The AIM of this prospective study was to identify a typical pattern for fluorodeoxyglucose (18F-FDG) uptake in positron emission tomography (PET) to differentiate aseptic loosening (tibial and / or femoral component) from prosthesis infection in painful knee prosthesis.
Patients, methods: 20 patients with painful knee prosthesis underwent PET imaging to evaluate aseptic loosening / prosthesis infection of their knee prosthesis. The interface between bone and surrounding soft tissue or bone was divided into 3 segments each for both the femur and the tibia and in addition for 4 segments reflecting the surrounding periprosthetic soft tissue. FDG uptake in each of the segments was scored (0-3) by two independent observers. The final diagnosis was based on operative findings with subsequent microbiological culture and histological examination.
Results: After surgical revision 6 femoral components and 5 tibial components were found to be loose and prosthetic infection was present in 9 prostheses. In 8 of 9 infected prostheses loosening of the femoral und tibial component occurred. There was no statistically significant correlation between the standardised uptake within each of the segments and the diagnosis of aseptic loosening or prosthesis infection. A differentiated qualitative and quantitative FDG-PET result interpretation divided into five categories was developed. Specificity / sensitivity / positive predictive value / negative predictive value were 93% / 83% / 83% / 93% for aseptic loosening of the femoral component, 87% / 80% / 67% / 93% for aseptic loosening of the tibial component and 82% / 89% / 80% / 90% for infection.
Conclusion: This pilot study shows that FDG-PET is a promising diagnostic tool for patients with painful knee prostheses. There is a good correlation between PET images and the intraoperative and pathology findings. Its clinical value, however, warrants further evaluation in a larger patient population.