Knee meniscus injury is frequent, resulting in over 1 million surgeries annually in the United States and Europe. Loss of meniscus tissue has been associated with early onset knee osteoarthritis due to an increase in joint contact pressures in meniscectomized knees; thus, meniscal injury also leads to damage on articular cartilage surfaces within the knee joint. Clinically available replacement strategies range from allograft transplantation to synthetic implants. Although short-term efficacy has been demonstrated with some of these treatments, factors such as long-term durability and chondroprotective efficacy remain unpredictable. Because of the near-avascularity of this fibrocartilaginous tissue and its intrinsic lack of healing, tissue engineering has been proposed as a solution for meniscus repair and replacement. In particular, bioactive and mechanical stimulation during culture can be used to enhance mechanical properties and drive extracellular matrix content toward native tissue levels. Before an effective tissue-engineering strategy for treating meniscal lesions can be translated to the clinic, the United States Food and Drug Administration (FDA) requires rigorous preclinical testing of the safety and efficacy of these technologies in a large animal model. However, guidance documents for meniscus repair technologies are nonexistent and no gold-standard animal model has been established for preclinical meniscus research. Thus, toward translating tissue engineering technologies to clinical applications, the global objectives of this research are: 1) to enhance self-assembled neomeniscus and neocartilage mechanical and biochemical properties through application of bioactive or mechanical stimuli, and 2) to identify appropriate implantation and integration methods in a large animal model to validate the repair capacity of the tissue-engineering strategies developed in vitro. To address these objectives, this research: 1) enhanced the mechanical and extracellular matrix properties of neomenisci using bioactive factors TGF-1, chondroitinase ABC, and lysyl oxidase-like 2 (collectively termed “TCL”), in addition to lysophosphatidic acid (LPA); 2) improved neocartilage mechanical and biochemical properties through sequential application of two forms of mechanical stimuli (uniaxial tension and fluid-induced shear); 3) established the Yucatan minipig as a suitable preclinical animal model for meniscus research by showing that several gross morphological, mechanical, and biochemical properties were within ranges of values reported in human menisci; and 4) evaluated the efficacy of neocartilage implanted in the medial meniscus of Yucatan minipigs toward repairing meniscal lesions.
An approach employing bioactive stimuli to enhance both extracellular matrix content and organization of neomenisci toward augmenting their mechanical properties was investigated. Specifically, self-assembled neomenisci were treated with TCL+LPA. Supporting our hypothesis, TCL+LPA treatment synergistically improved circumferential tensile stiffness and strength, significantly enhanced collagen and pyridinoline crosslink contents per dry weight by 61% and 81% over controls, respectively, and achieved tensile anisotropy (circumferential/radial) values of neomenisci close to 4. This study utilized a combination of bioactive stimuli for use in neomeniscus tissue engineering studies that improved functional properties to achieve anisotropic tensile properties, which is a crucial mechanical aspect of the native meniscus, providing a promising path toward deploying these neomenisci as functional repair and replacement tissues.
To investigate whether a hyperelastic model could capture changes to native and engineered meniscus functional properties to better inform meniscus tissue engineering strategies, three different hyperelastic models were applied to mechanical and biochemical data from native tissue treated with bioactive treatments, namely collagenase. Experimental data from neomenisci treated with bioactive factors in a previously published study, specifically TCL+LPA treated neomenisci, were also examined using hyperelastic analysis. Small-strain analysis, which is largely phenomenological, is typically used to model the meniscus; however, the meniscus experiences large strains (~40%) under normal loading conditions. Collagenase treatment on native meniscus samples led to significant decreases in tensile properties and collagen content compared to untreated controls. The three hyperelasticity models examined were Neo-Hookean, Yeoh, and fiber-reinforced neo-Hookean models; it was hypothesized that a microstructural, hyperelastic model would best describe the experimental data and provide model parameters that would correlate with the biochemical content of both engineered and native tissues. Out of the three, the fiber-reinforced Neo-Hookean model, which incorporates tissue microstructural properties, was found to be the best model based on goodness-of-fit. Positive correlations between both collagen content (ρ=0.81) and pyridinoline crosslinking (ρ=0.69) and the fiber modulus (γ), which is a stress-like material property determined from mechanical tests of the tissue, were identified. Interestingly, the strongest correlation existed between the collagen to GAG ratio (ρ=0.84) and the nonlinearity parameter (α). Together, these data provide a hyperelastic model that allows for deeper understanding of meniscal function with regard to its structural properties, and aids tissue engineers in the design of functional neomenisci toward their use in repair and replacement technologies.
The manipulation of neocartilage construct mechanical properties toward native tissue values can also be achieved with applied mechanical stimuli during culture. Uniaxial tensile stress, for example, has been found to improve tensile stiffness and strength of bovine-derived neocartilage constructs, while fluid-induced shear (FIS) improved constructs’ compressive stiffness. It was hypothesized, first, that combining two mechanical stimulation strategies, specifically, uniaxial tension and FIS, would improve multiple neocartilage mechanical properties more effectively compared to using one stimulus alone and, secondly, that order of stimulus application would lead to differences in neocartilage construct properties. It was found that the combination of both mechanical stimuli led to synergistic improvements to tensile properties and compressive stiffness. Specifically, constructs exhibited tensile Young’s modulus, ultimate tensile strength, and compressive aggregate modulus values that were 180%, 161%, and 131% higher than nonstimulated controls, respectively. Furthermore, combining the stimuli had additive effects on the extracellular matrix content of constructs, compared to unstimulated controls. Finally, it was determined that applying tension before FIS was more effective toward improving tissue mechanical properties, specifically tensile properties, when compared to applying FIS before tension. Overall, the use of complementary dual mechanical stimuli synergistically increased neocartilage properties and a dependence on the order of application was identified; thus, researchers should consider applying these or other forms of complementary mechanical stimuli toward engineering neocartilage with robust properties.
The frequency of knee meniscus injuries and surgical procedures motivates tissue engineering attempts and the need for suitable animal models. Despite their extensive use in cardiovascular research and the existence of characterization data for the menisci of farm pigs, the farm pig may not be a desirable preclinical model for the meniscus due to its rapid weight gain. However, minipigs, such as the Yucatan breed, are suitable for in vivo experiments due to their slower growth rate compared to farm pigs and similarity in body weight to humans. Despite this, characterization of minipig knee menisci is lacking. Both medial and lateral Yucatan minipig knee menisci were extensively characterized in terms of structural and functional properties within different regions to inform the Yucatan minipig’s suitability as a preclinical model for meniscal therapies. Gross morphological properties of minipig menisci that fell within ranges seen in native human tissue included meniscal width and peripheral height. Additionally, per wet weight, biochemical evaluation revealed 23.9-31.3% collagen (COL; 22% for human) and 1.20-2.57% glycosaminoglycans (GAG; 0.8% for human). Also, per dry weight, pyridinoline crosslinks (PYR) were 0.12-0.16% (0.12% for human). Biomechanical testing revealed circumferential Young’s modulus of 78.4-116.2MPa (100-300MPa for human), circumferential ultimate tensile strength (UTS) of 18.2-25.9MPa (12-18MPa for human), radial Young’s modulus of 2.5-10.9MPa (10-30MPa for human), radial UTS of 2.5-4.2MPa (1-4MPa for human), aggregate modulus of 157-287kPa (100-150kPa for human), and shear modulus of 91-147kPa (120kPa for human). Anisotropy indices ranged from 11.2-49.4 and 6.3-11.2 for tensile stiffness and strength (approximately 10 for human), respectively. Regional differences in mechanical and biochemical properties within the minipig medial meniscus were observed; specifically, GAG, PYR, PYR/COL, radial stiffness, and Young’s modulus anisotropy varied by region. The posterior region of the medial meniscus exhibited the lowest radial stiffness, which mirrors what is seen in humans and corresponds to the most prevalent location for meniscal lesions. Overall, similarities between minipig and human menisci support the use of minipigs for meniscus translational research.
Finally, to investigate the repair capacity of the approaches developed to this point, this work concluded with a large animal study examining the effects of tissue-engineered constructs in a meniscus defect. Allogeneic, self-assembled constructs were implanted into partial-thickness medial meniscus defects in the Yucatan minipig using novel surgical methods. Implants showed an exceptional safety profile and did not lead to a systemic immune response. As hypothesized, the surgical approach that was developed allowed for defect creation and implant delivery; additionally, implant treatment increased Young’s modulus values for the interface between native tissue and repair tissue in the pocket where the tissue was embedded by 51% compared to untreated controls. However, the allogeneic implants did not lead to increased defect repair tissue mechanical and biochemical properties because defects in the untreated control group also exhibited robust healing. Thus, modifications to surgical techniques used to implant engineered tissues within preclinical animal models, in addition to changes to the defect model, might be required to better investigate of the repair capacity of self-assembled constructs and translate our approach from the bench to the clinical bedside.