Osteoarthritis and Cartilage (1995) 3, 47-59 © 1995 Osteoarthritis Research Society 1063-4584/95/010047 + 13 $08.00/0

OSTEOARTHRITIS and

CARTILAGE Chondrocyte- laden col lagen scaffolds for resurfacing extens ive

articular cartilage defects BY ANDREW E. SAMS AND ALAN J. NIXON

Comparative Orthopaedics Laboratory, Department of Clinical Sciences, College of Veterinary Medicine, Cornell University, Ithaca, New York 14853, U.S.A.

Summary

Chondrocyte-collagen composites were evaluated for resurfacing of large articular defects. Isolated chondrocytes were cultured in expanded collagen scaffolds for 7-10 days to provide a composite containing 3.6 × 104 cells/mm 3. The graft was transplanted into 15 mm full thickness articular defects in the femoropatellar joint of 12 horses using arthroscopic techniques. Ungrafted defects in the opposite femoropatellar joint served as controls. Synovial fluid, clinical progress and pain responses were evaluated in groups of 6 horses over 4-month and 8-month periods. Following termination, gross, histochemical and histologic evaluations of the repair tissues and synovial membrane were performed. Arthroscopic defect debridement and chondrocyte implantation resulted in minimal post-operative effusion or pain, and synovial fluid constituents were not significantly different in grafted and ungrafted joints. Gross differences in grafted defects were not apparent. Increased chondrocyte numbers and cartilage histochemical staining were evident in the deeper layers of grafted defects, whereas ungrafted defects were almost entirely fibrous tissue. The surface layers of grafted defects were fibrous tissue. There were no synovial fluid cellular responses, synovial membrane histiocytic reaction or subchondral bone infiltrates to suggest immune-related reaction to the allograft cells. Chondrocyte- collagen grafts were arthroscopically implanted and resulted in improved cartilage healing in extensive defects. However, the structural organization of the surface layers was inadequate and suggested poor long-term durability. Key words: Cartilage, Resurfacing, Chondrocyte, Collagen, Transplantation.

Introduction

NORMAL a r t i cu l a r ca r t i l age is a h igh ly organized, n o n h o m o g e n o u s t i ssue composed p r imar i ly of chondrocy te s , col lagen, p ro teog lycans , and non- co l l agenous prote ins . The p ropo r t i on of cells to ma t r i x in m a t u r e ca r t i l age is low compared to m a n y tissues. The immedia t e b iomechan ica l prop- er t ies of car t i lage , such as durabi l i ty , s t ress dissi- pa t ion and stiffness, are dependen t upon the s t r u c t u r e and type of co l lagen and the proteo- g lycan conf ined wi th in it. This h ighly o rdered s t r u c t u r e of a r t i cu l a r ca r t i l age gives i t bio- mechan i ca l p roper t i e s t ha t u n f o r t u n a t e l y have no t been dupl ica ted in any c u r r e n t l y ava i lab le r e p l a c e m e n t [1].

Superf ic ia l defects in a r t i c u l a r ca r t i l age do no t hea l to any s igni f icant degree [2-5]. Defects t ha t p e n e t r a t e the zone of calcif ied ca r t i l age hea l to some extent , bu t the new t issue is p r imar i ly f ibrous t i ssue in the sur face layers and f ib rocar t i l age in the deeper zones [6]. The r epa i r t issue t h a t does

Submitted 9 November 1993; accepted 17 June 1994. Supported by grants from the Harry M. Zweig Memorial Fund for Equine Research. Address correspondence to: Dr Nixon.

form provides l i t t le r e s i s t ance to obl ique compres- sive forces [7], and over t ime may also d e t e r i o r a t e or ded i f fe ren t i a t e to f ibrous t i ssue [6]. Resu r f ac ing t echn iques are r equ i red t h a t will p romote no t on ly a morpho log ica l ly o rgan ized hya l ine surface , bu t one capable of long- term func t i on w i t h o u t dis- i n t e g r a t i o n or ce l lu la r ded i f fe ren t ia t ion . F u r t h e r , such t echn iques need to c r ea t e minimal dono r s i te morb id i ty and be p rac t i ca l ly appl icable to c l in ica l diseases. S igni f ican t a r t i c u l a r damage resu l t s in o s t eoa r th r i t i s (OA), which is progress ive and a ma jo r cause of morbidi ty . Biologica l c a r t i l age r e su r f ac ing t echn iques i n t e r c e p t the p rogress ion of focal ca r t i l age loss to p r e v e n t such i r revers ib le OA.

Surgi~cal g ra f t ing p rocedures to resur face a r t icu- la r ca r t i l age defects have no t been p a r t i c u l a r l y successful . Pe r ios t ea l graf t s resu l ted in p o o r re- sults when appl ied to la rge defects [8, 9]. Auto- genous o s t eochondra l graf ts have shown some promise in r e sea rch models. However , the d o n o r a r t i c u l a r sur face is d i s rup ted and degene ra t i ve jo in t d isease can develop i f the graf t is ex tens ive [10,11]. F re sh o s t eochond ra l a l lograf ts i nduce excessive immune responses , l eading to f a i lu re [11, 12]. F reez ing decreases the i m m u n o g e n i c i t y of

47

48 Sams and Nixon: Chondrocyte-laden collagen scaffolds

the grafts [13]; however, the chondrocytes are severely damaged, which decreases the uti l i ty of the graft. Sternal carti lage autografts have shown promise in studies in animal models, but long-term durabili ty has been less satisfactory, and there is still a need for an arthroscopic technique for insertion before clinical use can be expected [14, 15].

Chondrocyte allografts combined with fibrin- based vehicles [16], hyaluronic acid [17], or colla- gen gels [18] have been used to successfully resurface full thickness art icular cartilage defects in rabbits and chickens. Studies on the isolation and cryopreservation of equine chondrocytes have allowed the pursuit of resurfacing techniques in much larger animal models, using allograft chondrocytes implanted in various vehicles [19]. Chondrocytes secured in 12-mm defects by a fibrin matrix resulted in organized repair tissue contain- ing 62% type II collagen [20,21]. The fibrin matrix provided 3-dimensional support, prevented chondrocyte dedifferentiation toward fibroblasts, and protected the cells from the host immune response [22]. Although allograft chondrocytes are antigenic, the surface antigens can be exten- sively masked if the cells are allowed to produce their own matrix by incubation prior to implan- tation [23, 24]. Fibrin vehicles make incubation prior to implantation unnecessary. However, fibrin is difficult to a t tach in the graft site [20]. Hyaluronan has also been used as a vehicle to implant cultured chondrocytes into small ar t icular defects in laboratory animals, with the same 3- dimensional benefits as fibrin-based vehicles [17]. However, hyaluronate compounds have no ad- hesive qualities and little stiffness to assist in anchoring the chondrocyte-laden vehicle into large defects.

Purified bovine type I collagen scaffolds have been used in vitro to culture murine chondro- progenitor cells, and have been embedded with chondrocytes and implanted to successfully re- surface art icular defects in rabbits [25, 26]. The collagen scaffolds provide stiffness to the implant, allowing it to be pressed securely into the defect. Equine chondrocytes embedded in collagen scaffolds and cultured in vitro have been shown to survive and synthesize carti lage matrix products [27]. The development of a collagen vehicle for arthroscopic chondrocyte implantation of exten- sive full thickness defects in horses would provide a suitable and challenging model to confirm and expand the positive results of these composites in laboratory animals. The objectives of this s tudy were to assess the clinical progress, synovial fluid response, and the gross and histologic appearance

of the repair t issue following chondrocyte- collagen composite t ransplantat ion.

Mater ia l s a n d m e t h o d s

CELL CULTURE

Donor art icular carti lage was aseptically harvested from the joints of a 3- and a 6-month-old foal. During the harvest procedure the carti lage was lavaged with Gey's balanced salt solution (GBSS; GIBCO-Life Technologies INC., Grand Island, NY). The carti lage was kept ~n GBSS on ice, weighed, t ransported to a laminar flow hood and then diced into approximately 1-mm squares. The carti lage matrix was then digested to release the chondrocytes as previously described [19]. Briefly, the diced carti lage waS added to filter sterilized 0.075% (w/v) collagenase (type CLS 1 from Clostridium histolyticum, 169U/mg, Wor- thington Biochemicals, FreeholdS, NJ) in Ham's F-12 medium (GIBCO-Life) supplemented with 50#g ascorbic acid/ml, 30pg ~-ket0glutaric acid (as the monopotassium salt)/ml (GIBCO-Life), 300 pg of L-glutamine/ml (GIBCO-Life)i. 10% fetal bovine serum (GIBCO-Life), 500 units sodium peni- cillin/ml (GIBCO-Life) and 500pg streptomycin sulfate/ml (GIBCO-Life). Twenty-five mM HEPES (GIBCO-Life) was used to buffer the solutions.

The cartilage and collagenase solution were placed in an erlenmeyer flask and stirred on a magnetic stirrer for 18 h at 37°C. The cell enzyme mixture was then filtered through four layers of sterile gauze and one layer of nylon mesh (44 #m) to remove the extracellular matrix debris. The filtrate containing chondrocytes was then centri- fuged at 260g for 10rain. The supernatant was discarded and the cell pellet resuspended. An ali- quot of cell suspension was removed for a cell count and viabili ty determination u s ing a dye exclusion test (erythrosin B; Sigma Chemical Co., St. Louis, MO). The freshly isolated cells were resuspended in supplemented Ham's F-12 contain- ing 10°//o DMSO, and l ml aliquots containing 10 × 106 cells/ml were placed in cryotubes and frozen at a rate of - 1°C per minute to - 80°C. The chondrocytes were then stored at -196°C until needed.

COLLAGEN SCAFFOLD P R E P A R A T I O N

The porous collagen matrices were commercially available for investigational use and were supplied as sterile packages of six 16-mm diameter discs (Colla-Tec Inc., Plainsboro, NY). The matrices were predominantly composed of type I collagen

Osteoarthrit is and Cartilage Vol. 3 No. 1 49

purified from bovine Achilles ' tendon. The mater ia l was processed to el iminate contac t sensit ization or tissue i rr i ta t ion, and yet main ta in its triple helical configurat ion [28]. Left in situ it is progressively and completely resorbed from implantat ion sites. Previous light microscopic studies describe the open mesh porous ne twork of these collagen sponges [27].

Six days prior to incorpora t ion of the cells into the collagen scaffold, the suspensions were thawed, and number and viabil i ty determined using the erythrosin-B exclusion method. The cells were cul tured in 25- or 75-cm 2 flasks with 10 ml or 30 ml of supplemented Ham's F-12, respectively. A high seeding density of 4 x 10 ~ cells/cm 2 was used in the flasks. The cells were incubated in 5~/o CO2 and 90% humidi ty for 6 days. Supplemented Ham's F-12 media was replaced on day 3, or sooner if the media was exhausted. Between the 6th and 9th day, the chondrocyte monolayer was trypsinized using 0.05% trypsin (GIBCO-Life), a cell count was per- formed, and 8 × 10 ~ cells were resuspended in 0.25 ml of supplemented Ham's F-12 and drawn into the interst ices of a 16-mm diameter collagen scaffold using a gentle vacuum. The concent ra t ion of cells in the scaffold was 3.6 × 104 cells/mm 3. The disc shaped chondrocyte composite was then cov- ered on both sides by 0.25 ml neutralized, chilled (4°C), liquid collagen (vitrogen; Collagen Corpor- ation, Palo Alto, CA). Vi t rogen is a 99.9% pure collagen solution, derived from pepsin-solubilized bovine dermal collagen, and dissolved in 0.012 N

HC1 to establish a 3 mg/ml collagen solut ion with a pH of 2. This product is 95-98% type I collagen with the remainder being type III collagen. The collagen-coated disc was a l l o w e d t o polymerize in a 16-mm 12-well plate in an incubator for 30 min, and was then covered with l m l of supplemented Ham's F-12. The chondrocyte- laden collagen discs (Fig. 1) were incubated for 6-12 days pr ior to implantat ion in joints, with media change every 3 days.

S U R G I C A L D I S K I M P L A N T A T I O N

Twelve normal horses, weighing between 410-500 kg, and between the ages of 2-8 years, were used for this study. Clinical examinat ion verified there were no significant joint effusions or hind- limb lameness present before the animals were included in the trial.

For the implantat ion procedure the horses were anesthetized, and placed in dorsal recumbency. Both femoropatel lar joints were prepared for asep- tic surgery and the left or r ight leg was chosen at random to be the site for chondrocyte graft. The cont ra la tera l limb served as the sham control. A sample of joint fluid was obtained from each femoropatel lar joint. The joint was distended with saline and an ar throscope inserted into the distal aspect of the joint between the middle and la teral patel lar l igaments as described in s tandard texts [29]. The ar t icu lar surfaces were examined to deter- mine there were no previous cart i lage lesions or

FIG. 1. A collagen scaffold laden with chondrocytes (small arrows). The fibres of the scaffold (large arrows) and the coating of near transparent polymerized liquid collagen on the surface (open arrow) can be seen (H & E; × 312).

50 Sams and Nixon: Chondrocyte- laden col lagen scaffolds

FIG. 2. Arthroscopic approach to the equine lateral femoral trochlea. The 16mm cannula for instrument access and graft implantation is shown.

aligned 1-mm diameter holes were drilled 10 mm apart to a depth of 20 mm into the subchondral bone using a 1.14-mm K-wire. The collagen disc containing chondrocytes was then removed from the incubator and inserted down the cannula into the joint. The disc was gently manipulated into position and pressed into the defect to form a smooth surface approximately level with the per- imeter cartilage. It was then secured by insert ing a 1.1mm × 15mm polylactic acid tack (Biofix Minitack, Biocon Ltd, Tampere, Finland) into each pre-drilled hole. The tacks were driven into the bone to hold the disc in place wi thout compressing the disc substance. The gas was removed from the joint and lactated Ringer 's solution :was used to i rr igate the cart i lage and chondrocyte-polymer surfaces and interfaces. The residual fluid was expressed from the joint and the skin closed with nylon sutures. The control hindlimb was examined in an identical manner and a cart i lage lesion made using the same technique. The axSal holes were drilled into the subchondral bone as in the exper- imental side; however, no chondroc~te-col lagen disc was placed in the control side defect. The joint was flushed and closed as on the experimental side. Procaine penicillin (22 000 IU/kg i.m.) was adminis- tered 2 h prior to surgery and 6 h post-operatively. Phenylbutazone (4.4mg/kg p.o.) and detomidine (16 # g/kg i.m.) were given post-operatively as anal- gesics. Synovial fluid samples were drawn at the time of surgery, and on days 4, 7, 14, 30 and 120 or 240 following surgery .

synovial inflammation. An 18-gauge, 8.9-cm spinal needle was placed through the skin approximately 4 cm distal to the apex of the patella and into the joint between the middle and lateral patel lar liga- ments, to locate the central aspect of the lateral t rochlea of the femur. An 18-20 mm incision was made through the skin, subcutaneous tissues, and joint capsule at the point indicated by the spinal needle. A 16-ram internal diameter cannula was placed into the joint, centered on the lateral troch- lear ridge of the femur and 2 cm distal to the apex of the patella. A modified spade bit 16mm in diameter was used to remove a full-thickness layer of cart i lage and 1 mm of subchondral bone (Figs 2 & 3). The lactated Ringer 's solution used to distend the joint was replaced by sterile helium gas and the subchondral bed of the cart i lage lesion dried by inserting several lint-free sponges. Two axially

FIG. 3. Intraoperative arthroscopic view of a 16mm defect in the equine lateral femoral trochlea prior to graft implantation. Two holes can be seen (arrows) where tacks that anchor the graft were placed. Control defects appeared as this photograph depicts.

Osteoarthrit is and Cartilage Vol. 3 No. 1 51

P O S T - O P E R A T I V E PROTOCOL

Four weeks post-operatively, walking exercise for 10rain daily was commenced and increased each week by 5min daily, over a period of 4 additional weeks. The horses were then allowed out on a half acre pasture unti l the terminat ion of the experiment. The horses were divided randomly into two equal groups; one group of horses were euthanized at 120 days post-operatively and one group at 240 days post-operatively. Immediately prior to euthanasia another lameness exam was performed on each horse. The degree of lameness was scored using a degree/5-scale system.

GROSS PATHOLOGICAL E X A M I N A T I O N

The horses were euthanized and the stifle joints removed without opening the joint capsule. Synovial fluid was aspirated, quantitated, and ana- lyzed as described later and the gross anatomic appearance of the healing cartilage lesions recorded by photography. The general appearances of the repair tissue and surrounding tissue of the experimental and control joints were evaluated. The defect tissue was examined for homogeneity, level of repair tissue relative to the surrounding carti lage surface, color, and junct ion with the surrounding cartilage. The surrounding tissue was examined for adhesions, fibrillation, and color.

SAMPLE HARVESTING AND P R E P A R A T I O N

A section of synovial membrane from the area lateral to the cannula penetrat ion site in the joint capsule was harvested and fixed in 10°//o buffered formalin. A sterile bone saw was used to remove a rectangular block of carti lage and bone from each femur, encompassing the cartilage lesion and extending 1.5 cm beyond the edge of the defect into normal cartilage and I cm into the subchondral bone. The circular repair defect was divided parasagit tal ly into thirds. The repair tissue in the axial third was harvested and frozen for biochemi- cal analysis. The central third of the sample was harvested and fixed in 10% buffered formalin.

THIN SECTION HIGH DETAIL RADIOGRAPHY

High detail radiography was performed on the 3-mm bone slabs, using a low energy X-ray field (Faxitron X-ray, Hewlett Packard, McMinnville, OR) and high detail radiographic film, to evaluate the response of the subchondral t rabecular bone to the implan t . The samples were then placed under vacuum in citrate buffered 22% formic acid for

demineralization. Radiographic 'examination every 3 days determined when demineralization was com- plete. The demineralized samples were placed in 10% buffered formalin unti l embedded in paraffin for histology.

HISTOLOGY

Decalcified specimens were embedded in paraffin and 6-#m sections made in a sagittal plane for histology and histochemistry. Hematoxylin & eosin-stained sections were used for morphologic evaluation, and sections stained with safranin O (counter stained with fast green), and alcian blue (counter stained with nuclear fast red), were used for histochemical evaluation. The morphology of the cell types in the repair tissue was characterized as fibrocytic, fibrochondrocytic or chondrocytic. Repair tissue was characterized as fibrous, fibro- carti laginous or hyal ine cartilage-like. The inten- sity of the matrix staining with the cationic stains (safranin 0 and alcian blue) was characterized and used as an indicator of proteoglycan production.

The synovial membrane samples were embedded in paraffin, and sectioned at 6 pm prior to s taining with hematoxylin & eosin. The sections were then read blind at × 125 magnification by the same observer. Villus proliferation, white blood cell infiltration, vascular proliferation, perivascular connective tissue thickness, subintimal fibrosis and mineralization were evaluated. A score from 1 (absent) to 5 (severe) was assigned to quant i ta te changes in the villi.

SYNOVIAL FLUID ANALYSIS

Red blood cell (RBC) and white blood cell (WBC) numbers were determined using a Coulter counter (Coulter IV, Hialeah, FL). Synovial fluid protein was quantified using a coomassie blue spectro- photometric analysis [30] . Absorption was measured at 595nm. The concentrat ion of hyal- uronan in the synovial fluid was determined using an alcian blue precipitation spectrophotometric analysis [31], with absorbance measured at a wave- length of 620 nm. Cytology of the synovial fluid was performed by an experienced clinical pathologist.

STATISTICS

Lameness score data from control and grafted limbs was compared at each time period using a Student 's paired t-test. Comparison of the histo- logic and synovial fluid data from the chondrocyte grafted stifles and the controls at each time period was done using the Student 's paired t-test, since

52 S a m s and Nixon: C h o n d r o c y t e - l a d e n co l l agen scaffo lds

FIG. 4. Gross appearance of a pair of 240 day specimens with repair tissue from the lateral trochlea. Both the control (left) and grafted (right) have irregular repair tissue in the defect which does not fill either defect completely. The junction between the repair tissue and surrounding cartilage is easily discernable in both specimens.

each animal served as its own control . Compar i son of the changes in da ta over the sampl ing t ime periods was then per fo rmed by repea ted measures analysis of var iance . S igni f ican t i n t e rac t ions of t r e a t m e n t (graft or control) , or time, were assessed. The level of s ignif icance was P ~< 0.05.

R e s u l t s

CLINICAL E X A M I N A T I O N

There were no d i f ferences in the sever i ty of j o in t effusion or the degree of lameness be tween graf ted and cont ro l jo in ts over the course of the s tudy. At t e rmina t ion , all horses excep t one were sound at a walk and trot . Hind limb flexion for 1 .5min produced a low-grade lameness in 13 jo in ts of 11 horses.

All incis ions except one hea led by p r imary inten- t ion; par t i a l wound deh i scence occu r r ed in one and hea led wi thou t compl ica t ion . The re were no sig- n i f icant di f ferences (P > 0.05) in lameness, effusion and wound hea l ing pa r ame te r s be tween exper- imenta l and control , and be tween 4- and 8-month- old groups.

GROSS PATHOLOGIC EXAM

The genera l appea rance of the r epa i r t issue and s u r r o u n d i n g ca r t i l age was s imilar in expe r imen ta l and con t ro l jo in ts at each t ime period. The repa i r t i ssue in most jo in ts was opaque and white , wi th an u n e v e n sur face t ha t r a r e ly filled the defect to the level of the su r round ing t issue (Fig. 4). Bo th stifle jo in ts in one horse had a smooth whi te r epa i r t i ssue t ha t comple te ly filled the defect . A c lear d e m a r c a t i o n l ine was ev ident be tween the r epa i r t issue and the su r round ing no rma l car t i lage. F ib r i l l a t ion of the ca r t i l age s u r r o u n d i n g the defect was no t no ted in any joint . T h e r e were no gross lesions on the a r t i cu l a r sur face of the patel la , or ev idence of synovia l p ro l i fe ra t ion . The synov ium loca ted at the c a n n u l a en t ry site had a f ibrous scar p resen t wi th c o n t r a c t i o n of the s u r r o u n d i n g tis- sues bu t no o the r abnormal i t i e s were noted.

HISTOLOGIC AND HISTOCHEMICAL E X A M I N A T I O N

Synovia l m e m b r a n e h is to logic scores are pre- sen ted in Table I. The re were no s igni f icant differ- ences in any of the ca tegor ies eva lua ted . The

O s t e o a r t h r i t i s and Cart i lage Vol . 3 No. 1 53

synovial membrane intimal layer was one to two cell layers thick (Fig. 5). Few white blood cells were present in the subintimal layer. Fibroplasia and subintimal mineral izat ion were rare in the experimental and control joints.

Grafted and ungrafted defect repair tissue was either fibrous or fibrocartilaginous tissue (Fig. 6). Grafted defects had greater numbers of chondro- cytic cells in the base of the defect compared to nongrafted controls (Fig. 7). However, fibroblasts predominated in the surface layers of both grafted and nongrafted defects. Chondrocytes in the deeper zones of the grafted defects occurred in small groups, and on occasion in columns. There were no accumulations of WBCs in the repair tissue or the subchondral bone of any of the defects. Islands of chondrocytes occurred in the lower 20-30~/o of the defects, and these had strong uptake of safranin O (Fig. 8). Safranin-stained tissue in the grafted defects was more extensive than in nongrafted defects. The a t tachment of the experimental repair tissue to the surrounding cartilage was frequently better than the attach- ment of control repair tissue. There was less cleft- ing as evidence of demarcation in the experimental defects. The surface of the defects was rough and uneven in appearance at high power magnifi- cation• Minor cloning of chondrocytes had occurred in the surrounding carti lage at the edge of the defect• A mild to moderate loss of safranin metachromasia was apparent in this cartilage in the control and experimental sections.

SYNOVIAL F L U I D ANALYSIS

Results of synovial fluid analyses are presented in Table II, There was no significant difference in the median volume of synovial fluid retrieved at the time of necropsy from grafted (5.0 ml) vs non- grafted joints (4.0 ml). There were no significant differences in RBC, WBC, protein or hyaluronan values between grafted and nongrafted joints at any sampling time. There was a trend towards higher protein values in grafted vs nongrafted; however, this difference was not significant (P > 0.05). RBC values were less than 3500 at days 120 and 240. Mean WBC values peaked on day 4 and declined to normal by day 14. On days 120 and 240 mean WBC values were within normal values for both experimental and control joints.

D i s c u s s i o n

Deep frozen chondrocytes were successfully thawed, embedded and cultured in semi-synthetic collagen scaffold vehicles, supporting the results of

~ +~

g

~ +~

+~

+l+l+l+l

Y Y Y

II

v

Y Y Y

54 Sams and Nixon: Chondrocyte- laden col lagen scaffolds

previous in vitro studies [27]. Arthroscopic implan- tat ion of these collagen-chondrocyte composites was performed with minimal associated morbidity. Polylactic acid tacks firmly anchored the com- posite in the defect, preventing the loss of grafted chondrocytes which may have occurred in pre- vious studies with other graft vehicles [20]. Clini- cal, gross histologic and synovial fluid evaluations indicated that the chondrocyte implant and tacks were tolerated without significant inflammation.

Immature chondrocytes were used for transplan- tat ion due to their elevated synthetic activity compared to chondrocytes from mature animals [32]. However, a l though immature cells are more synthetically active, they are less hardy and do not survive the harvesting, freezing and culture pro- cess as well as mature cells [19]. Culture studies of equine chondrocytes indicate that the optimal age of chondrocyte donor is approximately 3 ~ months, where the cells remain synthetically active and are more resilient to the harvest and storage procedures.

Culturing in monolayer prior to embedding in the collagen scaffold allowed thawed cells with

marginal viability to be eliminated. Nonviable cells fail to a t tach to the flask base and are removed with the exchange of exhausted media. The viable cells tha t survive the freezing, thawing and initial culture were the hardiest of the cell populat ion and the most synthetical ly active [22]. Pre-implantation high density short-term mono- layer culture, wi thout cell passage, improves Chon- drocyte function and increases the chance of survival in the collagen scaffold and recipient graft bed [22].

Collagen scaffolds have been used for dental procedures, hemostasis, body cavity filling and reconstruct ive surgery [28, 33-35]. Intra-art icular collagen scaffold react ion has not been described in detail [36, 37]. The composite provided a biologi- cally tolerated resorbable framework that allowed the chondrocytes to persist in a 3-dimensional vehicle, facilitating maintenance of the chondro- cyte phenotype [20, 22, 38]. Histologic evaluat ion and glycosaminoglycan (GAG) assay of collagen scaffold disks laden with chondrocytes and as- sessed in vitro over 21 days confirmed chondrocyte survival and synthetic activity of the graft cells

FIG. 5. Synovial membrane photomicrographs from (a) an 8-month grafted and (b) a control joint. The intimal layer (arrow) is one to two cells thick in both experimental and control joints. No cellular infiltrate is seen in the subintimal layers of either sample (H & E; × 312).

Osteoarthrit is and Cartilage Vol. 3 No. 1 55

FIo. 6. Photomicrographs of sections of (a) 8-month grafted and (b) control repair tissue and surrounding cartilage. Clefting can be seen in the control tissue, and a bonding of the repair tissue to the surrounding cartilage in the grafted section. Pericellular glycosaminoglycan basophilic staining can be seen deep within both defects, and a loss of matrix is evident in the surrounding cartilage immediately adjacent to the repair tissue in both the grafted and control sections (H & E; x 8).

[27]. Culturing chondrocytes in the scaffold allowed the ceils to surround themselves with a layer of matrix, masking surface antigens [23, 24]. Coating the scaffold with liquid collagen gave added protect ion from the host immune response, provided some addit ional stiffness, and prevented loss of cells from the chondrocyte-col lagen com- posite.

The lateral t rochlea of the femur was chosen as the implantat ion site in this model because it was sufficiently large to allow formation of an exten- sive defect. Defects of less than 9-mm diameter in horses have been shown to heal bet ter than large defects, solely as a funct ion of lesion size [39]. Matr ix flow likely plays a much greater role in the healing of small defects. To assess the success of this graft in a clinically re levant lesion, large defects were necessary. Most research on ar t icu lar resurfacing has been done using ei ther rabbits or dogs as a model for man. However, defect size and

locat ion are limited in labora tory animals, and do not represent the clinical or biomechanical challenge of large defects.

A graft ing procedure tha t utilized ar throscopic ra the r than a r th ro tomy techniques for placement was an impor tant goal. In man and animals, a r throscopy has made a r th ro tomy largely obsolete, par t icular ly in the complex joints such as the knee [29, 40]. Surg ica l procedures tha t are less invasive lead t6~I0wer morbidity and less intensive pat ient care. Additionally, the ar throscope allows more extensive visualizat ion of the joint compared to ar throtomy. Complete visual izat ion of the t rochlea was possible using s tandard techniques [29]. A modified spade bit made a precise c i rcular defect in the cart i lage and~penetrated the subchondral bone only 1-2 mm. In this model, penet ra t ion of the subchondral bone and the holes drilled for the tacks allowed access to vessels and undifferenti- ated mesenchymal cells beneath the subchondral

58 Sams and Nixon: Chondrocyte-laden collagen scaffolds

and the primary cell types in both defects were fibroblasts. Similarly, histochemical staining was evident over a th icker zone in the deeper regions of the chondrocyte-graf ted repair tissue. The inter- terr i tor ial areas had less stain uptake and this varied directly with the number of chondrocytes per unit area. The chondrocytes in cart i lage sur- rounding the defects, in both the experimental and control joints, had reduced in ter ter r i tor ia l meta- chromasia, suggesting ei ther decreased GAG pro- duction or loss of GAG through matr ix flow into the defect. Chondrocyte cloning was seen in this region but no evidence of an effective increase in GAG product ion was evident histochemically, which is consistent with findings in the l i terature. The lack of apparent influx of white blood cells, histiocytes, or mul t inucleated giant cells in the repair tissue, subchondral bone or synovial membrane suggests tha t there was ei ther no, or minimal, immune response to the chondrocyte graft.

Conc lus ion

Chondrocyte-col lagen composite grafts moder- ately improved the gross and histologic qual i ty of the repair tissue in full thickness defects at 4 and 8 months post-operatively. An immune mediated reject ion react ion was not evident on histologic or synovial fluid evaluation. Despite the fact tha t in vitro studies showed chondrocytes were able te survive in the collagen disk and produce matr ix [27], the t ransplanta t ion procedure resulted in poorer cart i lage in this large defect model than similar chondrocyte grafts in laboratory animals. Although the graft resulted in improved cart i lage healing, the quali ty of the tissue was not normal.

A c k n o w l e d g m e n t s

Technical assistance was provided by Ms Janice Williams, Ms Carol Smith, and Mr Robert Foley.

R e f e r e n c e s

1. Buckwalter J, Hunziker E, Rosenberg L, Coutts R, Adams M, Eyre D. Articular cartilage: compo- sition and structure. In: Woo SL-Y, Buckwalter JA, Eds. Injury and Repair of the Musculoskeletal Soft Tissues. Park Ridge, IL: American Academy of Orthopaedic Surgeons 1988: 405-25.

2. Fuller JA, Ghadially FN. Ultrastructural obser- vations on surgically produced partial-thickness defects in articular cartilage. Clin Orthop 1972;86:193-205

3. Shamis LD, Bramlage LR, Gabel AA, Weisbrode S. Effect of subchondral drilling on repair of partial- thickness cartilage defects of third carpal bones in horses. Am J Vet Res 1989;50:290-5.

4. Meachim G, Roberts C. Repair of the joint surface from subarticular tissue in the rabbit knee. J A n a t 1971;109:317-27.

5. Mankin HJ. The response of articular cartilage to mechanical injury. J Bone Joint Surg [Am] 1982;64:460~.

6. Mitchell N, Shepard N. The resurfacing of adult rabbit articular cartilage by multiple perforations through the subchondral bone. J Bone Joint Surg [Am] 1976;58:230-3.

7. Coletti JM, Akeson WH, Wood SL-Y. A comparison of the physical behavior of normal articular carti- lage and the arthroplasty surface. J Bone Joint Surg [Am]1972;54:147-60.

8. Sullins KE, McIlwraith CW, Powers BE, Norrdin RW. Evaluation of periosteal grafts in articular cartilage repair in horses. Vet Surg 1985;14:66-7.

9. Vachon A, McIlwraith CW, Trotter GW, Norrdin RW, Powers BE. Neochondrogenesis in free intra- articular, periosteal, and perichondrial autografts in horses. Am J Vet Res 1989150:1787-94.

10. Girdler NM. Repair of articular defects with auto- logous mandibular condylar cartilage. J Bone Joint Surg [Br] 1993;75:710-14.

11. Hurtig MB. Experimental use of small osteochondral grafts for resurfacing the equine ~hird carpal bone. Equine Vet J 1988;(suppl 6):23-7.

12. Czitrom AA, Keating S, Gross AE. The viability of articular cartilage in fresh osteochondral allo- grafts after clinical transplantation. J Bone Joint Surg [Am] 1990;72:574-81.

13. Rodrigo JJ, Thompson E, Travis C. Deep-freezing versus 4 degree preservation of avascul~ar osteo- cartilaginous shell allografts in rats. Clin Orthop 1987;218:268-75.

14. Vachon AM, McIlwraith CW, Powers BE et al. Morphologic and biochemical study of sternal cartilage autografts for resurfacing induced osteochondral defects in horses. A m J Vet Res 1992;53:1038-47.

15. McIlwraith CW, Howard GW, Trotter GW, Powers BE, McFadden PR, Harwood FL. Sternal cartilage autografts for repair of large osteochondral defects in horses: Long term fate and effects of exercise. Transactions of the 39th Annual Meeting of the Orthopaedic Research Society 1993;18:729.

16. Itay S, Abramovici A, Nevo 7_, Use of cultured embryonal chick epiphyseal chondrocytes as grafts for defects in chick articular cartilage. Clin Orthop 1987;220:284-303.

17. Robinson D, Halperin N, Nevo Z. Regenerating hyaline cartilage in articular defects of old chick- ens using implants of embryonal chick chondro- cytes embedded i n a new natural delivery substance. Calcif Tiss Int 1990;46:246-53.

18. Wakitani S, Kimura T, Hirooka A et al. Repair of rabbit articular surfaces with allograft chondro- cytes embedded in collagen gel. J Bone Joint Surg [Br] 1989;71:74-80.

19. Nixon AJ, Lust G, Vernier-Singer M. Isolation, propagation and cryopreservation of equine articular chondrocytes. A m J Vet Res 1992; 53:2364-70.

20. Hendrickson DH, Nixon AJ, Grande DA et al. Chon- drocyte-Fibrin matrix transplants for resurfacing extensive articular cartilage defects. J Orthop Res 1994;12:485-97.

O s t e o a r t h r i t i s and Cart i lage Vol . 3 No. 1 59

21. Hendrickson DA. In vitro and in vivo studies of a fibrin polymer as a vehicle for chondrocyte grafting of large articular cartilage defects in horses. MS Thesis, Cornell University, NY, U.S.A., 1993.

22. Hendrickson DA, Nixon AJ, Erb HN, Lust G. Pheno- type and biological activity of neonatal equine chondrocytes cultured in a three-dimensional fibrin matrix. A m J Vet Res 1994;55:410-14.

23. Elves MW. A study of the transplantat ion antigens on chondrocytes from art icular cartilage. J Bone Joint Surg [Br] 1974;56:178-85.

24. Heyner S. The significance of the intercellular matrix in the survival of cartilage allografts. Transplantation 1969;8:666-77.

25. Ben-Yishay A, Grande DA, Menche D, Pitman M. Repair of ar t icular cartilage defects using collagen-chondrocyte allografts Transactions of the 38th Annual Meeting of the Orthopaedic Research Society 1992;17:174.

26. Maor G, von der Mark K, Reddi H, Heinegard D, Franzen A, Silbermann M. Acceleration of cartilage and bone differentiation on collagenous substrata. Collagen Rel Res 1987;7:351-70.

27. Nixon AJ, Sams AE, Lust G, Grande DA, Mohammed H. Temporal matrix synthesis and histologic features of a chondrocyte-laden porous collagen cartilage analogue. A m J Vet Res 1993;54:349-56.

28. Stein MD, Salkin LM, Freedman AL, Glushko V. Collagen sponge as a topical hemostatic agent in mucogingival surgery. J Periodontol 1985;56:35-8.

29. McIlwraith CW. Diagnostic and Surgical Arthroscopy in the Horse, 2nd ed. Philadelphia: Lea & Febiger 1990.

30. Bradford MM. A rapid and sensitive method for the quanti tat ion of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 1976;72:248-54.

31. Smith RL, Gilkerson E, Kohatsu N e t al. Quantitat- ive microanalysis of synovial fluid and art icular cartilage glycosaminoglycans. Anal Biochem 1980;103:191-200.

32. Madsen K, Moskalewski S, vonder Mark K, Friberg U. Synthesis of proteoglycans, collagen, and elastin by cultures of rabbit auricular chondro- cytes--relat ion to age of the donor. Dev Biol 1983;96:63-73.

33. Hansbrough JF, Boyce ST, Cooper ML, Foreman TJ. Burn wound closure with cultured autologous keratinocytes and fibroblasts at tached to a collagen-glycosaminoglycan substrate. J A M A 1989;262:2125-30.

34. Mannai C, Leake D, Pizzoferrato A, Ciapetti G, Sangiorgi C. Histologic evaluation of purified bovine tendon collagen sponge in tooth extract ion sites in dogs. Oral Surg Oral Med Oral Path 1986;61:315-23.

35. Murphy GF, Orgill DP, Yannas IV. Part ial dermal regeneration is induced by biodegradable collagen-glycosaminoglycan grafts. Lab Invest 1990;63:305-13.

36. Speer DP, Chvapil M, Volz RG et al. Enhancement of healing in osteochondral defects by collagen sponge implants. Clin Orthop 1979;144:326-35.

37. Stone KR, Rodkey WG, Webber R, McKinney L, Steadman JR. Meniscal regeneration with copoly- meric collagen scaffolds. In vitro and in vivo studies evaluated clinically, histologically, and biochemically. A m J Sports Med 1992;20:104-11.

38. Glowacki J, Trepman E, Folkman J. Cell shape and phenotypic expression in chondrocytes (41533). Proceedings Soc Exp Bio Med 1983;172:93-8.

39. Convery FR, Akeson WH, Keown GH. The repair of large osteochondral defects. Clin Orthop 1972;82:253-62.

40. McGinty J, Johnson L, Jackson R, McBryde A, Goodfellow J. Uses and abuses of arthroscopy: a symposium. J Bone Joint Surg [Am] 1992;74:1563-77.

41. Kawabe N, Yoshinao M. The repair of full-thickness art icular cartilage defects. Immune responses to reparative tissue formed by allogeneic growth plate chondrocyte implants. Clin Orthop 1991;268:279-93.

42. Moskalewski S, Kawiak J, Rymaszewska T. Local cellular response evoked by cartilage formed after auto-and allogeneic transplantat ion of isolated chondrocytes. Transplantation 1966;4:572-81.

43. Roth JE, Nixon AJ. Pulsed carbon dioxide laser for cartilage vaporization and subchondral bone perforation in horses: 1. Technique, clinical results, and synovial fluid response. Vet Surg 1991 ;20:190-99.