Introduction

Although revision rates of unicondylar knee arthroplasty (UKA) are higher compared to total knee arthroplasty, two-thirds of UKA revisions are performed without clinical evidence other than patient reported pain [1]. Furthermore, the main UKA failure modes leading to revision can be linked to osteoblastic activity which can be quantified with SPECT-CT imaging. Such imaging thus has clear potential to further our understanding of such failures towards reduced revision rates. Therefore, this study aimed to compare post-operative osteoblastic activity patterns in the tibia between successful UKA’s and UKA’s slated for revision surgery.

Methods

SPECT-CT scans were acquired in 23 patients (Age=61±8 years; BMI=29.23±4.71 kg/m²) with well-functioning medial UKA’s at two years post-operatively. Additionally, SPECT-CT scans were also acquired in 10 patients slated for medial UKA revision surgery (Age=42±25 years; BMI=27.90±4.53 kg/m²). For these 10 patients, the interval time between initial UKA and scanning was on average 4.91 years (SD: 4.77, range: 0.96-13.96).
Using validated SCReg platform [2], spatial voxel-to-voxel correspondences were obtained between all SPECT-CT scans through a registration scheme. Osteoblastic activity was normalized according to activity in distal femur. Then, a statistical map was created by performing a Mann-Whitney U test on each voxel to compare successful UKA’s and UKA’s slated for revision. Finally, for each group, an aggregate map was created by averaging, for each voxel, normalized osteoblastic activity values over all specimens of the group.

Results

Main regions with significantly different normalized osteoblastic activity (p<0.05) were located on (1) the lateral tibial plateau (volume of 2448 voxels = 7044 mm³), (2) the antero-medial cortex at implant level (volume of 452 voxels = 1293 mm³) and (3) the lateral femoral epicondyle (volume of 962 voxels = 2752 mm³). For all these regions, higher activity was observed in the UKA’s slated for revision compared to successful UKA’s (2.46±0.40 vs 1.50±0.22 in (1); 3.07±0.75 vs 1.73±0.50 in (2); 2.30±0.25 vs 1.54±0.16 in (3)).

Discussion

Significantly higher normalized activity on the antero-medial cortex of the tibia was expected as that region is known to be directly involved in two primary modes of UKA failures, i.e. tibial loosening and anterior tibial pain. Concerning the observed differences in the lateral femoral and tibial compartment, we hypothesize this to be related to lateral osteoarthritis progression as another common mode of UKA failure. Finally, the fact that this study cohort included different UKA designs should be noted as a limitation.

Conclusions

Analysis of post-operative normalized osteoblastic activity using SPECT-CT imaging has diagnostic potential to confirm or infirm UKA revision surgery by elucidating failure mechanisms.

References:
[1] Kennedy JA et al. (2020). KSSTA, 28: 3926-3934.
[2] Dandois F et al. (2020). BMC Musculoskelet Disord, 21