Bone bruise assessment in knees with ACL lesions: semi-quantitative scores outperform quantitative evaluation

Bone bruise is a common consequence of trauma. Its presence, size, and shape correlate with articular damage and symptoms evolution. Purpose of this study was to identify, among the most used quantitative and semi-quantitative evaluation strategies, the most reliable method to evaluate bone bruise in knees with anterior cruciate ligament (ACL) injury.

Knee Magnetic Resonance Imaging (MRI) of 100 ACL-injured patients were selected. Bone bruise characteristics were assessed by five different evaluators applying Whole Organ Magnetic Resonance Scoring (WORMS) and MRI Osteoarthritis Knee Score (MOAKS). Bone bruise area was measured freehand using Intellispace^® PACS Enterprise. A topographic assessment in the coronal view was performed to evaluate reproducibility between different operators. Inter and intra-rater agreement were calculated with Fleiss-k for WORMS, MOAKS, and topographic evaluation, and with intra-class correlation (ICC) for area measurement.

Bone bruise presence was reported in 63-65% of cases in the lateral femoral condyle, 42-48% in the medial femoral condyle, 80-82% in the lateral tibial plateau, 53-61% in the tibial sulcus, and 40-42% in the medial tibial plateau, with a substantial to almost perfect agreement (Fleiss-k 0.67-0.89). Inter-rater agreement was classified as moderate for WORMS and MOAKS (Fleiss-k = 0.46 and 0.45, respectively), while the area measurement resulted in a poor agreement (ICC = 0.44). Intra-rater agreement was classified as almost perfect for WORMS and MOAKS (Fleiss-k = 0.91 and 0.90, respectively), while the area measurement resulted in a good agreement (ICC = 0.79).

Semi-quantitative scores outperformed the quantitative evaluation of bone bruise. WORMS and MOAKS both resulted in a moderate inter-rater agreement, while bone bruise area assessment showed poor reliability. Intra-rater reliability confirmed the advantages of the semi-quantitative approach versus the quantitative one, being good for the area measurement while almost perfect when both WORMS and MOAKS scores were used to assess bone bruise in ACL-injured knees.