gagCEST imaging at 3 T MRI in patients with articular cartilage lesions of the knee and intraoperative validation.

OBJECTIVE: The aim of this study was to compare glycosaminoglycan chemical exchange saturation transfer (gagCEST) of knee cartilage with intraoperative results for the assessment of early osteoarthritis (OA) and to define gagCEST values for the differentiation between healthy and degenerated cartilage. DESIGN: Twenty-one patients with cartilage lesions or moderate OA were examined using 3 T Magnetic Resonance Imaging (MRI). In this prospective study, regions of interest (ROIs) were examined by a sagittal gagCEST analysis and a morphological high-resolution three-dimensional, fat-saturated proton-density space sequence. Cartilage lesions were identified arthroscopically, graded by the International Cartilage Repair Society (ICRS) score in 42 defined ROIs per patient and consecutively compared with mean gagCEST values using analysis of variance and Spearman's rank correlation test. Receiver operating characteristics (ROC) curves were applied to identify gagCEST threshold values to differentiate between the ICRS grades. RESULTS: A total of 882 ROIs were examined and graduated in ICRS score 0 (67.3%), 1 (25.2%), 2 (6.2%) and the merged ICRS 3 and 4 (1.0%). gagCEST values decreased with increasing grade of cartilage damage with a negative correlation between gagCEST values and ICRS scores. A gagCEST value threshold of 3.55% was identified to differentiate between ICRS score 0 (normal) and all other grades. CONCLUSIONS: gagCEST reflects the content of glycosaminoglycan and might provide a diagnostic tool for the detection of early knee-joint cartilage damage and for the non-invasive subtle differentiation between ICRS grades by MRI even at early stages in clinical practice.

Autologous chondrocyte implantation (ACI) for cartilage defects of the knee: A guideline by the working group "Clinical Tissue Regeneration" of the German Society of Orthopaedics and Trauma (DGOU).

BACKGROUND: Autologous chondrocyte implantation (ACI) is an established and well-accepted procedure for the treatment of localised full-thickness cartilage defects of the knee. METHODS: The present review of the working group "Clinical Tissue Regeneration" of the German Society of Orthopaedics and Trauma (DGOU) describes the biology and function of healthy articular cartilage, the present state of knowledge concerning therapeutic consequences of primary cartilage lesions and the suitable indication for ACI. RESULTS: Based on best available scientific evidence, an indication for ACI is given for symptomatic cartilage defects starting from defect sizes of more than three to four square centimetres; in the case of young and active sports patients at 2.5cm(2), while advanced degenerative joint disease needs to be considered as the most important contraindication. CONCLUSION: The present review gives a concise overview on important scientific background and the results of clinical studies and discusses the advantages and disadvantages of ACI. LEVEL OF EVIDENCE: Non-systematic Review.

First clinical study of a novel complete metal-free ceramic total knee replacement system.

BACKGROUND: The aim of the study was to evaluate the safety and efficacy of a novel metal-free ceramic total knee replacement system. METHODS: Thirty-eight primary total knee arthroplasties (TKAs) were performed on 34 patients using the metal-free BPK-S ceramic total knee replacement system with both the femoral and tibial components of an alumina/zirconia ceramic composite. The clinical outcome was evaluated pre- and postoperatively at 3 (n = 32 TKA) and 12 months (n = 32 TKA) using the Knee Society Score (KSS), the Oxford Knee Score and the EQ-5D. Safety analysis was performed by radiological examination and assessment of adverse events. RESULTS: Postoperatively, the KSS, Oxford Knee Score and EQ-5D improved significantly at 3 and 12 months (p < 0.001). Non-progressive partial radiolucent lines were observed in six cases, but there was no osteolysis and no implant loosening. Induction or exacerbation of allergies did not occur during the follow-up. CONCLUSIONS: The metal-free BPK-S ceramic total knee replacement system proved to be a safe and clinically efficient alternative to metal implants in this short-term follow-up study.

Limited evidence of chondrocyte outgrowth from adult human articular cartilage.

OBJECTIVE: Cellular outgrowth from articular cartilage tissue has been described in a number of recent experimental studies. The aim of this study was to investigate the occurrence of cellular outgrowth from articular cartilage explants isolated from adult human donors. METHOD: Macroscopically intact articular cartilage specimens were isolated from adult human donors and cultured either in their native status, or in a cleansed status achieved by forced washing to minimize attaching cells. Additionally, the effect of chemotactic stimuli including cell lysate, High-Mobility-Group-Protein B1 (HMGB-1), Trefoil-factor 3 (TFF3), bone morphogenetic protein-2 (BMP-2), transforming growth factor-ß1 (TGF-ß1), or three-dimensional fibrin or collagen matrices were investigated. Co-cultures with synovial membrane served as a positive control for a source of migratory cells. The occurrence of cellular outgrowth was analyzed by histological examination after a culture period of 4 weeks. RESULTS: Spontaneous cellular outgrowth from cleansed cartilage specimens was not observed at a relevant level and could not significantly be induced by chemotactic stimuli or three-dimensional matrices either. A forming cartilage-adjoining cell layer was only apparent in the case of native cartilage explants with cellular remnants from surgical isolation or in co-culture experiments with synovial membrane. CONCLUSION: The relevance of cellular outgrowth from cartilage tissue is largely absent in the case of adult human articular cartilage samples. A cartilage-adjoining cell layer forming around the explants may instead originate from still attaching cells that remained from surgical isolation.

Functional ankle instability as a risk factor for osteoarthritis: using T2-mapping to analyze early cartilage degeneration in the ankle joint of young athletes.

OBJECTIVE: The aim of this study was to investigate, using T2-mapping, the impact of functional instability in the ankle joint on the development of early cartilage damage. METHODS: Ethical approval for this study was provided. Thirty-six volunteers from the university sports program were divided into three groups according to their ankle status: functional ankle instability (FAI, initial ankle sprain with residual instability); ankle sprain Copers (initial sprain, without residual instability); and controls (without a history of ankle injuries). Quantitative T2-mapping magnetic resonance imaging (MRI) was performed at the beginning ('early-unloading') and at the end ('late-unloading') of the MR-examination, with a mean time span of 27 min. Zonal region-of-interest T2-mapping was performed on the talar and tibial cartilage in the deep and superficial layers. The inter-group comparisons of T2-values were analyzed using paired and unpaired t-tests. Statistical analysis of variance was performed. RESULTS: T2-values showed significant to highly significant differences in 11 of 12 regions throughout the groups. In early-unloading, the FAI-group showed a significant increase in quantitative T2-values in the medial, talar regions (P = 0.008, P = 0.027), whereas the Coper-group showed this enhancement in the central-lateral regions (P = 0.05). Especially the comparison of early-loading to late-unloading values revealed significantly decreasing T2-values over time laterally and significantly increasing T2-values medially in the FAI-group, which were not present in the Coper- or control-group. CONCLUSION: Functional instability causes unbalanced loading in the ankle joint, resulting in cartilage alterations as assessed by quantitative T2-mapping. This approach can visualize and localize early cartilage abnormalities, possibly enabling specific treatment options to prevent osteoarthritis in young athletes.

[Autologous chondrocyte implantation (ACI) for cartilage defects of the knee: a guideline by the working group "Tissue Regeneration" of the German Society of Orthopaedic Surgery and Traumatology (DGOU)]. / Stellenwert der autologen Chondrozytentransplantation (ACT) in der Behandlung von Knorpelschäden des Kniegelenks - Empfehlungen der AG Klinische Geweberegeneration der DGOU.

Autologous chondrocyte transplantation/implantation (ACT/ACI) is an established and recognised procedure for the treatment of localised full-thickness cartilage defects of the knee. The present review of the working group "Clinical Tissue Regeneration" of the German Society of Orthopaedics and Traumatology (DGOU) describes the biology and function of healthy articular cartilage, the present state of knowledge concerning potential consequences of primary cartilage lesions and the suitable indication for ACI. Based on current evidence, an indication for ACI is given for symptomatic cartilage defects starting from defect sizes of more than 3-4 cm2; in the case of young and active sports patients at 2.5 cm2. Advanced degenerative joint disease is the single most important contraindication. The review gives a concise overview on important scientific background, the results of clinical studies and discusses advantages and disadvantages of ACI.

[Total knee arthroplasty for rheumatoid arthritis]. / Endoprothetik des Rheumatischen Kniegelenks.

A total of 150,000 primary total knee arthroplasties are performed in Germany each year. There is only a limited amount of evidence-based data available on possible surgery-related differences between osteoarthritis (OA) and rheumatoid arthritis (RA) of the knee joint. The following review summarizes the recent literature on total knee arthroplasty with a focus on special features of RA patients.

[The indication for radiosynoviorthesis. From the perspective of the nuclear medicine expert, rheumatic orthopedist and internist]. / Indikation zur Radiosynoviorthese. Aus Sicht des Nuklearmediziners, Rheumaorthopäden und Internisten.

Radiosynovectomy or radiosynoviorthesis (RSO), the intra-articular injection of beta-emitting radionuclides (e.g. colloidal preparations of 90-Yttrium, 186-Rhenium or 169-Erbium), is an approved, reliable and easily performed therapy for the treatment of chronic synovitis without harmful side effects. The best clinical results have been obtained in patients with predominantly inflammatory joint disease such as rheumatoid arthritis or reactive arthritis. But RSO is also established to treat pain and persistent effusions after total knee replacement. It also represents an adjuvant therapy in patients with pigmented villonodular synovitis to protect against recurrence following synovectomy. In patients with hemophilia and arthropathy a reduction in joint bleeding is seen and the use of coagulation factor is reduced. The indication for RSO should be made in close cooperation between the referring physician, the rheumatologist and the nuclear medicine expert in the context of a multimodal therapy concept. In this way, success rates of over 80%, with only few side effects, can be achieved, particularly in rheumatoid arthritis, reactive arthritis and hemophilic arthropathy.

Transplanted chondrocytes inhibit endochondral ossification within cartilage repair tissue.

The aim of this study was to investigate the effect of transplanted chondrocytes on endochondral bone formation in cartilage repair tissue. In the knee joint of miniature pigs, cartilage lesions were treated by microfracturing and were then either left empty, covered with a collagen membrane, or treated by matrix-associated autologous chondrocyte transplantation. In control lesions, the subchondral bone plate was left intact (partial-thickness lesion). The repair tissues were analyzed after 12 weeks by histological methods focusing on bone formation and vascularization. The effect of chondrocytes on angiogenesis was assessed by in vitro assays. The presence of antiangiogenic proteins in cartilage repair tissue, including thrombospondin-1 (TSP-1) and chondromodulin-I (ChM-I), was detected immunohistochemically and their expression in chondrocytes and bone marrow stromal cells was measured by quantitative RT-PCR. Significant outgrowths of subchondral bone and excessive endochondral ossification within the repair tissue were regularly observed in lesions with an exposed or microfractured subchondral bone plate. In contrast, such excessive bone formation was significantly inhibited by the additional transplantation of chondrocytes. Cartilaginous repair tissue that resisted ossification was strongly positive for the antiangiogenic proteins, TSP-1 and ChM-I, which were, however, not detectable in vascularized osseous outgrowths. Chondrocytes were identified to be the major source of TSP-1- and ChM-I expression and were shown to counteract the angiogenic activity of endothelial cells. These data suggest that the resistance of cartilaginous repair tissue against endochondral ossification following the transplantation of chondrocytes is associated with the presence of antiangiogenic proteins whose individual relevance has yet to be further explored.

Chondrogenic differentiation of growth factor-stimulated precursor cells in cartilage repair tissue is associated with increased HIF-1alpha activity.

OBJECTIVE: To investigate the chondrogenic potential of growth factor-stimulated periosteal cells with respect to the activity of Hypoxia-inducible Factor 1alpha (HIF-1alpha). METHODS: Scaffold-bound autologous periosteal cells, which had been activated by Insulin-like Growth Factor 1 (IGF-1) or Bone Morphogenetic Protein 2 (BMP-2) gene transfer using both adeno-associated virus (AAV) and adenoviral (Ad) vectors, were applied to chondral lesions in the knee joints of miniature pigs. Six weeks after transplantation, the repair tissues were investigated for collagen type I and type II content as well as for HIF-1alpha expression. The functional role of phosphatidylinositol 3-kinase (PI3K)/mammalian target of rapamycin (mTOR) and mitogen-activated protein kinase (MEK)/extracellular signal-regulated kinase (ERK) signaling on BMP-2/IGF-1-induced HIF-1alpha expression was assessed in vitro by employing specific inhibitors. RESULTS: Unstimulated periosteal cells formed a fibrous extracellular matrix in the superficial zone and a fibrocartilaginous matrix in deep zones of the repair tissue. This zonal difference was reflected by the absence of HIF-1alpha staining in superficial areas, but moderate HIF-1alpha expression in deep zones. In contrast, Ad/AAVBMP-2-stimulated periosteal cells, and to a lesser degree Ad/AAVIGF-1-infected cells, adopted a chondrocyte-like phenotype with strong intracellular HIF-1alpha staining throughout all zones of the repair tissue and formed a hyaline-like matrix. In vitro, BMP-2 and IGF-1 supplementation increased HIF-1alpha protein levels in periosteal cells, which was based on posttranscriptional mechanisms rather than de novo mRNA synthesis, involving predominantly the MEK/ERK pathway. CONCLUSION: This pilot experimental study on a relatively small number of animals indicated that chondrogenesis by precursor cells is facilitated in deeper hypoxic zones of cartilage repair tissue and is stimulated by growth factors which enhance HIF-1alpha activity.

Hypoxia-inducible factor 1alpha is involved in the prostaglandin metabolism of osteoarthritic cartilage through up-regulation of microsomal prostaglandin E synthase 1 in articular chondrocytes.

OBJECTIVE: To investigate crosslinks between catabolic and anabolic pathways in articular cartilage by examining the synthesis and distribution pattern of microsomal prostaglandin E synthase 1 (mPGES-1) in healthy and osteoarthritic (OA) cartilage and analyzing its functional relationship to hypoxia-inducible factor 1alpha (HIF-1alpha) in primary articular chondrocytes. METHODS: Normal cartilage and OA cartilage were subjected to immunohistochemical staining for mPGES-1 and HIF-1alpha. Isolated chondrocytes were cultivated under 21% or 1% O(2). Microarray analysis and quantitative reverse transcriptase-polymerase chain reaction were used to detect genes differentially expressed in chondrocytes cultured under normoxic compared with hypoxic conditions. Immunoblotting was conducted to evaluate intracellular protein levels of mPGES and nuclear accumulation of HIF-1alpha under different oxygen tension levels and with different stimulatory or inhibitory chemical agents. RESULTS: We found enhanced levels of expression of the mPGES-1 gene and an increased number of OA chondrocytes showing staining for mPGES-1 in OA cartilage. Microarray analysis demonstrated that mPGES-1 was among the genes that were up-regulated to the greatest degree in primary chondrocytes exposed to 1% O(2). In vitro, hypoxia led to an enhanced synthesis of mPGES-1, coinciding with a nuclear accumulation of the transcription factor HIF-1alpha. In chondrocyte culture, stimulation with dimethyloxaloylglycine promoted the expression of mPGES-1, phosphoglycerate kinase 1, and cyclooxygenase 2 (COX-2) by stabilizing HIF-1alpha protein levels. A reduction of mPGES-1 synthesis was detected after treatment with 2-methoxyestradiol, correlating with lower HIF-1alpha activity. In contrast, synthesis of mPGES-1 was not influenced by treatment with the specific COX-2 inhibitor NS398. CONCLUSION: These findings suggest that the transcription factor HIF-1alpha is involved in the up-regulation of mPGES-1 and may therefore play an important role in the metabolism of OA cartilage.

Mechanical properties of hyaline and repair cartilage studied by nanoindentation.

Articular cartilage is a highly organized tissue that is well adapted to the functional demands in joints but difficult to replicate via tissue engineering or regeneration. Its viscoelastic properties allow cartilage to adapt to both slow and rapid mechanical loading. Several cartilage repair strategies that aim to restore tissue and protect it from further degeneration have been introduced. The key to their success is the quality of the newly formed tissue. In this study, periosteal cells loaded on a scaffold were used to repair large partial-thickness cartilage defects in the knee joint of miniature pigs. The repair cartilage was analyzed 26 weeks after surgery and compared both morphologically and mechanically with healthy hyaline cartilage. Contact stiffness, reduced modulus and hardness as key mechanical properties were examined in vitro by nanoindentation in phosphate-buffered saline at room temperature. In addition, the influence of tissue fixation with paraformaldehyde on the biomechanical properties was investigated. Although the repair process resulted in the formation of a stable fibrocartilaginous tissue, its contact stiffness was lower than that of hyaline cartilage by a factor of 10. Fixation with paraformaldehyde significantly increased the stiffness of cartilaginous tissue by one order of magnitude, and therefore, should not be used when studying biomechanical properties of cartilage. Our study suggests a sensitive method for measuring the contact stiffness of articular cartilage and demonstrates the importance of mechanical analysis for proper evaluation of the success of cartilage repair strategies.

Ex vivo gene therapy approaches to cartilage repair.

Degeneration of articular cartilage is one of the great clinical challenges that still lack a convincing therapeutic solution. Traumatically induced lesions and final stages of osteoarthritis with substantial loss of cartilage tissue require the generation of new hyaline cartilage, because significant endogenous repair does not occur. Current joint-preserving surgical approaches, however, mostly lead to fibrous repair tissue that is rapidly degraded. In experimental studies, several differentiation factors have been shown to stimulate chondrogenesis, promoting the formation of functionally acceptable cartilage-like repair tissue. Cell-mediated transfer of the respective genes may ideally combine the supply of a chondrogenic cell population with the production of such factors directly at the site of the lesion. The treatment of earlier disease stages aims both at the protection of the remaining cartilage from further degradation and a stimulation of cartilage anabolism. Various studies proved the administration of anti-catabolic or anti-inflammatory cytokines into joints affected by cartilage destruction to be beneficial. However, the clinical utility of intraarticular protein application is limited by the short half-lives of such proteins in vivo. The transfer of cells over-expressing the respective genes may provide a more sustained delivery of the therapeutic molecules and thus be the more economic option.

Cartilage regeneration by gene therapy.

Damage of articular cartilage is a frequent clinical problem and is commonly considered to be irreversible. Full-thickness defects may lead to the formation of fibrous repair tissue of minor mechanical quality, while partial-thickness lesions hardly show any repair response. Surgical approaches often fail to restore the articular surface, facing the problem of incomplete chondrogenesis or rapid degradation of the repair tissue. However, advances in molecular biology have revealed the potential of growth factors, differentiation factors, and cytokines in directing cellular differentiation and metabolic activity. Anabolic factors including members of the TGF-beta superfamily, IGF-1, FGF, or HGF have proven their potential to stimulate chondrogenesis and synthesis of cartilage-specific matrix components, allowing the formation of a hyaline cartilage-like repair tissue in experimental studies. In addition, anti-catabolic or anti-inflammatory molecules, such as IL-4, IL-10, IL-1Ra, and TNFsR may also exert beneficial effects by inhibiting excessive cartilage degradation. Although it is questionable whether regeneration of hyaline cartilage implying a complete restoration of the articular surface by a tissue that is identical with the original can ever be achieved, all these molecules have been considered as suitable tools for cartilage repair. The transfer of the respective genes into the joint, possibly in combination with the supply of chondroprogenitor cells, might be an elegant method to achieve a sustained delivery of such therapeutic factors at the required location in vivo. This review focuses on the therapeutic molecules, the suitability of different viral and non-viral vectors for intraarticular gene transfer and the lessons learned from gene therapy studies on various animal models.

Bone regeneration in critical size defects by cell-mediated BMP-2 gene transfer: a comparison of adenoviral vectors and liposomes.

Large bone defects resulting from nonunion fractures or tumour resections are common clinical problems. Recent studies have shown bone morphogenetic protein-2 (BMP-2) gene transfer using adenoviral vectors to be a promising new therapeutic approach. However, comparative studies of different vectors are required to identify the optimal system for possible clinical trials. This study compares the use of liposome-mediated and adenoviral gene transfer for the generation of autologous BMP-2-producing bone marrow stromal cells (BMSC). Primary BMSC isolated from the rat femur were treated ex vivo with either an adenovirus or a liposome carrying human BMP-2 cDNA. The genetically modified cells were evaluated in vitro and transplanted into critical size defects in the rat mandible in vivo. BMSC treated with a reporter gene vector or untreated BMSC served as controls. The newly formed tissue was analysed by in situ hybridization, radiography and immunohistochemistry. Both groups of genetically modified cells produced BMP-2 for at least 2 weeks, and markers of new bone matrix such as osteopontin and osteocalcin were observed within 2 weeks following gene transfer. In the liposome group, the critical size defects were found completely healed at 6 weeks after the gene transfer, whereas the more efficient adenoviral gene transfer allowed for complete bone healing within 4 weeks. None of the three control groups showed bone healing, not even after 8 weeks. Thus, both liposome-mediated and adenoviral BMP-2 gene transfer to primary BMSC are suitable methods to achieve the healing of critical size bone defects in rats. As liposomes have proven sufficient for this purpose and offer several advantages over any other vector, such as ease of preparation, theoretically no limitation of the size of the DNA, and less immunological and safety problems, they may represent the best vector system for future clinical trials of bone regeneration by BMP-2 gene therapy.

Osteophyte development--molecular characterization of differentiation stages.

OBJECTIVE: Osteophytes are non-neoplastic osteo-cartilaginous protrusions growing at the margins of osteoarthritic joints. They can not only be considered as in situ repair tissue, but also represent an excellent in vivo model for induced cartilage repair processes. Our focus was to identify different steps of osteophyte development via analysis of expression patterns of marker genes of chondrocytic differentiation. DESIGN: We performed an extensive analysis of the presence and expression of matrix components using histochemical, immunohistochemical and in situ hybridization technology. RESULTS: Four different stages of osteophyte formation could be identified based on histomorphological and cell biological parameters: starting from mesenchymal condensates, chondrogenic differentiation is indicated by the onset of Col2A and aggrecan expression (stage I). Stage II shows fibrocartilage with an admixture of cartilaginous and fibrous matrix components such as Col2 and aggrecan on the one hand and Col1 on the other hand. The proliferating osteophyte (stage III) shows a zonal organization similar to the fetal growth plate cartilage with extensive chondrocyte hypertrophy in the zones next to ongoing endochondral bone formation. 'Mature' osteophytes (stage IV) resembled largely articular hyaline cartilage with a predominance of Col2 and aggrecan and Col6 found mainly pericellularily. CONCLUSIONS: The development of osteophytes is a good in vivo model to pursue chondrocyte differentiation from pluripotent mesenchymal cells to mature or hypertrophic chondrocytes in situ in the adult. The analysis of marker molecules of mesenchymal differentiation allows to identify different stages of repair tissue development and the transformation from fibrous tissue to neo-cartilage. Tissue architecture and matrix composition in mature osteophytes suggests that metaplastic neo-cartilagenous tissue might be one potential source of cartilage repair tissue in the adult joint.

Fibroblast-mediated delivery of growth factor complementary DNA into mouse joints induces chondrogenesis but avoids the disadvantages of direct viral gene transfer.

OBJECTIVE: To assess the advantages and disadvantages of a direct adenoviral and a cell-mediated approach to the induction of cartilage formation in joints by transfer of growth factor genes. METHODS: Adenoviral vectors carrying insulin-like growth factor 1 (IGF-1) or bone morphogenetic protein 2 (BMP-2) complementary DNA were constructed and applied to primary human and murine chondrocytes or fibroblasts. Transgene expression was quantified by enzyme-linked immunosorbent assay. Direct injection of these vectors or AdLacZ, a reporter gene vector, into mouse knee joints was compared with the transplantation of syngeneic fibroblasts (infected ex vivo with the same vectors) with respect to virus spread, immune response, and cartilage formation by use of histologic, immunohistochemical, and molecular analyses. RESULTS: AdIGF-1 and AdBMP-2 efficiently infected all cell types tested. Human cells secreted biologically relevant levels of protein over a period of at least 28 days. Direct transfer of AdLacZ into mouse knee joints resulted in positively stained synovial tissues, whereas AdLacZ-infected fibroblasts settled on the surface of the synovial membranes. Inadvertent spread of vector DNA into the liver, lung, and spleen was identified by nested polymerase chain reaction in all mice that had received the vector directly; this rarely occurred following fibroblast-mediated gene transfer. Direct injection of AdBMP-2 induced the synthesis of new cartilage in periarticular mesenchyme, accompanied by extensive osteophyte formation. When AdBMP-2 was administered by injecting ex vivo-infected fibroblasts, cartilage formation was observed only in regions near the injected cells. AdIGF-1 treatment did not lead to morphologic changes. Importantly, fibroblast-mediated gene transfer avoided the strong immune response to adenovirus that was elicited following direct application of the vector. CONCLUSION: Our results indicate that cell-mediated gene transfer provides sufficient BMP-2 levels in the joint to induce cartilage formation while avoiding inadvertent vector spread and immune reactions.