JL, Terry GC, McLeod WD, Mad- sen N Comparison of over-the-top and tunnel techniques for anterior cruciate ligament replacement. Clin Orthop 231:144-153, 1988

16. Shino K, Kawasaki T, Hirose H, Gotoh I, Inoue M, Ono K: Replacement of the anterior cruciate ligament by an al- logeneic tendon graft: an experimental study in the dog. J Bone Joint Surg [Br] 66:672-681,1984

17. Shoemaker SC, Rechl H, Campbell P, Kram HB, Sanchez M: Effects of fibrin sealant on incorporation of autograft and xenograft tendons within bone tunnels: a preliminary study. Am J Sports Med 17:318-324,1989

18. Silver FH, Tria AJ, Zawadsky JP, Dunn MG: Anterior cruci- ate ligament replacement: a review. J Long-Term Effects Med Imp1 1:135-154,1991

19. Wasserman AJ, Kato YP, Christiansen D, Dunn MG, Silver FH: Achilles tendon replacement by a collagen fiber pros- thesis: morphological evaluation of neotendon formation. Scanning Microsc 3:1183-1200,1989

20. Weadock K, Olson RM, Silver FH: Evaluation of collagen crosslinking techniques. Biomater Med Devices Artif Or- guns 11:293-318,1983-1984

21. Woo S, Maynard J, Butler D, Lyon R, Torzilli P, Akeson W, Cooper R, Oakes B: Ligament, tendon, and joint capsule insertions to bone. In: Injury and Repair of the Musculo- skeletul Soft Tissues: Workshop, Suvannuh, Georgia, June 1987, pp 133-166. Ed by SL-Y Woo and JA Buckwalter. Park Ridge, Illinois, American Academy of Orthopaedic Surgeons, 1988

J Orthop Rer, Vol. 12, No. I , 1994