T1ρ, T2 mapping, and EPIC-µCT imaging in a canine model of knee osteochondral injury

Authors: 
Franklin SP, Stoker AM, Lin ASP, Pownder S, Burke E, Bozynski CC, Kuroki K, Guldberg RE, Cook JL, Holmes SP
J Orthop Res. 2019 Aug 20. doi: 10.1002/jor.24450.

The dog is the most commonly used large animal model for the study of osteoarthritis. Optimizing methods for assessing cartilage health would prove useful in reducing the number of dogs needed for valid study of osteoarthritis and cartilage repair.

Twelve beagles had critical-sized osteochondral defects created in the medial femoral condyle of both knees. Eight dogs had T1ρ and T2 MRI performed approximately 6 months after defect creation.

Following MRI evaluations, all twelve dogs were humanely euthanatized and cartilage samples were obtained from the medial and lateral femoral condyles, medial and lateral tibial plateaus, trochlear groove, and patella for proteoglycan and collagen quantification. Equilibrium partitioning of an ionic contrast (EPIC)-µCT was then performed followed by histologic assessment of the knees. Correlations between T1ρ, T2, EPIC-µCT and proteoglycan, collagen, and histology scores were assessed using a multivariate analysis accounting for correlations from samples within the same knee and in the same dog. Pearson correlation coefficients were calculated to assess strength of significant relationships. Correlations between µCT values and biochemical or histologic assessment were weak to moderately strong (0.09-0.41; p<0.0001 to 0.66). There was a weak correlation between the T2 values and cartilage proteoglycan (-0.32; p=0.04). The correlation between T1ρ values and cartilage proteoglycan were moderately strong (-0.38; p<0.05) while the strongest correlation was between the T1ρ values and histological assessment of cartilage with a correlation coefficient of 0.58 (p<0.0001).

These data suggest that T1ρ shows promise for possible utility in translational study of cartilage health and warrant further development in this species.