JBJS | Formation of Phalanges and Small Joints by Tissue-Engineering*†

The Journal of Bone & Joint Surgery, Volume 81, Issue 3

Articles | March 01, 1999

Formation of Phalanges and Small Joints by Tissue-Engineering*†

NORITAKA ISOGAI, M.D., PH.D.‡; WILLIAM LANDIS, PH.D.§; TAE HO KIM, M.D.#; LOUIS C. GERSTENFELD, PH.D.**; JOSEPH UPTON, M.D.††; JOSEPH P. VACANTI, M.D.‡‡, BOSTON, MASSACHUSETTS

View Disclosures and Other Information

Investigation performed at the Departments of Surgery and Orthopaedics, Children's Hospital and Harvard Medical School, Boston

The Journal of Bone & Joint Surgery. 1999; 81:306-16

5 Recommendations (Recommend) | 3 Comments | Saved by 3 Users Save Case

Article

Figures

References

text A A A

Abstract

Background: This report describes the formation of small phalanges and whole joints from three types of bovine-cell sources transplanted onto biodegradable polymer matrices. The resulting structures had the shape and composition of human phalanges with joints.Methods: Fresh bovine periosteum was wrapped around a copolymer of polyglycolic and poly-L-lactic acid. Separate sheets of polyglycolic acid polymer were then seeded with chondrocytes and tenocytes isolated from the shoulders of freshly killed calves. The gross form of a composite tissue structure was constituted in vitro by assembling the parts and suturing them to create models of a distal phalanx, a middle phalanx, and a distal interphalangeal joint.Results: Subcutaneous implantation of the sutured composite tissues into athymic mice resulted in the formation, after twenty weeks, of new tissue with the shape and dimensions of human phalanges with joints. Histological examination revealed mature articular cartilage and subchondral bone with a tenocapsule that had a structure similar to that of human phalanges and joints. There was continuous cell differentiation at the ectopic site even after extended periods.Conclusions: These findings suggest that the formation of phalanges and small joints is possible with the selective placement of periosteum, chondrocytes, and tenocytes into a biodegradable synthetic polymer scaffold.Clinical Relevance: The formation of a joint construct of this nature is an example of a growing list of tissue-engineering techniques that, in general, offer alternatives to obtaining autogenous tissue for reconstructive operations in humans. Tissue-engineering holds promise for the treatment of loss of tissue or organ function as well as congenital malformations. Difficult reconstructions in symphalangism, arthrogryposis multiplex congenita, brachydactyly, or traumatically fixed joint contractures may someday be performed with this approach.

Figures in this Article

Introduction

Introduction | Materials and Methods | Results | Discussion | References

The reconstruction of severely traumatized or congenitally deficient joints in young patients who need multiple complicated reconstructive procedures is a challenging problem. Prosthetic joint reconstruction often may be compromised by the limited durability of the nonbiological materials used or by intractable infection¹¹. Another problem with use of artificial prostheses in children is that these devices may need to be replaced as the child grows. In recent years, free whole-joint transfer with use of a microvascular anastomosis has been employed successfully to replace joints or digits in the hand^{2,13,16,21,24,31,35}. Vascularized autogenous transplantation of a joint from the foot to the hand has also been performed. The predictable vascular patterns and sufficient diameter of the first dorsal metatarsal and dorsalis pedis arteries make transplantation of the toe feasible^12,22.

The ideal arthroplasty should provide stability, motion, durability, and freedom from pain. It should also allow for normal growth in the child. Unfortunately, although a vascularized autogenous whole-joint transfer has the potential to meet these demands, it is limited by donor-site morbidity and the availability of tissue. From an aesthetic standpoint, the sacrifice of an entire big toe or the second toe is undesirable. The major problem with regard to the transplantation of allografts and xenografts remains the necessity for lifelong immunosuppression without serious side effects. Another solution is needed.

The engineering of new tissue by means of cell transplantation is another possibility. Small biopsy specimens can be obtained from the patient, cells can be isolated and grown in culture in large numbers, and these cells then can be placed on biodegradable polymer matrices that provide scaffolds to guide tissue ingrowth. The long-term foreign-body response associated with use of engineered autogenous tissue or alloplastic material can be eliminated if the polymer is completely degraded and removed, leaving only naturally regenerated tissue. In this regard, a nonwoven polymer mesh of polyglycolic acid was recently shown to be biocompatible, biodegradable, and effective as a scaffold for cell delivery to generate new tissue^7,19,29,33.

To date, there have been few reports describing the formation of an intact joint with use of cell transplantation on artificial matrices. The current study presents the results of a novel approach in which three different cell types, derived from bovine periosteum, cartilage, and tendon, were used to form a joint with a predetermined shape and composition.

*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. Funds were received in total or partial support of the research or clinical study presented in this article. The funding sources were a grant from Advanced Tissue Sciences, Incorporated, La Jolla, California, and National Institutes of Health Grant AR 41452 (W. L.).

†Read at the Annual Meeting of the Plastic Surgery Research Council, Galveston, Texas, February 1997.

‡Plastic and Reconstructive Surgery, Kinki University Hospital, 377-2 Onohigashi, Osaka-sayama, Osaka 589-8511, Japan.

§Department of Biochemistry and Molecular Pathology, Northeastern Ohio Universities College of Medicine, 4209 State Road 44, P.O. Box 95, Rootstown, Ohio 44272.

#Department of Surgery, University of Massachusetts Memorial Health Care, 55 Lake Avenue North, Worcester, Massachusetts 01655.

**Musculoskeletal Research Laboratory/Orthopedic Surgery, Boston University School of Medicine, 715 Albany Street, R205, Boston, Massachusetts 02118.

††Department of Plastic Surgery, Beth Israel-Deaconess Medical Center, 330 Brookline Avenue, Boston, Massachusetts 02115.

‡‡Department of Surgery, Warren 11, Massachusetts General Hospital, 55 Fruit Street, Boston, Massachusetts 02114.

†Read at the Annual Meeting of the Plastic Surgery Research Council, Galveston, Texas, February 1997.

‡Plastic and Reconstructive Surgery, Kinki University Hospital, 377-2 Onohigashi, Osaka-sayama, Osaka 589-8511, Japan.

§Department of Biochemistry and Molecular Pathology, Northeastern Ohio Universities College of Medicine, 4209 State Road 44, P.O. Box 95, Rootstown, Ohio 44272.

#Department of Surgery, University of Massachusetts Memorial Health Care, 55 Lake Avenue North, Worcester, Massachusetts 01655.

**Musculoskeletal Research Laboratory/Orthopedic Surgery, Boston University School of Medicine, 715 Albany Street, R205, Boston, Massachusetts 02118.

††Department of Plastic Surgery, Beth Israel-Deaconess Medical Center, 330 Brookline Avenue, Boston, Massachusetts 02115.

‡‡Department of Surgery, Warren 11, Massachusetts General Hospital, 55 Fruit Street, Boston, Massachusetts 02114.

View Large | Download Slide(.ppt)
Add to My JBJS

FIG1:: Fig. 1 Diagram showing the basic procedure for the fabrication of polyglycolic-poly-L-lactic acid copolymer. A nonwoven mesh of polyglycolic acid (PGA) fibers was placed in a Petri dish containing poly-L-lactic acid (PLLA) polymer dissolved in methylene chloride. After the solvent had evaporated, the resulting polyglycolic-poly-L-lactic acid composite was subjected to a temperature of at least 195 degrees Celsius for ninety minutes. This process yields a scaffold of bonded polyglycolic acid—polyglycolic acid and polyglycolic-poly-L-lactic acid fibers.

View Large | Download Slide(.ppt)
Add to My JBJS

FIG2:: Fig. 2 Schematic drawings of the three different types of composite tissue structures in the experimental groups. The structures were constituted in vitro by suturing to create models of a distal phalanx (Group I), a middle phalanx (Group II), and a distal interphalangeal joint (Group III). The sutured tissues then were implanted subcutaneously in athymic mice.

View Large | Download Slide(.ppt)
Add to My JBJS

FIG3:: Fig. 3 Photographs of a tissue-engineered distal phalanx (Group I), made after the implant had been in vivo for twenty weeks (x 1.9; bar = 1.2 centimeters). The construct was fabricated by suturing an articular chondrocyte-polymer sheet to the proximal end of a periosteum-polymer model of a distal phalanx with use of Vicryl (Ethicon). A: Macroscopic view of the construct. B: Longitudinal section showing articular cartilage that is clearly distinguishable from the subchondral bone. Cartilage has formed through the full thickness and width of the construct, which has become vascularized (red) both externally (A) and internally (B). A dark central channel along the length of the hemisections in B is the result of abutment of two halves of the polymer during formation of the phalanx. The channel is filled with vascular elements. (Regions a, b, and c in B are enlarged in Fig. 4.)

View Large | Download Slide(.ppt)
Add to My JBJS

FIG4:: Fig. 4 Light photomicrographs of specimens from the tissue-engineered distal phalanx. a: Enlarged view of region a in Fig. 3, B, showing smooth articular cartilage stained with safranin O. There is no disintegration at the well defined, intact osteochondral junction (J) between the articular cartilage (C) and the subchondral area (S) of the phalanx. Small, irregularly shaped inclusions in the subchondral bone are residual polymer. A layer of fine connective tissue (blue) of host origin surrounds the construct (x 30). b: Enlarged view of a, corresponding in part to region b in Fig. 3, B, revealing three zones of articular cartilage: tangential (TA), transitional (TR), and radial (RA). The zones may be distinguished by staining. Note the presence of chondrocytes in the cartilage (x 80). c: The diaphyseal bone corresponding to region c in Fig. 3, B. Trabeculae (T) of new bone, possibly formed by periosteum-derived osteoblasts, are apparent near and within the surface of the polyglycolic-poly-L-lactic acid copolymer construct of the shaft. The outermost surface of the construct is surrounded by host connective tissue. The polymer is being resorbed, but there are numerous remnants (P) (hematoxylin and eosin, x 30). d: Higher-magnification view of a trabecular area similar to that shown in c. There are many osteoblasts, some organized in a palisade (arrowheads) adjacent to osteoid and new bone. Presumptive vascular channels (V), residual polymer (P), and trabeculae (T) are also seen (hematoxylin and eosin, x 160).

View Large | Download Slide(.ppt)
Add to My JBJS

FIG5:: Fig. 5 Photographs of a tissue-engineered middle phalanx (Group II) after twenty weeks in vivo (x 1.4; bar = one centimeter). Two articular chondrocyte-polymer constructs were sutured onto the periosteum-polymer construct in the shape of a middle phalanx. Additional polyglycolic acid polymer seeded with tenocytes then was sutured to the proximal end of the composite. A: Macroscopic view showing two articular cartilage surfaces with round borders and vascular formation over the central shaft. B: Longitudinal section showing vascular ingrowth and extensive development of bone tissue in the middle portion of the phalanx (a), cartilage at the articular portions of both the distal end and the proximal end (b), a tenochondral junction (c), and formation of neotendon at the proximal point of suture (d).

View Large | Download Slide(.ppt)
Add to My JBJS

FIG6:: Fig. 6 Photomicrographs of specimens from the tissue-engineered middle phalanx, made twenty weeks after implantation. a: The center of the shaft, corresponding to region a in Fig. 5, B, is occupied by numerous spatially independent sites comprised of chondrocytes undergoing hypertrophy (HC). Osteoblasts and trabecular bone (T) are also present as are some polymer remnants (P). Bone formation is extensive. No inflammatory response was detected (hematoxylin and eosin, x 80). b: Articular cartilage corresponding to region b in Fig. 5, B, is stained with safranin O. The tangential (TA) and transitional (TR) zones are identifiable; each contains chondrocytes that are widely distributed but are smaller and flatter in the tangential regions. There is some indication that the cells in the tangential zone lie parallel to the articular surface. A thin layer of connective tissue (blue) is apparent around the construct (x 80). c: Photomicrograph corresponding to region c in Fig. 5, B, showing newly formed articular cartilage (C), fibrocartilage, and tendon (TE). An insertion of the fibers of the collagenous tendon (tenocytes seeded on a polyglycolic acid sheet) into the articular cartilage (chondrocytes seeded on a polyglycolic acid sheet) is visible (arrow). Fibrocartilage, also stained with safranin O, has formed at the tenochondral junction, indicating a normal biological response and development by the two different cell types in this portion of the construct (x 30). d: Enlarged view of c, showing the extension of the tenocyte-seeded polyglycolic acid polymer and the cells in linear alignment along the neotendon (TE) (hematoxylin and eosin, x 100). Inset: higher-magnification view showing the parallel nature of the tendon fibers and the presence of numerous darkly staining tenocytes, again disposed in linear fashion as seen in vivo. The clear area at the top of the image is the edge of the surface of the developing neotendon (safranin O, x 500).

View Large | Download Slide(.ppt)
Add to My JBJS

FIG7:: Fig. 7 Photographs showing a tissue-engineered distal interphalangeal joint (Group III) at twenty weeks. Articular chondrocyte-polymer sheets were sutured separately to the periosteum-polymer constructs of the distal and middle phalanges to create articular surfaces. The joint then was fabricated by wrapping these two composites with additional polyglycolic acid polymer sheets seeded with tenocytes. A: Macroscopic view showing considerable vascularity over the surface (x 1.5; bar = one centimeter). B: Longitudinal section of a portion of the construct, demonstrating that the tenocapsule of the joint was well formed from the tenocyte-polymer construct. The joint space is green (x 1.65; bar = one centimeter). (Regions a through d are enlarged in Fig. 8.)

View Large | Download Slide(.ppt)
Add to My JBJS

FIG8:: Fig. 8 Light photomicrographs of specimens from the tissue-engineered joint. a: The tenocapsule region (arrowheads) corresponding to regions a and b in Fig. 7, B, is intact between the distal (D) and middle (M) phalanges of the composite. The adjacent cartilaginous articular surfaces are visible on staining with hematoxylin and eosin (x 12). b: Enlarged view of a, showing the insertion of tendon into the bone model as in Fig. 6, c; a joint cavity (JC); and a portion of the joint capsule (arrow). Articular cartilage (C) adjacent to the joint capsule is apparent on staining with Masson's trichrome (x 80). c: Staining with safranin O shows that the tissue-engineered articular surface corresponding to region c in Fig. 7, B, and also seen in a may be separated into superficial tangential (TA) and transitional zones (TR) along the joint cavity (JC). A narrow layer of fibrous tissue (blue) lines the articular cartilage (x 80). d: The interior spaces of the diaphyseal shaft, such as those corresponding to region d in Fig. 7, B, are seen to contain numerous hypertrophic chondrocytes (HC) as well as remnants of residual polymer (P). The formation of hypertrophic cells was similar to that in the model of a middle phalanx alone, shown in Fig. 6, a (hematoxylin and eosin, x 330).

View Large | Download Slide(.ppt)
Add to My JBJS

FIG9:: Fig. 9 Histological changes in a Group-I implant after forty weeks in vivo (x 2.5; bar = one centimeter). A: Photograph of a longitudinal hemisection, showing a mature periosteum, cortical bone (the triangular-shaped tissue that is partially revealed beneath the uncut vascular surface of the construct), and abundant cancellous bone within the construct. Only a portion of the construct is visible because the heavily mineralized composite was difficult to section. B: Photomicrograph of the implant stained with safranin O, showing a distinct region of articular cartilage (b, red), the envelope of cortical bone (near c), and extensive cancellous bone (a) interior to the construct. The relative locations and contours of these three features are similar to those seen in a long bone in vivo. No equivalent marrow cavity has formed.

View Large | Download Slide(.ppt)
Add to My JBJS

FIG10:: Fig. 10 Higher-magnification photomicrographs of portions of the construct shown in Fig. 9. a: There is evidence of the development of a growth plate and the presence of articular cartilage (C) on the extreme end of the model, subchondral cancellous bone (T), and an osteochondral junction (J) revealed by staining with safranin O. The area corresponds to that between sites a and b in Fig. 9, B (x 30). b: Enlarged view of a, showing articular cartilage with hypertrophic chondrocytes partly organized in columns in the radial zone (RA); rounded, individual chondrocytes in the transitional zone (TR); and relatively small, flat reserve cells in the tangential zone (TA) of the tissue (safranin O, x 80). c: Enlarged view of region c in Fig. 9, B, demonstrating extensive cortical bone (CO) overlaid with a thick periosteum and adjacent to regions of cancellous-bone (CA) formation. The periosteum and the cancellous bone are highly vascularized (hematoxylin and eosin, x 30). d: Enlarged view of an area of cortical bone similar to that in c, showing numerous osteocytes in their lacunae, lamellar bone (L), and a presumptive haversian system (H) established in the construct. There are almost no polymer remnants (hematoxylin and eosin, x 160).

Materials and Methods

Introduction | Materials and Methods | Results | Discussion | References

Polymer Molds of the Phalanges

Two forms of biodegradable polymers were used for fabrication of the phalanges and phalangeal joints. In both, a nonwoven mesh of polyglycolic acid fibers (Albany International, Mansfield, Massachusetts) with a diameter of fifteen micrometers and interfiber spaces averaging seventy-five to 100 micrometers was used. The mesh served as a supportive structure for cell attachment and matrix formation. The diameter of the fiber provides mechanical strength while the interfiber spaces provide areas for matrix formation. One type of polymer, used for cartilage and tendon support, was simply a mesh of polyglycolic acid fibers alone, prepared as described previously^6,7,19,32,33; this mesh provided a flat surface with no specific shape. The second type of polymer was created by cutting a polyglycolic acid mesh into rectangles with dimensions appropriate for the formation of phalanges. The polyglycolic acid mesh then was immersed in a 2 percent solution of poly-L-lactic acid (Polysciences, Warrington, Pennsylvania), resulting in the formation of covalent cross-links between the poly-L-lactic acid chains and the polyglycolic acid fibers as well as between the polyglycolic acid fibers themselves. Cross-linking reactions cause the constructs to retain specific shapes²³ (Fig. 1). The wet mesh then was shaped into the form of a human phalanx, modeled from cadaveric human distal and middle phalangeal bones. Negative impression molds of the shaped mesh were composed of vinyl polysiloxane (Impression material putty; 3M Dental Products, St. Paul, Minnesota). After evaporation of the solvent, the copolymer was removed from the mold. The polymers were sterilized with ethylene oxide.

Source of the Cells

Preparation of the periosteum¹⁸: Shoulders and forelimbs from newborn calves were obtained from a slaughterhouse within six hours after the animals had been killed. Fresh bovine periosteum from the radial diaphyses was removed in sterile fashion with a periosteal elevator and was wrapped around the central portion of the polyglycolic-poly-L-lactic acid copolymer molds of the finger.

Isolation of chondrocytes^7,19,32,33: Chondrocytes were isolated from the articular cartilage of shoulder and elbow joints obtained from calves, as described by Klagsbrun²⁰. The cartilage was enzymatically digested with use of 0.3 percent collagenase (Worthington, Freehold, New Jersey) on a shaker for twelve hours, and the released cells were passed through a sterile 150-micrometer nylon-mesh filter. This filtered solution was mixed with an equal volume of Ham's F12 medium (Gibco, Grand Island, New York) supplemented with 10 percent fetal bovine serum (Gibco) and was washed twice. The isolated cells were counted with a hemocytometer, and their viability was determined with use of the trypan-blue exclusion method. The cells were suspended in Ham's F12 medium and then were concentrated by centrifugation at 900 repetitions per minute for five minutes at room temperature. The resulting cell suspension was diluted to a final concentration of 150 x 10⁶ cells per milliliter. Polyglycolic acid polymer mesh, one by one centimeter in size and two millimeters in thickness, was seeded with 100 microliters of the concentrated cell suspension.

Isolation of tenocytes: Tenocytes were obtained and isolated as described previously⁶. Flexor tendons of the forelimbs of calves were obtained when the animals were killed and were diced into pieces that were approximately five cubic millimeters in size. The tissue was digested and the cells were isolated as described with regard to the chondrocytes. The cells were suspended in Ham's F12 medium, concentrated in a suspension, and diluted to 150 x 10⁶ cells per milliliter as noted earlier. Two hundred microliters of the suspension (30 x 10⁶ cells) was then seeded onto polymer-mesh sheets that were one by two centimeters in size and two millimeters in thickness.

Preparation and Implantation of the Tissue Constructs

The periosteum-polymer constructs were placed in M199 medium (Gibco), and the cell-polymer constructs were placed in Ham's F12 medium (Gibco). Both media contained 10 percent fetal bovine serum, L-glutamine (292 micrograms per milliliter), penicillin (100 units per milliliter), streptomycin (100 micrograms per milliliter), and ascorbic acid (fifty micrograms per milliliter). All constructs were incubated at 37 degrees Celsius in a humidified 5 percent CO₂ incubator for one week before implantation.

Three experimental models of phalanges and phalangeal joints were designed (Fig. 2). In Group I (six implants), a distal phalanx was constructed by suturing a chondrocyte-polymer sheet that was fit and trimmed over the proximal end of the periosteum-polymer composite. The suture material was unbraided biodegradable polyglactin (5-0 Vicryl; Ethicon, Somerville, New Jersey). In Group II (five implants), a chondrocyte-polymer sheet was sutured over both ends of a periosteum-polymer construct in the shape of a middle phalanx. Additional polyglycolic acid polymer seeded with tenocytes then was sutured to the proximal end of the composite to create a middle phalanx with a tendon. In Group III (five implants), chondrocyte-polymer sheets were sutured to periosteum-polymer constructs of the distal and middle phalanges to create apposing articular surfaces, and a joint then was formed by wrapping this composite with additional polyglycolic acid polymer sheets seeded with tenocytes. This created a distal interphalangeal joint. A silicone sheet with a thickness of 0.5 millimeter was inserted between the phalanges to prevent contact between the two adjacent equivalent articular surfaces.

All constructs were implanted into the dorsal subcutaneous space of four-to-six-week-old athymic nude mice. A total of sixteen composite constructs (one per animal) was implanted. Three implants from Group I were removed after twenty weeks and three, after forty weeks, in order to examine time-dependent changes in the morphology of the implants and the phenotypic expression of the cells. Samples were dissected in half and were examined histologically. The implants from Groups II and III also were dissected in half and were examined histologically after twenty weeks in vivo.

All animals were treated in compliance with the guidelines for the care and use of laboratory animals, developed by the National Institutes of Health⁸.

Histological Examination

The implants were fixed in 10 percent neutral buffered formalin and embedded in paraffin. Paraffin sections with a thickness of six micrometers were stained with hematoxylin and eosin to define the general morphology and with safranin O to reveal proteoglycans in the constructs. Since the polyglycolic-poly-L-lactic acid copolymer is highly birefringent, its retention and biodegradation over time were assessed with polarizing microscopy (Ortholux II light microscope; Leitz Wetzlar, Wetzlar, Germany). The percentage of residual polymer was determined, with use of computer-assisted image-analysis software (NIH Image 1.55 graphics program; National Institutes of Health, Bethesda, Maryland), on the basis of positive birefringent areas of photographs made under low-power magnification (x 40). Measurements were made for five fields per slide for the distal phalanx constructs that were removed twenty and forty weeks after implantation. Three constructs were examined at each time-period.

Morphometric and Statistical Analysis

The dimensions of the Group-I implants were analyzed by measuring the width and the length of each. Statistical analysis of the size of the implants was performed with use of a Power Macintosh computer (Apple Computer, Cupertino, California), and the values were expressed as the mean and the standard error of the mean. The unpaired t test was used for the analysis (performed with StatView 4.0; Abacus Concepts, Berkeley, California), and p values of less than 0.05 were considered significant.

Results

Introduction | Materials and Methods | Results | Discussion | References

Histological Characteristics of the Cells and the Polymer

Tissue cultures of cells on polymer scaffolds for one week resulted in the proliferation of periosteal osteoblasts, chondrocytes, and tenocytes in the polyglycolic acid models and sheets. There was no development of bone, cartilage, or tendon before the constructs were implanted into the nude mice. These morphological findings are identical to those that were reported previously by us and other authors^6,7,19,32,33.

Tissue-Engineered Constructs

In the initial experiments, the basic Group-I construct consisting of the chondrocyte-containing polymer sheet sutured to the periosteum-wrapped core polymer was examined. On removal from the athymic mouse after twenty weeks, this construct had a sharp, well defined configuration of a distal phalanx (Fig. 3, A). On gross examination, a clearly delineated cartilaginous tissue was observed over the surface onto which the chondrocyte-seeded polymer had been sutured (Fig. 3, A). When sectioned longitudinally, these specimens, which had been created by implantation of periosteum onto polymer mesh, showed a greater degree of mineralization peripherally with polymer remnants still present centrally. The margins of the articular cartilage were readily distinguishable from subchondral bone (Figs. 3, B, and 4, a and b). Cartilage development was limited to the articular surface that had been seeded with chondrocytes (Fig. 3, B).

On histological examination, the articular surface of the Group-I constructs was smooth and formation of subchondral bone was evident (Fig. 4, a). A distinct interface was seen at the junction of the bone and the cartilage on staining with safranin O, and there was no discontinuity between the articular cartilage and the regions of subchondral bone. Such interfacial zones had a histological appearance consistent with that of metaphyseal bone surfaces in vivo. Chondrocytes were located throughout the depth of the articular cartilage and had no distinct organization; however, larger and more numerous cells were present in the deeper regions of the cartilage (Fig. 4, b). At twenty weeks, new bone trabeculae had formed along the periphery of the polyglycolic-poly-L-lactic acid copolymer (Fig. 4, c) and many osteoblasts, vascular channels, and hematopoietic elements were visible within the construct (Fig. 4, d).

The Group-II implants resulted in the generation of a middle phalanx-like bone twenty weeks after implantation (Figs. 5 and 6). As described, a chondrocyte-containing polymer sheet had been sutured to both ends of a periosteum-core polymer composite and a tenocyte-seeded polymer had been sutured to the proximal end of the construct. The two articular cartilage surfaces had well circumscribed or rounded borders (Fig. 5, A). These constructs eventually received a blood supply from the surrounding tissue, which was rich in vascular elements; the blood supply was derived especially from the bottom margins of underlying muscle (Fig. 5, A). Longitudinal sections showed that articular cartilage had formed consistently throughout the entire thickness and width of the chondrocyte-polymer sheet (Fig. 5, B) and that bone had formed throughout the entire width and length of the middle phalanx. A tenochondral junction was clearly visible at the proximal border of the phalanx (Fig. 5, B).

At higher magnification, the polyglycolic-poly-L-lactic acid copolymer that made up the shaft of the phalanx demonstrated numerous sites containing hypertrophic chondrocytes, osteoblasts, and newly formed trabecular bone (Fig. 6, a). At the proximal end of the phalanx, newly formed fibrocartilage was observed in direct apposition to the neighboring articular cartilage and the fibrocartilage was continuous with newly formed tendon (Fig. 6, c). The tendon tissue was composed of spindle-shaped fibroblasts organized in linear arrays as in normal tendon. The collagen bundles of the tendon tissue also appeared to be organized in a parallel manner and oriented in the same direction as the longest dimension of the constituent fibroblasts (Fig. 6, d and inset).

The composite whole-joint constructs in Group III retained their original shape and formed a phalangeal joint resembling their normal counterparts (Fig. 7, A). A longitudinal hemisection of one such specimen showed, after twenty weeks, that the periosteum-polymer composite had developed into bone tissue, whereas the articular chondrocyte-polymer construct had developed into cartilage. In these constructs, a tenocyte-containing polymer had also been added, and this formed a tenocapsule of the joint that was distinguishable from the surrounding connective tissue. The arrangement of the engineered tissue was quite similar to that of a normal joint (Fig. 7, B). Histologically, a tenocapsule was seen to have formed between the two phalanges and the adjacent cartilaginous articular surfaces remained intact (Fig. 8, a and b). The chondrocytes had formed an articular surface along the joint space (Fig. 8, c), and this region stained intensively with safranin O as did the related constructs of the distal phalanx (Fig. 4, a and b) and the middle phalanx (Fig. 6, b and c). The joint surface appeared smooth and congruent, although a thin fibrous layer was detected in the innermost articular surface at higher magnification. This layer presumably resulted from a foreign-body reaction to the silicone sheet that had been inserted into the joint space. A distinct transitional zone at the interface between the subchondral bone and the articular cartilage was clearly observed and was found to be smooth (Fig. 8, a). These findings indicate that biological development and cell response in these two different regions occurred in a manner similar to that in normal tissue. There was no histological evidence of inflammation around this or the other types of implants.

After forty weeks in vivo, the distal phalanx model had formation of a distinct articular cartilage with retention of its original contour and a periosteum (Fig. 9) that was thicker than that of the constructs at twenty weeks (Figs. 3 and 4). Longitudinal sections revealed well organized bone tissue throughout the constructs (Fig. 9), whereas calcified bone matrix had been limited principally to the peripheral regions of these implants at twenty weeks (Fig. 4). At forty weeks, an intact junction had formed between the articular cartilage and the subchondral bone; this subchondral bone contained bone spicules and cartilage remnants that stained positively with safranin O (Fig. 10, a). Notably, a growth plate consisting of chondrocytes, having a phenotype and arrangement similar to those seen in vivo, and a matrix rich in proteoglycans appeared to have developed in the articular regions (Fig. 10, b). There was cortical bone around the constructs and abundant cancellous bone within them (Figs. 9, B, and 10, a, b, and c). The tissue constructs were heavily vascularized (Fig. 10, c), and haversian canals were noted in areas of lamellar bone formation (Fig. 10, d).

Morphometric data on changes in the dimensions of the distal phalanx constructs between twenty and forty weeks were determined. Three constructs at twenty weeks and three constructs at forty weeks were measured. At the time that the specimens were implanted, all six constructs had a width of five millimeters and a length of sixteen millimeters. At twenty weeks, these implants had a mean width (and standard error) of 5.3 ± 1.2 millimeters compared with 4.6 ± 0.5 millimeters at forty weeks and a mean length of 18.0 ± 1.2 millimeters compared with 22.0 ± 3.0 millimeters at forty weeks. Thus, there was little change in either dimension during the twenty-week interval, although there was a trend toward lengthening of the construct. With the numbers available, neither of these changes was found to be significant, with use of an unpaired t test (p = 0.655 and p = 0.233 for width and length, respectively). No attempt was made to determine the volume of the specimens, but this parameter increased with time as assessed visually and as suggested by changes in relative area based on length multiplied by width alone.

Fate of the Polyglycolic-poly-L-lactic Acid Copolymer

Since the physical and chemical properties of implanted materials appear to influence osteoclastic adherence and to modify osteoclastic activity²⁸, the retention and biodegradation of the polyglycolic-poly-L-lactic acid material in the six distal phalanx constructs were determined on the basis of residual polymer birefringence. At twenty weeks, polarizing microscopy of three implants revealed a marked reduction in birefringence attributable to polymer biodegradation. The percentage of residual polymer, determined by dividing the area of residual polymer by the total area of the periosteum-polymer construct, was a mean (and standard error) of 36.5 ± 7.1 at twenty weeks. At forty weeks, the continued hydrolysis of the polymer had extensively degraded it, in three other implants, to about 5 percent of its original content, leaving only small pieces within the constructs; at this time-period, the mean percentage of residual polymer was 5.3 ± 0.9. The difference between the percentages of residual polymer at the two time-periods was found to be significant with use of an unpaired t test (p = 0.0059).

Discussion

Introduction | Materials and Methods | Results | Discussion | References

Whole-joint construction presents special challenges in tissue-engineering since the joint is a complex structure composed of multiple cell types and their particular extracellular matrices. The joint also has the capacity to perform a specific function in an organism. The individual cells are further defined by their differentiated states at any particular time-point. The anatomical and physiological variables are confounding in their potential to develop proper arrangements of the multiple cell types and matrices that make up the joint. In any joint, uniquely configured apposing articular surfaces are sheathed by synovial cells, which in turn are encased in an osteoarticular skeleton held intact by complex ligamentous and capsular structures³³. Moreover, the articular surfaces are composed of hyaline cartilage and an underlying subchondral bone. All of these cell types, the tissues that they form, and the synchronous, integrated function that they execute at any time should be considered in joint-engineering.

In the current study, three sources of distinct structural cells on biodegradable polymers were introduced in vivo in order to generate an engineered intact whole joint. The use of several specific differentiated cell types produced a final construct that had an articular surface of hyaline cartilage and an underlying subchondral bone resembling normal phalangeal tissue. First, periosteum was used to generate bone as this tissue has osteogenic potency through release of periosteum-derived osteoblastic cells. Grafts of periosteum are known to induce new-bone formation²⁶; vascularized periosteum has a constant level of osteogenic capacity and the amount of new bone increases over time^{5,10,17,30,34}. When the periosteum was placed directly onto a polyglycolic acid polymer, periosteal cells migrated from the tissue, attached and spread readily on the polymer, and formed bone through an endochondral ossification process. However, periosteal cells did not attach as readily to the polymer fiber when they were seeded as a concentrated cell suspension¹⁸. Secondarily, chondrocytes and tenocytes were seeded directly onto the polymer scaffold. Transplantation of chondrocytes and tenocytes on polyglycolic acid polymer mesh has been shown to regenerate cartilage^19,32 and tendon tissue⁶ on amorphous fibrous mesh, and the adaptation here to polyglycolic and poly-L-lactic acid polymer was successful.

For clinical applications, it is essential that polymer have a defined shape that remains stable after transplantation and until new tissue forms appropriately. A synthetic polymer system of polyglycolic-poly-L-lactic acid for use in cell transplantation promotes cell adhesion and provides a structural template for tissue organization^7,19,32. This material degrades through nonspecific hydrolysis as tissue organization proceeds, leading to a natural tissue without foreign materials and with little or no observable inflammatory response. Analysis of birefringence in this study showed that the original polymer scaffold had been nearly completely hydrolyzed in this manner. The mechanism of polymer resorption is not known, but it probably involves nonspecific degradation. Osteoclasts do not appear, although osteoblasts are present on the polymer, an observation consistent with that in other reports^6,7,19,32,33. The factors that are responsible for the migration of osteoblasts to the polymer are unclear. Perhaps the direct nature of the polymer surface or the surface adsorption of certain host proteins serves as a chemoattractant to the cells.

Previous studies of nonvascularized autogenous joint transplantation have demonstrated the survival of articular cartilage^3,4,9,25. However, failure because of delays in revascularization has led to subchondral collapse with subsequent disintegration of the cartilage^13,15,27. A vascularized environment is necessary for the construction of a large volume of tissue. The findings of the current study suggest that an angiogenic response of the host to the implanted construct allows the bovine cells to survive and proliferate after extended periods. The rate of bone morphogenesis was related to the site of implantation of the periosteum-polymer construct and also to the vascularity at this location. The viability of the chondrocytes is interesting to consider. While these cells and articular cartilage do not normally require a vascular supply for survival, previous attempts to repair articular cartilage defects with use of isolated chondrocytes have failed¹. In the present study, in contrast, the full-thickness structure of the articular cartilage was reproduced by the implantation of both periosteum and articular chondrocyte-seeded polymers. This suggests the possibility that the multiple cell types of the constructs play an important role in the maintenance of the articular cartilage. The specific biological mechanisms that control the regeneration of articular cartilage are not known. Perhaps the cells or the subchondral bone releases vascular factors or local cytokines that are essential to the development of chondrocytes and cartilage in the constructs.

Throughout the course of this study, the two engineered tissue layers of the bone-cartilage composites remained distinct, and the interface was without discontinuity. Similarly, the junction between the cartilage-tendon composites showed formation of fibrocartilage. These findings indicate that the polymers support the cells that are responsible for reorientation of the total construct into structures that retain the tissue architecture seen in vivo.

In summary, this study provides evidence that whole-joint structures can be created de novo from dissociated cells in vivo with use of biodegradable polymers as delivery vehicles. Unwoven polyglycolic acid, used alone and as a copolymer with poly-L-lactic acid, provided a satisfactory shape and composition for the scaffold and was suitable for cell survival and growth.

As suggested earlier, a vascular supply appears to be critical to cell viability and to the development of the tissues that make up the phalangeal joint construct. However, no attempt was made to identify the precise host source of vascularity or the fate of the cell types in the current study. Regardless of the vascular and cellular origins, the lamellar and cancellous bone, cartilage, and tendon that formed in the constructs appeared similar to those seen in vivo. The mechanisms of such tissue formation, which leads to endochondral ossification within the central portion of the phalanx or to formation of cartilage rich in proteoglycans, also may be closely related to those that occur in vivo.

With regard to tissue development, it is interesting that a putative growth plate appeared in the cartilaginous regions of the bone constructs. This engineered tissue contained immature, proliferating, and hypertrophic chondrocytes of increasing size, apparently organized as they are in vivo, with the respective cells arranged in a columnar fashion and maintained in specific zones of cartilage without migrating beyond them. Normally, such growth regions of cartilage are thought to form in response to the mechanical forces of compressive loading, as generated by animal locomotion for example. In the current model, the constructs were implanted into the dorsal subcutaneous space of mice, where mechanical forces appear to be absent. Nonetheless, a potential growth plate developed, although the mechanism for its formation is unknown. It is possible that movement by the mouse introduces strain, shear stresses, or fluid flow, or a combination of these factors, which in turn produces streaming potentials as sources of mechanical forces that act on the constructs. However, this explanation remains entirely conjectural. Perhaps mechanical forces simply are not required.

While an engineered whole-joint structure was successfully generated in this study, the construct still requires optimization with regard to shape and function. Since the study was performed with use of immunologically incompetent mice, special attention must be paid to the possible deterioration of the implant caused by immunological rejection in a fully competent model. To limit this potential problem of donor rejection in a clinical setting, patients can serve as hosts for the regeneration of their own tissue-engineered components. Non-weight-bearing joints of upper limbs are the joints for which the possible uses of the technology are most applicable. Studies of the response of tissue-engineered joints to dynamic forces and to innervation of these constructs are ongoing. With regard to innervation, other research in this laboratory is focusing on tissue constructs engineered to contain small channels that can serve as guides for nerve location and direction¹⁴.

References

Introduction | Materials and Methods | Results | Discussion | References

Bentley, G., and Greer, R. B., III: Homotransplantation of isolated epiphyseal and articular cartilage chondrocytes into joint surfaces of rabbits. Nature,230: 385-388, 1971.230385 1971 [PubMed]

Buncke, H. J., Jr.; Daniller, A. I.; Schulz, W. P.; and Chase, R. A.: The fate of autogenous whole joints transplanted by microvascular anastomoses. Plast. and Reconstr. Surg.,39: 333-341, 1967.39333 1967

Campbell, C. J.; Ishida, H.; Takahashi, H.; and Kelly, F.: The transplantation of articular cartilage. An experimental study in dogs. J. Bone and Joint Surg.,45-A: 1579-1592, Dec. 1963.45-A1579 1963

Campbell, C. J.: Homotransplantation of a half or whole joint. Clin. Orthop.,87: 146-155, 1972.87146 1972 [PubMed]

Canalis, R. F., and Burstein, F. D.: Osteogenesis in vascularized periosteum. Interactions with underlying bone. Arch. Otolaryngol.,111: 511-516, 1985.111511 1985 [PubMed]

Cao, Y.; Vacanti, J. P.; Ma, X.; Paige, K. T.; Upton, J.; Chowanski, Z.; Schloo, B.; Langer, R.; and Vacanti, C. A.: Generation of neo-tendon using synthetic polymers seeded with tenocytes. Transplant. Proc.,26: 3390-3392, 1994.263390 1994 [PubMed]

Cima, L. G.; Vacanti, J. P.; Vacanti, C.; Ingber, D.; Mooney, D.; and Langer, R.: Tissue engineering by cell transplantation using degradable polymer substrates. J. Biomech. Eng.,113: 143-151, 1991.113143 1991 [PubMed]

Committee on Care and Use of Laboratory Animals of the Institute of Laboratory Animal Resources, National Research Council: Guide for the Care and Use of Laboratory Animals. Publication 85-23, revised 1985. Washington, D.C., United States Department of Health, Education, and Welfare, National Institutes of Health, 1985.

DePalma, A. F.; Tsaltas, T. T.; and Mauler, G. G.: Viability of osteochondral grafts as determined by uptake of S³⁵. J. Bone and Joint Surg.,45-A: 1565-1578, Dec. 1963.45-A1565 1963

Finley, J. M.; Acland, R. D.; and Wood, M. B.: Revascularized periosteal grafts—a new method to produce functional new bone without bone grafting. Plast. and Reconstr. Surg.,61: 1-6, 1978.611 1978

Flatt, A. E.: Studies in finger joint replacement. A review of the present position. Arch. Surg.,107: 437-443, 1973.107437 1973 [PubMed]

Foucher, G.; Merle, M.; Maneaud, M.; and Michon, J.: Microsurgical free partial toe transfer in hand reconstruction: a report of 12 cases. Plast. and Reconstr. Surg.,65: 616-627, 1980.65616 1980

Goldberg, V. M.; Porter, B. B.; and Lance, E. M.: Transplantation of the canine knee joint on a vascular pedicle. A preliminary study. J. Bone and Joint Surg.,62-A: 414-424, April 1980.62-A414 1980

Hadlock, T.; Elisseeff, J.; Langer, R.; Vacanti, J.; and Cheney, M.: A tissue-engineered conduit for peripheral nerve repair. Arch. Otolaryngol. Head and Neck Surg.,124: 1081-1086, 1998.1241081 1998

Herndon, C. H., and Chase, S. W.: Experimental studies in the transplantation of whole joints. J. Bone and Joint Surg.,34-A: 564-578, July 1952.34-A564 1952

Hurwitz, P. J.: Experimental transplantation of small joints by microvascular anastomoses. Plast. and Reconstr. Surg.,64: 221-231, 1979.64221 1979

Ishida, H.; Tamai, S.; Yajima, H.; Inoue, K.; Ohgushi, H.; and Dohi, Y.: Histologic and biochemical analysis of osteogenic capacity of vascularized periosteum. Plast. and Reconstr. Surg.,97: 512-518, 1996.97512 1996

Kim, T. H.; Jannetta, C.; Vacanti, J. P.; Upton, J.; and Vacanti, C. A.: Engineered bone from polyglycolic acid polymer scaffold and periosteum. Mater. Res. Soc. Sympos. Proc.,394: 91-97, 1995.39491 1995

Kim, W. S.; Vacanti, J. P.; Cima, L.; Mooney, D.; Upton, J.; Puelacher, W. C.; and Vacanti, C. A.: Cartilage engineered in predetermined shapes employing cell transplantation on synthetic biodegradable polymers. Plast. and Reconstr. Surg.,94: 233-237, 1994.94233 1994

Klagsbrun, M.: Large-scale preparation of chondrocytes. Meth. Enzymol.,58: 560-564, 1979.58560 1979 [PubMed]

Mathes, S. J.; Buchannan, R.; and Weeks, P. M.: Microvascular joint transplantation with epiphyseal growth. J. Hand Surg.,5: 586-589, 1980.5586 1980

May, J. W., Jr.; Chait, L. A.; Cohen, B. E.; and O'Brien, B. M.: Free neurovascular flap from the first web of the foot in hand reconstruction. J. Hand Surg.,2: 387-393, 1977.2387 1977

Mikos, A. G.; Bao, Y.; Cima, L. G.; Ingber, D. E.; Vacanti, J. P.; and Langer, R.: Preparation of poly(glycolic acid) bonded fiber structures for cell attachment and transplantation. J. Biomed. Mater. Res.,27: 183-189, 1993.27183 1993 [PubMed]

O'Brien, B. M. [editor]: Microvascular free small joint transfer. In Microvascular Reconstructive Surgery, pp. 284-289. New York, Churchill Livingstone, 1977.

Pap, K., and Krompecher, S.: Arthroplasty of the knee. Experimental and clinical experiences. J. Bone and Joint Surg.,43-A: 523-537, June 1961.43-A523 1961

Ritsila, V.; Alhopuro, S.; and Rintala, A.: Bone formation with free periosteum. An experimental study. Scandinavian J. Plast. and Reconstr. Surg.,6: 51-56, 1972.651 1972

Seligman, G. M.; George, E.; Yablon, I.; Nutik, G.; and Cruess, R. L.: Transplantation of whole knee joints in the dog. Clin. Orthop.,87: 332-344, 1972.87332 1972 [PubMed]

Shelton, R. M.; Rasmussen, A. C.; and Davies, J. E.: Protein adsorption at the interface between charged polymer substrata and migrating osteoblasts. Biomaterials,9: 24-29, 1988.924 1988 [PubMed]

Shinoka, T.; Ma, P. X.; Shum-Tim, D.; Breuer, C. K.; Cusick, R. A.; Zund, G.; Langer, R.; Vacanti, J. P.; and Mayer, J. E., Jr.: Tissue engineered heart valves. Autologous valve leaflet replacement study in a lamb model. Circulation,94: 164-168, 1996.94164 1996

Takato, T.; Harii, K.; Nakatsuka, T.; Ueda, K.; and Ootake, T.: Vascularized periosteal grafts: an experimental study using two different forms of tibial periosteum in rabbits. Plast. and Reconstr. Surg.,78: 489-497, 1986.78489 1986

Tsai, T. M.; Jupiter, J. B.; Kutz, J. E.; and Kleinert, H. E.: Vascularized autogenous whole joint transfer in the hand—a clinical study. J. Hand Surg.,7: 335-342, 1982.7335 1982

Vacanti, C. A.; Langer, R.; Schloo, B.; and Vacanti, J. P.: Synthetic polymers seeded with chondrocytes provide a template for new cartilage formation. Plast. and Reconstr. Surg.,88: 753-759, 1991.88753 1991

Vacanti, C. A., and Upton, J.: Tissue-engineered morphogenesis of cartilage and bone by means of cell transplantation using synthetic biodegradable polymer matrices. Clin. Plast. Surg.,21: 445-462, 1994.21445 1994 [PubMed]

van den Wildenberg, F. A.; Goris, R. J.; and Tutein Nolthenius-Puylaert, M. B.: Free revascularized periosteum transplantation: an experimental study. British J. Plast. Surg.,37: 226-235, 1984.37226 1984

Wray, R. C., Jr.; Mathes, S. M.; Young, V. L.; and Weeks, P. M.: Free vascularized whole-joint transplants with ununited epiphyses. Plast. and Reconstr. Surg.,67: 519-525, 1981.67519 1981

Topics