Biomechanical Characteristics of Bovine Articular Cartilage under a Novel Wear Regime

Introduction: Damaged or diseased articular cartilage has limited repair opportunities. The development of functional articular cartilage repair techniques is conditional on the attainment of biotribological and biomechanical properties similar to native tissue. Typically, studies have been limited to either the biomechanical aspect or the biotribological aspect. Until recently, biotribological studies of soft tissue have also been restricted to only one motion instead of the concomitant translation and rotation experienced in vivo.[1-2] A method for subjecting an articular surface to both translational and rotational wear was developed.[3-4] This study describes characterizing the viscoelastic behavior of the tissue through indentation stress relaxation and dynamic cyclic testing. Materials and Methods: Bovine femoral condyles were sectioned transversely into four quadrants, anchored in a bath of artificial lubricant (20% bovine calf serum, 20mM EDTA, and 0.9% saline), and equilibrated for 20min.[6] The Instron 1321 servohydraulic materials testing equipment (Instron Corp., Canton MA) retrofitted with MTS TestStar II controller (MTS Corp., Eden Prairie MN) provided a means for completing tissue thickness measurements, stress relaxation (SR), dynamic cyclic testing (DCT), and inducing wear. The applied strains during cyclical testing were based on maintaining the tissue within the linear range of viscoelasticity and preventing liftoff of the indenter.[7-9] Tissue Thickness Measurement: Tissue thickness was determined by translating a needle probe through the cartilage surface to the subchondral bone.[10] Stress Relaxation and Dynamic Cyclical Testing: SR and DCT were performed at the same location as the thickness measurement preand post-wear. To properly seat the porous, 1.5mm Φ indenter, the tissue was compressed 10% of the thickness following a 10gm preload. SR for 20min was followed by 20min of 1Hz DCT at 2.5% compression. Defect: A 5mm Φ defect was created at the apex of curvature in randomly chosen specimens. Inducing Wear: An initial normal stress of 0.5 to 1.5MPa (25N and 40N, respectively) was applied to the specimen via an Airpel air cylinder (Model E16, Airpot Corp., Norwalk CT). For 1hr, a stainless steel wear plate was translated 10mm for linear wear while the specimen was oscillated 10° at 1Hz for rotational wear. SR and DCT were performed a second time on the affected surface after 1hr of recovery. Property Determination: The stress response was characterized by the convolution of the reduced stress relaxation function, G(t), and an elastic response, σe, to allow a complete determination of the stress over all time, σ[ε(t);t]=G(t)*σe[ε(t)], where the elastic response is a function of strain, ε. The continuous form of G(t) proposed by Fung [5] was used, while the elastic response, σe, was represented by the exponential equation, σe(ε)=A(eBε–1). A MATLAB 7.1 (The MathWorks, Natick MA) program was developed to curve-fit the SR data obtaining the QLV parameters (A, B, c, τ1, and τ2).[11-13] The constant, A, was determined by curve-fitting the elastic response to the ramping portion of the SR data using a least squares nonlinear algorithm. The remaining terms, B, c, τ1, and τ2 were determined by simultaneously curve-fitting the SR response. A separate program was created to evaluate the dynamic data using FFT analysis to obtain |G| and tan δ at the testing frequency of 1Hz. Statistical Analysis: A repeated measures ANOVA was performed to determine significance with p<0.05. Results: Five specimens were tested in each of 8 permutation groups: high / low load, no defect / defect, rotation / no rotation. Visually, only specimens with a defect showed fibrillated tissue along a defect edge. Comparisons of the curvefit parameters showed a significant decrease in prevs post-wear with the elastic response, A, and viscous response, C.(Table 1) In addition, the short term relaxation response, τ1, showed a significant decrease between no defect (0.801±0.13 sec) and a defect (0.679±0.16 sec).