Residual strain development in biological tissue is believed to result from remodeling in response to repetitive loading. This study hypothesized that differences in in-vivo loading between levels of the bovine tail result in differences in intervertebral disc (IVD) annulus fibrosus (AF) microstructural remodeling. The hypothesis was tested by quantifying tail musculature using clinical computed tomography and tissue microstructure using collagen fiber crimp period, which has previously been correlated with residual strain.

Three bovine tail segments (levels c1 through c6) were imaged using a clinical computed tomography (CT) scanner followed by removal of muscle and harvest of IVDs. The discs were frozen, and transverse cryosections were obtained. Additionally, tangential plane cryosections were obtained from the inner and outer zones of the AF.

Transverse CT slices corresponding to each joint level thresholded for both disc and muscle tissue and analyzed in MATLAB. First, the centroid of the disc image was calculated to use as an origin. Then the disc area and moments of inertia about the flexion extension axis and lateral bending axis were calculated. Total muscle area was then calculated, along with muscle moments of inertia relative to the disc centroid. All muscle parameters were normalized by those of the corresponding disc.

Cryosections were imaged using an inverted light microscope equipped with crossed polarizing filters and a digital camera. A MATLAB routine was used to perform Fourier transform analysis on user selected lines of interest in the transverse micrographs, yielding average fiber crimp period in the inner and outer AF. Micrographs from tangential sections were opened in ImageJ, and fiber orientation angles were measured manually.

Muscle moments of inertia were analyzed using a two-way ANOVA with disc level and axis as dependent variables. Normalized muscle area was analyzed with a one-way ANOVA with disc level as a dependent variable. A two-way ANOVA, with disc level and zone (inner versus outer) was used to analyze collagen fiber crimp period and collagen fiber angle.

Normalized muscle moment of inertia showed significant effects of both level and axis (p < 0 .001), decreasing at distal levels, and being lower about the flexion-extension axis than the lateral bending axis. Normalized muscle cross section showed a visible, but not significant (p=0.0721) decreasing trend with disc level. Fiber crimp period had significant effects of both level and zone (p < 0 .001), and was significantly longer in the outer zone than inner at all levels. Significant decrease in crimp period at distal levels were seen in the outer AF, but not the inner. While fiber angle was significantly (p < 0 .001) higher in the inner AF (36±6.6°) than outer AF (24±3.5°)), there was no significant effect of level.

Fiber crimp period in the AF has previously been correlated with residual circumferential strain, with larger crimp period corresponding to increased residual tension. The present study suggests that at proximal levels of the tail, where peak compressive and bending stresses in the AF (as inferred from normalized muscle area and moments of inertia respectively) are greatest, there is more accumulation of residual strain.

As the intervertebral disc is largely avascular, needle injection is the most practical method for delivery of therapeutic agents used in treatments for degenerative disc disease. Intradiscal pressure increases during injection, and insufficient recovery time prior to needle retraction may result in injectate leakage. In order to determine the maximum pressure and post-injection recovery time for a given injection volume and rate, an analytical model of intradiscal injection was developed and calibrated experimentally.

A governing equation was derived defining intradiscal pressure as a function of effective permeability, initial elastic stiffness, nonlinear stiffness term, and injection rate. The equation was solved using a fourth order Runge-Kutta routine with a 0.05s time step and a ramp-dwell injection.

The model was calibrated by performing controlled intradiscal injections on five bovine caudal intervertebral discs. Three had adjacent vertebrae intact, while two were separated from vertebrae and constrained between porous stainless steel platens. A syringe driven by a linear actuator was used to inject phosphate buffered saline through a 21g hypodermic needle inserted radially into the disc to a depth of one half of the disc diameter. Injection was performed at a rate of 75μL/s to a volume of 250μL followed by a 240s dwell. Fluid pressure was recorded during both the injection phase and subsequent recovery phase. For each experimental pressure vs time trace, model parameters were varied in order to obtain an optimal fit.

The model was run with the average parameter values across a grid of possible injection protocols, with injection volume ranging from 30 to 300μL and injection time ranging from 0.1 to 5s. For each case, peak pressure and time required to reach a 1kPa threshold were recorded.

Experimentally measured peak pressure ranged from 68 to 88kPa. Pressure at the end of the 240s dwell ranged from 49 to 69kPa. There was no apparent difference between discs with and without endplates. Leakage of fluid following needle retraction was observed in all specimens. Experimental data were well fit by the analytical model, which predicted higher peak pressure and longer recovery time with increasing volume, from approximately 1500s at 30μL to nearly 3000s at 300μL. The model was nearly insensitive to injection rate.

The experimental data confirm pressurization of the disc during injection and injectate leakage resulting from insufficient recovery time. The model predicts that the time required to recover to below threshold leakage pressure is impractically long for both laboratory and clinical injection protocols. Similar behavior with and without endplates confirms that fluid flow is limited by permeability of the tissue itself, not the boundary conditions. Slow recovery is likely attributable to the fact that peak injection pressures were lower than the hydraulic swelling pressure of the nucleus pulposus, which has been reported to be approximately 140kPa.

Due to the high swelling pressure of the nucleus pulposus, it is unlikely that intradiscal injection procedures can be performed without substantial injectate leakage following needle retraction.