Articular Cartilage Healing

General Concepts

Healing refers to restoration of the structural integrity and function of the tissue after injury or disease. Repair refers to the replacement of damaged or lost cells and matrix with new cells and matrix, a process that does not necessarily restore the original structure or function of a tissue [1]. Regeneration may be considered a special form of repair in which the cells replace lost or damaged tissue with a tissue identical to the original tissue. It has been suggested that, with exception of bone fractures, most injuries and diseases of musculoskeletal tissues do not stimulate regeneration of the original tissue. The limited potential of articular cartilage for regeneration and healing has been appreciated for over 2 centuries. In 1743, Hunter stated "From Hippocrates to the present age, it is universally allowed that ulcerated cartilage is a troublesome thing and that when once destroyed it is not repaired". The response of cartilage to tissue damage is limited, and natural repair responses from adjacent tissues are unable to produce tissue with the same morphologic, biochemical, and biomechanical properties of articular cartilage.

Some of the earlier experimental work provided us with classic concepts. Bennett, Bauer, and Maddock studied the repair of articular cartilage defects, as well as the reaction of normal joints of adult dogs to surgically created defects of articular cartilage and implantation of joint mice. No defects completely healed in the 28-week period. Calandruccio and Gilmer studied the healing of defects in immature dogs.

They recognized the phenomena of matrix flow, proliferation of superficial cartilage cells, proliferation of connective tissue from the base of the defect, and formation of new chondrones as phenomena in the healing process. Using radioactive labeling, other research has recognized increased collagen synthesis, as well as increased proteoglycan synthesis adjacent to the area of cartilage injury. However, despite this evidence for increased proteoglycan and collagen synthesis, no significant contribution was made to the healing of lacerated defects in mature rabbit cartilage. Biochemical studies have been directed at characterization of the collagen types in the repair tissue. Furukawa et al. were the first to recognize the early formation of type I collagen with later increases in type II collagen content (the characteristic collagen for articular cartilage). Although type II later predominated, type I collagen always persisted significantly. It addition, there was less hexosamine in the repair tissue, suggesting that the fibrous nature that usually developed in the healing tissue was due to a lack of proteoglycans rather than a change in the type of collagen [7].

Natural Healing of Articular Cartilage Defects

The major limiting factor in the successful rehabilitation of any joint after injury and disease is the failure of osteochondral defects to heal. Three mechanisms have been recognized as possible contributors to articular cartilage repair. Intrinsic repair (from within the cartilage) relies on the limited mitotic capability of the chondrocyte and a somewhat ineffective increase in collagen and proteoglycan production. Extrinsic repair comes from mesenchymal elements from the subchondral bone participating in the formation of new connective tissue that may undergo some metaplastic change to form cartilage elements. The third phenomenon known as "matrix flow" may contribute to equine articular cartilage repair by forming lips of cartilage from the perimeter of the lesion that migrate toward the center of the defect [9,10]. A publication has suggested that it is possible for synovial membrane (and the subsynovial tissue) to act a source of repair [11]. This experimental study was done with partial-thickness defects in rabbits. With the use of chondroitinase ABC to digest the GAG chains of cartilage proteoglycans (which enhances adhesions) an addition of growth factors and mesenchymal cells that were considered synovial in origin came into the defect. There was no transformation of these cells into chondrocytic elements and at this stage it would seem, although synovial membrane could be a potential source of cells, effective healing is not going to be achieved.

The depth of the injury (full or partial thickness), size of defect, location and relation to weight-bearing or non-weight-bearing areas, and the age of the animal influence the repair and remodeling of an injured joint surface [4,6,12<].

With a partial-thickness defect, some degree of repair with increased glycosaminoglycan synthesis and increased collagen synthesis occurs [6]. However, the repair process is never completely effective. It has been reported in humans that complete repair of chondromalacia of the patella is thought to occur if matrical depletion and surface breakdown are minimal [13]. However, more recent work with arthroscopic debridement of partial-thickness defects in humans questions any actual regeneration [14]. It should also be recognized that superficial defects are not necessarily progressive and do not necessarily compromise joint function.

With full-thickness defects, the response from the adjacent articular cartilage varies little from that of superficial lesions and provides only limited repair necessary to replace dead cells and damaged matrix at the margins of the wound. These defects heal by in-growth of subchondral fibrous tissue that may or may not undergo metaplasia to fibrocartilage [9,12,15-17]. Subchondral bone defects either heal with bone that grows up into the defect or fill in with fibrocartilaginous in-growth. Duplication of the tide mark in the calcified cartilage layer is rare, and adherence of the repair tissue to surrounding non-injured cartilage is often incomplete.

The size and location of articular defects have a significant effect on the degree of healing achieved; several studies in the horse demonstrate this. Convery et al first reported that large defects were less likely to heal [12]. A more recent study distinguished between large (15 mm2) and small (5 mm2) full-thickness lesions in weight-bearing and non-weight-bearing areas of the radiocarpal, intercarpal and femoropatellar joints [9]. At 1 month, small defects were filled with poorly organized fibrovascular repair tissue; by 4 months, repair was limited to an increase in the amount of organization of this fibrous tissue; and by 5 months, small radiocarpal and femoropatellar lesions were hardly detectable because of the combinations of matrix flow and extrinsic repair mechanisms. Large lesions showed good initial repair but at 5 months perilesional and intralesional subchondral clefts developed.

It has also been demonstrated that induced osteochondral defects (6.5 mm) on the non-weight-bearing aspect of the distal lateral trochlea in the horse healed with fibrocartilaginous tissue at a faster rate, and more completely, than those on the weight-bearing proximomedial trochlear ridge of the talus [18].

The repair tissue that forms after full-thickness injury to hyaline cartilage, or as a natural repair process in joints with OA, is primarily composed of type I compared to type II collagen at least at 4 months [19,20]. Identification of type II collagen is a critical biochemical factor distinguishing hyaline cartilage from repair fibrous tissue and fibrocartilage. It is believed that the presence of an abnormal subchondral bone plate and the absence of a tide mark reforming may create a stiffness gradient, and that shear stresses of the junction of the repair tissue and underlying bone develop. The propagation of such shear stresses would lead to the degradation of repair fibrocartilage and exposure of the bone. This mechanical failure has been observed experimentally and clinically [9,21]. Of some interest are data from the author’s laboratory on the analysis of tissue 12 months after creation of full-thickness articular cartilage defects in a weight-bearing area. In a study looking at the effectiveness of sternal cartilage grafting long-term, the repair tissue in the non-grafted defects at 12 months consisted of fibrocartilaginous tissue with fibrous tissue in the surface layers, as had been seen in control defects at 4 months (Fig. 115-1). On biochemical analysis, the repair tissue of the non-grafted defects had a mean type II collagen percentage of 79% compared with being non-detectable at 4 months [20]. On the other hand, the glycosaminoglycan content expressed as milligrams of total hexosamine per gram of dried tissue was 20.6 +/- 1.85 mg/gm compared with 26.4 +/- 3.1 mg/gm at 4 months and 41.8 +/- 4.3 mg/gm DW in normal equine articular cartilage [20,22].

Figure 115-1. Photomicrographs of repair tissue in full-thickness articular cartilage defects in equine radial carpal bone. (A) at 4 months, (B) at 12 months. There is a mixture of fibrous (superficially) and fibrocartilaginous (deeper) tissue in the defect.

The fibrocartilaginous repair seen in normal full-thickness defects, therefore, is biomechanically unsuitable as a replacement-bearing surface and has been shown to undergo mechanical failure with use. The lack of durability may be related to faulty biochemical composition of the old matrix and incomplete remodeling of the interface between old and repaired cartilage or to increased stress in the regenerated cartilage because of abnormal remodeling of the subchondral bone plate and calcified cartilage layer. Although some work implies that it may be possible to reconstitute the normal collagen type in articular repair [22] it is clear that low levels of glycosaminoglycan persist and these are important components in the overall composition of the cartilage matrix.

It does need to be recognized that the presence of a cartilage defect may not represent clinical compromise. In the equine carpus, it has been shown that up to 30% of articular surface loss of an individual bone may not compromise the successful return of a horse to racing [21]. However, a loss of 50% of the articular surface or severe loss of subchondral bone leads to a significantly decreased prognosis.

It is also to be recognized that the inadequate healing response may not necessarily apply to immature animals nor to non-weight bearing defects. An example is the young patient after surgery for osteochondritis dissecans (OCD) in which impressive healing responses or at least functional responses are obtained. This may be related to increased chondrocytic capacity for mitosis and matrix synthesis and the presence of intracartilaginous vascularity. Complete restoration of the ultrastructure and surface configuration in a hinge-like gliding joint surface such as the femoropatellar joint may be unnecessary for clinical soundness compared with the more severe loading on an osteochondral defect located on the weight-bearing portion of the medial condyle of the femur, for instance. It has been suggested that increasing age may affect the response of cartilage to injury in humans because the ability of the chondrocytes to synthesize and assemble matrix micromolecules could decline with age [1]. Buckwalter and Mow cite a study of transplanted chondrocytes, suggesting that older chondrocytes produce a more poorly organized matrix than do younger chondrocytes [1], and other studies demonstrate that the proteoglycan synthesized by the chondrocytes change with age [5,23].

Techniques Attempted for Modulation of Articular Cartilage Healing

Defect Debridement

An old clinical generalization is that superficial defects in articular cartilage do not heal and that full-thickness defects heal through metaplasia of granulation tissue arising at either the articular margin or from the subchondral marrow spaces below [16]. This led to the common practice of clinicians curetting partial-thickness defects into full-thickness ones to stimulate "healing". As mentioned previously, the importance of small partial-thickness defects has been questioned and the quality of the new replacement tissue after full-thickness curettage is now recognized as uncertain and commonly defective [8]. In addition to experimental work, arthroscopic observations (follow-up examinations) by the author of the healing of clinical articular defects confirm the failure of these defects to heal effectively in weight-bearing areas. Because of this inadequate healing response, it is suggested that debridement of partial-thickness defects to full-thickness defects in aggressive debridement to bone is contraindicated [24]. This opinion is also influenced by the observation that superficial defects are not necessarily progressive and do not necessarily compromise joint function.

Having said that, cartilage and bone debridement are common during arthroscopic surgery. As a simple rule, loose fibrous tissue or exposed loose bone should be removed from full-thickness defects. If cartilage is elevated or separated from the bone, it should be debrided back to where it attaches. Debridement should continue down to firm, normal-appearing subchondral bone. Maintaining as much subchondral bone as possible keeps the bone plate and overlying cartilage repair tissue contoured to the normal congruency of the opposing joint surface, thereby enhancing the chance of healing cartilage tissue persisting. It is important, however, that the remaining bone is viable. Crumbly, brownish bone should also be removed by debridement, either using hand instruments or motorized equipment. As has been mentioned, consensus appears to favor not debriding partial-thickness fibrillated cartilage. An argument exists that chondroplasty reduces the possibility of damaged cartilage leaching degraded cartilage-matrix fragments, including collagen, proteoglycan, and cellular components, to the synovial fluid where they increase synovitis and concurrent lameness [25].

Full-thickness cartilage lesions are debrided to remove residual portions of the calcified cartilage layer. Research has confirmed that persistence of calcified cartilage retards the development of well attached cartilage repair tissue from the subchondral bone and surrounding cartilage (Fig. 115-2) [26].

Figure 115-2. Photomicrographs of repair tissue at 12 months in full-thickness articular cartilage defects on the medial femoral condyle where (A) the calcified cartilage layer has been completely removed, and (B) the calcified cartilage layer has not been removed.

Chondroplasty

Resection of the protruding surface strands of partial-thickness cartilage fibrillation has been promoted as a mechanism to reduce cartilage-derived debris entering the synovial environment [25,27-29]. Motorized synovial abraders are used smooth the surface of the more seriously damaged cartilage. The residual cartilage then presents a more uniform non-clefted surface, which may be more durable and incite less synovitis than the large surface area presented by multiple strands of fibrillated cartilage. Although this seems a simple concept, a paucity of evidence exists documenting any discrete benefit, either in reducing synovial levels of fragmented proteoglycan and collagen or in decreasing the symptoms of synovitis. Clinically, chondroplasty to smooth articular cartilage in areas of fibrillation seems to reduce the evidence of persistent effusion when patients go back into exercise, but because of a lack of controlled experimental data, the technique is controversial. It is important that, if done, resection should only involve the fibrillated surface and not be aggressively pursued down to the subchondral bone for reasons previously discussed. Articular cartilage was shaved on the underside of the rabbit patella with no evidence for repair (and no evidence of degenerative changes) in either the superficially or deeply shaved areas [23,30]. Ultrastructural studies after arthroscopic cartilage shaving however, question any regeneration and suggest deleterious effects [14].

Spongialization

Removal of sclerotic subchondral bone from the base of full-thickness defects has been advocated to achieve the formation of new tissue in the defect [31]. In defects associated with OA, the subchondral plate is often sclerotic and ischemic, and complete removal of the subchondral plate (spongialization) could potentially provide an increased opportunity for tissue filling of cartilage defects. However, long-term follow-up of experimental defects in horses with significant debridement of subchondral bone implies that the biomechanical changes result in stresses that disrupt the new repair tissue [22]. Spongialization has been replaced by less invasive techniques, particularly microfracture.

Subchondral Bone Drilling

Subchondral bone drilling has been used with rationale similar to spongialization (access to cancellous bone through a sclerotic or ischemic subchondral bone plate). Controlled studies have been done with experimentally created full-thickness and partial-thickness defects in horses [32,33]. Subchondral bone drilling of full-thickness cartilage defects of the equine third carpal bone was followed by fibrocartilaginous repair tissue of quality and quantity superior to the fibrous tissue of non-drilled defects, but satisfactory functional healing was not achieved [32]. Subchondral drilling did not significantly improve partial-thickness cartilage healing [33].

Abrasion Arthroplasty

The use of superficial intracortical debridement (arthroscopic abrasion arthroplasty), as opposed to deep cancellous debridement, for sclerotic degenerative lesions has been advocated in humans [34]. The concept is somewhat controversial and argues the necessity to expose cancellous bone to reach both bone supply and pluripotential cells. Abrasion arthroplasty is an arthroscopic technique used to remove eburnated bone by exposing intracortical vessels in the tibial and femoral surfaces of the knee. The organizing hematoma forms over these abraded surfaces and then differentiates into fibrocartilage. The abrasion goes approximately 2 mm into cortical bone. Deep exposure into cancellous bone with removal of subchondral bone is considered counter-productive as it causes a fibrous repair without cartilage [35]. Microfracture is the favored technique for gaining access to cancellous bone elements now.

Subchondral Bone Microfracture

The use of microfracture, or micropicking, as it has been referred to in equine arthroscopy, has many of the advantages associated with subchondral drilling, including focal penetration of the dense subchondral plate to expose cartilage defects to the benefits of cellular and growth-factor influx, as well as improving anchorage of the new tissue to the underlying subchondral bone and, to some extent, surrounding cartilage [36-38]. The simplicity of microfracture comes from the use of a tapered awl (Linvatec, Largo, FL; Arthrex, Naples, FL) instead of a parallel-sided twist drill. Using the awl eliminates the need for powered instrumentation and gives accurate control. A tapered entry into the subchondral marrow space is achieved. The microfracture awl should penetrate the subchondral bone deeply enough to provide ready access to the marrow spaces, thereby maximizing cellular and anabolic growth-factor delivery (Fig. 115-3). The microfracture awl also tends to make a crater in the subchondral bone, which may play a role in better attachment of the cartilage repair tissue [39]. Microfracture holes are generally placed 3 to 5 mm apart and cover the entire debrided area in a cartilage lesion. It is also important to microfracture the subchondral bone on the perimeter of the cartilage lesion to encourage new tissue at the junction of repair tissue and residual cartilage. The technique has become popular in human arthroscopy and is now frequently compared to chondrocyte transplantation as one of the two most frequently employed techniques to improve cartilage healing. [36,40,41]. One study compared use of microfracture to autologous chondrocyte implantation in a randomized trial. At 2 years’ follow-up, both groups had significant clinical improvement, but according to the SF-36 physical component score, the improvement in the microfracture group was significantly better than that in the autologous implantation group [42].

Figure 115-3. Arthroscopic view of a microfractured articular defect.

An experimental study in the horse documented improvement in the quantity of tissue and the type II collagen content at 4 and 12 months after microfracture of full-thick37]. Improvements in early gene expression of type II collagen and mRNA at 8 weeks have also been demonstrated [43].ness cartilage defects (Fig. 115-4) [

Figure 115-4. Photomicrographs (same level of magnification) of full-thickness defects on the medial femoral condyle that have been (A) simply debrided and (B) debrided and microfractured, demonstrating significantly more filling of articular defect that has been microfractured.

The technique clearly has advantages over forage and transplantation methods, including ease of application using arthroscopy, use of a simple hand tool, the relative economics of the equipment required, and the apparent increase of cartilage repair tissue that develops after the procedure.

Transplantation Procedures

It is generally accepted that most debriding and marrow-stimulating techniques result in fibrocartilage formation with modest biomechanical capabilities. The use of supplemental free cells, various vehicles containing cells or entire tissue such as periosteum or cartilage grafts have been advocated to improve on the modest impact that local manipulative procedures have on both the quality and quantity of cartilage repair tissue. Transplantation procedures can be classified according to the origin of the transplanted tissue: 1) periosteal grafting; 2) periochondral grafting; 3) autogenous cartilage (articular, sternal, or auricular transplantation); 4) osteochondral transplantation; 5) chondrocyte transplantation; and 6) pluripotential stem cell transplantation. These transplantation techniques have practical limitations. Arthrotomy is required for insertion of periosteum, perichondrium, intact cartilage, and osteochondral grafts. Similarly, tissue-engineered analogs such as chondrocytes cultured on collagen, PGA or PGA/PLA, or other synthetic materials such as hyaluronan membranes are also difficult or impossible to implant arthroscopically.

Periosteum

The chondrogenic potentials of perichondrium and periosteum have been used experimentally to restore large osteochondral defects in the rabbit and dog [44-47]. Work in the horse showed that chondroid tissue could be produced after implantation of free periosteal autografts (but not perichondreal autografts) [48]. However, when periosteal autografts were placed into osteochondral defects and fixed with fibrin glue, results were unsatisfactory, with the predominant tissue in most defects being fibrous tissue. The poor results and high type I collagen content were related to the formation of adhesions [17].

Sternal Cartilage Autografts

The use of sternal cartilage autografts has been investigated in horses with fixation of the autografts using biodegradable pins [20,22]. On histologic examination at 4 months, the repair tissue appeared morphologically similar to hyaline cartilage compared with the non-grafted tissue, which was fibrocartilage tissue with fibrous tissue on the surface. On biochemical analysis, the repair tissue with sternal grafting was composed predominantly of type II collagen (non-grafted predominantly type I collagen) and the total GAG content of repair tissue (42.6 =/- 5.9 mg/g DW) was not significantly different from that of normal articular cartilage and was significantly greater than that of non-grafted defects (26.4 +/- 3.1 mg/g DW). Unfortunately, a long-term (12-month) study (together with high-speed treadmill exercise) resulted in repair tissue degenerating and the formation of horizontal and vertical clefts. The type II collagen content was not significantly different from control defects and the mean total GAG content had dropped to 29.1 mg/g DW in grafted defects compared with 19.1 mg/g DW in controls.

Osteochondral Grafts

The use of osteochondral autografts and allografts has cycled through several periods of clinical interest. Originally, autogenous osteochondral shell allografts were used, but there were major problems of limited availability of autogenous osteochondral graft tissue and donor site morbidity. Few locations in humans or animals can sacrifice considerable areas of a joint as osteochondral grafts. The use of osteochondral shell allografts overcomes the limitations of donor site morbidity, and fresh osteochondral hemiarthroplasty shell allografts have been used for advanced degenerative osteoarthritis in humans [49]. These are entire femoral condylar grafts and immunogenicity is always of concern.

Mosaicplasty using autogenous osteochondral dowel grafts has become popular. Osteochondral dowels are harvested arthroscopically from less weight-bearing regions of the same joint, and these dowels are inserted to reconstruct a relatively congruous joint surface with articular cartilage [50-53].

Several instrument systems are available for the harvest and implantation of osteochondral dowels in humans and animals, including the mosaicplasty system (ACUFEX-Smith and Nephew, Andover, MA); the osteochondral autografts transfer system (OATS-Arthrex, Naples, FL); and the consistent osteochondral repair system (COR-Innovasif BE). Frequently "mosaicplasty" is used as an umbrella term when it is specific to the Smith and Nephew instrumentation. The benefits of this technique include immediate weight-bearing capabilities, relatively good integration of the bony portion of the dowel, and the long-term data available from clinical trials [52,53].

The mosaicplasty technique has been used in the horse and it has also been shown that age influences the outcome [54]. A case report has been published of the use of mosaicplasty for the repair of a subchondral bone cyst in the medial femoral condyle in the horse [55]. The technical difficulties associated with careful graft harvesting and the precision and crafting needed for heterotopic graft insertion in the recipient bed will probably detract from wider clinical application [56]. Also empty spaces that naturally form between inserted osteochondral dowels heal poorly and allow synovial fluid entry into the bony tunnels of adjacent dowels.

Chondrocyte Transplantation

Autogenous chondrocyte implantation is one of the few FDA-approved tissue engineering techniques to treat articular cartilage injury in humans. It is a two-stage procedure in which articular cartilage biopsies are harvested arthroscopically from minimally weight-bearing regions of the injured knee, propagated ex vivo in cell culture, and later implanted under an autogenous periosteal tissue flap [57]. The indications include focal defects and osteochondritis dissecans [58,59]. The delivery of cells requires an arthrotomy and the harvest and suture attachment of a periosteal flap, which is tedious and technically demanding. One study compared chondrocyte transplantations secured with a periosteal patch to defects treated with periosteal patches without chondrocyte implantation and found no difference between the healing tissues after 1 year.60 Long-term studies for patients treated with autogenous chondrocyte repair have reported good outcomes [61-63].

In horses, chondrocyte implantation techniques have been examined in a variety of matrix carrier vehicles [64-66]. Initial research trials indicated enhancement of healing with chondrocyte implantation using a fibrin vehicle [64] but further tissue engineering approaches with collagen matrix scaffolds did not provide a satisfactory improvement in repair [65].

Experimental work in the horse with a modified ACI technique (culturing of chondrocytes on a collagen membrane) has yielded positive results at 12 and 18 months [67]. However, more recently, even better results have been obtained with re-implantation of morselized autogenous articular cartilage particles in a one-step procedure.68 Such techniques have real potential application in dogs and horses.

Pluripotential Mesenchymal Stem Cell Transplantation

The use of pluripotent cells to enhance cartilage repair has been investigated for several years. Initial studies in the rabbit indicated mesenchymal stem cells (MSCs) could enhance cartilage repair [69]. Follow-up work in small animals demonstrated that MSCs can be partially induced down chondrocyte lineages [70]. Studies in the horse indicate bone marrow-derived MSCs can be harvested and cultured for sufficient time and defined media to differentiate toward a chondrocytic lineage [71], but in vivo studies in the horse report little advantage in the healing at 8 months in a femoral trochlear ridge cartilage defect model [72]. Further work is being done in this area.

Growth Factors

Several naturally occurring polypeptide growth factors play an important role in cartilage homeostasis. The differentiating and matrix anabolic-promoting activity of IGF-1 and TGF-beta are particularly important in counteracting the degradatory and catabolic activities of cytokines, serine proteases, and neutral proteinases. Studies in the horse have largely focused on IGF-1 as TGF-beta has shown synovitis and osteophyte development in animal studies. Slow release of IGF-1 from fibrin composites has been shown to allow for enhanced cartilage repair [73]; however, ideal repair tissue is not obtained and it has been suggested that IGF-1 seems to have better application in combination chondrocyte implantation where more complete cartilage repair develops [74].

Conclusion

At the present time most cartilage repair techniques in small animals would revolve around endogenous manipulation of healing, such as the use of debridement and microfracture. More advanced techniques including cartilage transplantation and autogenous chondrocyte implantation, as well as autogenous osteochondral grafting techniques, show promise.