Can Walking Recover Cartilage T2 Relaxation Time in Healthy and Osteoarthritic Knees

Introduction: Articular cartilage is the connective tissue that covers the ends of long bones located within diarthrodial joints. Its primary functions are to enable low friction articulation while providing a surface capable of bearing loads that can often exceed a person’s body weight. These biomechanical functions are driven by the tissue’s poroelastic structure and the resulting interstitial fluid pressure. However, this same fluid pressure also drives fluid from the cartilage, which leads to a reduction in pressure, load transfer to the solid matrix, and increased cartilage deformation. Previous work has shown that these ex-vivo observations of acute loading induced deformation and change in composition can also be observed in vivo. To date, the only in vivo mechanism known to recover cartilage deformation and composition is static unloading (e.g., lying supine).

Recent ex vivo research has suggested that cartilage sliding is also capable of deformation and composition recovery [2]. In this work we aimed to evaluate whether joint activity (i.e., walking), which induces knee cartilage sliding in vivo, is capable of composition recovery. Furthermore, we aimed to determine if the response to joint activity differs between participants with and without symptomatic knee osteoarthritis (OA). We hypothesize that joint activity will restore cartilage composition following a period of static loading, and that this recovery response is severely compromised in osteoarthritis.

Materials and

Methods: Following Institutional Review Board approval, we recruited 8 participants without symptomatic knee OA and 8 with symptomatic knee OA. Each completed the Knee Injury and Osteoarthritis Outcome Score (KOOS) for pain and symptoms. Participants then underwent a fixed task series, followed by magnetic resonance (MR) imaging, to examine the relationship between task and cartilage composition.

Task Series. Participants performed the following fixed task series: static loading (standing for 30 min), active loading (walking at 110 steps/min for 10 min), a second period of static loading, and finally static unloading (lying supine for 50 min), shown visually in Fig 2.

MR Scanning. After each task a multi-echo multi-slice T2-weighted scan (TE = 12.5…75 ms and TR = 4480 ms) was acquired using a 3T MR scanner with a 15 channel Tx/Rx knee coil. The T2-weighted scan had a resolution of 0.5 x 0.5 x 3 mm.


Analysis. MR scans were pre-processed by first up sampling to 0.25 x 0.25 x 1.5 mm followed by registration to a reference image via a 6-degree rigid transformation. T2 relaxation times were calculated using a mono-exponential decay function (Fig 1). T2 relaxation times over 100 ms and R2 values under 70% were excluded from analysis, consistent with previous studies.

We defined a region of interest as the cartilage-cartilage contact area between the tibia and femur (Fig 1). Segmentation was performed on three contiguous slices for both the medial and lateral compartments. Non-parametric statistical tests were conducted due to non-normal and heteroscedastic data.

Results, Conclusions, and Discussions: The impact of task on T2 relaxation time in healthy and symptomatic participants is shown in Fig 2. Qualitatively it appears that higher T2 relaxation times are associated with OA and the medial compartment. Task type appears to have a minor influence on T2 relaxation time.

Effect of joint health. OA cartilage in the medial and lateral compartment was 17 ms and 8 ms greater than healthy cartilage. A Kruskal-Wallis test revealed significant effects of joint health on T2 relaxation time for both medial (p < 0.0008) and lateral compartments (p < 0.0002). Effect of compartment. Medial compartment T2 relaxation time was 3.9 ms (healthy) and 12.8 ms (OA) greater than the lateral compartment. A Wilcoxon signed-rank test detected significant effects of compartment for both healthy (p < 0.008) and OA (p < 0.008).

Effect of task. A Friedman test (p > 0.1) confirmed our observation that task did not have a significant effect on T2 relaxation time in both the health and OA cohort; however, two notable observations emerged: (1) T2 relaxation time increased during standing and (2) T2 relaxation time decreased during walking in healthy participants. These preliminary trends are interesting as they are opposite to what we anticipated. Given the well-supported assumption that T2 is directly related to water content and that standing will drive fluid from cartilage, we expected lower T2 relaxation times following standing for 30 min. In addition, we hypothesized that walking would recover fluid and increase T2 relaxation time relative to standing. In future work we will expand our participant population and regional analysis to determine if these trends are significant and if they correlate with cartilage deformation.

Conclusions: Task variation had a minor, non-significant effect on T2 relaxation time for both cohorts. The effects of joint health and compartment greatly outweighed those of task demonstrating that while loading history may influence T2 relaxation time it may not be necessary to control for this when attempting to distinguish healthy and OA cartilage.

Acknowledgements (Optional): Tejus Surendran/tsurendr@andrew.cmu.edu - Carnegie Mellon University
Daniel K. White /dkw@udel.edu - University of Delaware
Axel C. Moore/axelm@andrew.cmu.edu - Carnegie Mellon University