ABSTRACT
UNLABELLED: In not infected knee prostheses bone scintigraphy is a possible method to diagnose mechanical loosening, and therefore, to affect treatment regimes in symptomatic patients. However, hitherto studies showed controversial results for the reliability of bone scintigraphy in diagnosing loosened knee prostheses by using asymptomatic control groups. Therefore, the AIM of our study was to optimize the interpretation procedure and to evaluate the accuracy using results from revision surgery as standard. METHODS: Retrospectively, we were able to examine the tibial component in 31 cemented prostheses. In this prostheses infection was excluded by histological or bacteriological examination during revision surgery. To quantify bone scintigraphy, we used medial and lateral tibial regions with a reference region from the contralateral femur. RESULTS: To differentiate between loosened and intact prostheses we found a threshold of 5.0 for the maximum tibia to femur ratio of the both tibial regions and a threshold of 18% for the difference of the ratio of both tibial regions. Using these thresholds, values of 0.9, 1, 0.85, 1, and 0.94 were calculated for sensitivity, specificity, negative predictive value, positive predictive value, and accuracy, respectively. To get a sensitivity of 1, we found a lower threshold of 3.3 for the maximum tibia to femur ratio. CONCLUSION: Quantitative bone scintigraphy appears to be a reliable diagnostic tool for aseptic loosening of knee prostheses with thresholds evaluated by revision surgery results being the golden standard.