Relationships between human patellar tendon and tibial cartilage viscoelastic properties in osteoarthritic knees: an exploratory investigation

Osteoarthritis is well established to be a whole joint disease, and previous studies have found degenerative changes in periarticular tendons around osteoarthritic joints. Thus, we aimed to assess whether potential periarticular adaptations (patellar tendon) are related to intraarticular changes (tibial cartilage) at material-level. As they are both connected to the tibia, we explored the relationships between patellar tendon and tibial cartilage viscoelastic properties obtained from eight osteoarthritic cadaver knees. Six patellar tendon regions and six to eight tibial cartilage samples per knee underwent tensile and indentation sinusoidal measurements, respectively, at 0.1, 0.5, and 1 Hz. Osteoarthritis Research Society International (OARSI) grades were obtained from tibial cartilage histological sections and tested for associations with patellar tendon properties. All knees were deemed osteoarthritic according to tibial cartilage OARSI grades. Comparing the two tissues, we found strong inverse correlations in the phase difference between stress and strain (r: −0.887 to −0.934). Patellar tendon phase difference was also strongly, inversely correlated with average tibial cartilage OARSI grade (r: −0.889 to −0.890). Patellar tendon dynamic modulus was moderately correlated with tibial cartilage dynamic modulus (r: −0.561 to −0.575) and OARSI grade (r: 0.631–0.646). For the first time, we show that the viscoelastic properties of the human patellar tendon are strongly, inversely related to tibial cartilage viscoelastic properties and OARSI grades in osteoarthritic knees, suggesting these two tissues may undergo diverging adaptations with osteoarthritis. These results provide a foundation for more detailed future investigations on patellar tendon-cartilage interactions in knee osteoarthritis.

Sixty cartilage plugs (8 mm diameter, 2.38 ± 0.69 mm thickness) were extracted from up to eight locations (four medial, four lateral) of the tibial plateau for each knee (6–8 samples per cadaver; Fig. 1A; Table 2). Cartilage samples could not always be harvested for all locations of a given knee due to excessive degeneration of some regions. Mechanical indentation testing (Mach-1 v500css, Biomomentum, Laval, QC, Canada) of tibial cartilage samples immersed in phosphate-buffered solution was performed using a custom-made flat-ended steel indenter (1 mm diameter) attached to a 17 N multiaxial load cell (MA235 + MA239, Biomomentum, Laval, QC, Canada). After applying a pre-stress of 12.5 kPa for 15 min, samples underwent a three-step stress-relaxation protocol (each step: strain increments of −5 % current tissue thickness at 100 %/s, followed by 15 min relaxation) (Linus et al., 2024). From final strain, sinusoidal testing including 0.1, 0.5, and 1.0 Hz (4 cycles per frequency) was performed at ±2 % strain amplitude. Phase difference was determined by applying a Fast Fourier transform to the stress and strain data, selecting the phase angle of stress and strain by finding the maximum power amplitude of the Fast Fourier transform (spectra), and then subtracting the strain phase angle from the stress phase angle. After dividing peak-to-peak stresses by strains, dynamic modulus was calculated with the Hayes et al. (1972) equation (assumed incompressible material; Poisson’s ratio: 0.5; Jurvelin et al., 1997) to account for sample and indentation geometry. One overall value for each parameter was obtained by averaging all samples from a given knee.

Patellar tendons (N = 8) were prepared into six dumbbell-shaped samples (three mediolateral × two anteroposterior) of ∼10 mm gauge length and ∼2 mm width. Samples (n = 48) were imaged using an ex vivo micro-computed tomography scanner (SkyScan 1172, Bruker, Kontich, Belgium). Cross-sectional area was obtained with a custom-made, semi-automated MATLAB (R2019b, MathWorks, Natick, MA, USA) code by finding the smallest moving average value of 3 mm length. Patellar tendon samples underwent tensile mechanical testing in a phosphate-buffered solution bath using the same device as for tibial cartilage measurements but with a 250 N uniaxial load cell (MA297, Biomomentum, Laval, QC, Canada). After a comprehensive preconditioning protocol consisting of five sequences of 10 sinusoidal cycles to 6 % strain amplitude (2 %/s), sinusoidal testing including 0.1, 0.5, and 1.0 Hz was performed at ± 0.5 % strain amplitude after 30 min relaxation at 8 % strain (20 cycles per frequency) (Kositsky et al., 2025).

Patellar tendon phase angle and dynamic modulus were originally obtained using sinusoidal fitting (Kositsky et al., 2025). However, for consistency across tissues, tendon data reported in this study were reanalyzed in the same manner as described in Tibial cartilage experiments (analysis method had minimal influence on viscoelastic properties; Supplementary material). As no significant region-specific differences were found for any patellar tendon property (Kositsky et al., 2025), the average of all six regions was taken as the value for each individual knee.

Biomechanics; Connective tissue; Crosstalk; Joint; Osteoarthritis.