Summary
Non-invasive measurement of knee implant loosening using knee loading, CT scans and 3D-image analysis.
Abstract
Introduction
Despite its success up to 13% of the patients will require revision surgery within 10 years after Total Knee Arthroplasty (TKA). In the majority, the reason for revision is aseptic loosening of the tibial component. The main tests used to diagnose aseptic TKA loosening currently have a sensitivity or specificity higher than 70-80% and merely demonstrate secondary and a-specific effects. This leads to 20-30% of patients undergoing unnecessary, risky and expensive revision surgery. Detecting actual displacement of the implant with respect to the bone may be a more reliable and direct approach to evaluate TKA loosening.
Therefore, a new non-invasive technique was developed to evaluate implant loosening. This study presents this new technique and evaluates its reproducibility and reliability in a laboratory cadaveric study.
Methods
For the purpose of this study a prototype loading device was developed to apply consecutive varus- and valgus loads to the knee (20 Nm). The advanced 3D-image analysis software was developed by the Biomedical Engineering and Physics department. It consists of a three-step approach to visualize and quantify prosthesis loosening using CT-images, i.e. (1) segmentation of the tibia and tibial TKA component in the varus scan, (2) registration of the tibia and tibial segmentations to the valgus scan and (3) calculation of varus-to-valgus displacements and rotations of the tibial component relative to the tibial bone. These displacements and rotations are quantified using the mean Target Registration Error (mTRE), helical axis rotation and Maximum Total Point Motion (MTPM).
In this experimental study, first the reproducibility errors of this technique were quantified using a single frozen solid specimen in which implant displacement and rotation were assumed to be absent. This specimen was implanted with a TKA and scanned in ten slightly different orientations without load. Any observed implant displacement was designated as reproducibility errors caused by the noise content of the CT images, the segmentation and/or registration procedure.
Secondly, to evaluate reliability, ten thawed cadaveric specimens were first implanted with loosely fitted TKA’s and scanned under varus load first followed by a valgus load scan. Thereafter, the implants were fixed to the bone using bone cement and scanned in the same manner to investigate possible differences between the loose and the fixed tibial component. Differences were tested using a paired sample t-test.
Results
Reproducibility errors, expressed in terms of mTRE, helical-axis rotation and MTPM were 0.073 mm (SD 0.033), 0.129 degrees (SD 0.039 degrees) and 0.116 mm (SD 0.031), respectively.
Comparing the mTRE, helical axis rotation and MTPM of the loose situation relative to the fixed situation resulted in mean differences of 0.463 mm (SD 0.279 mm; p=0.001), 1.769 degrees (SD 0.868 degrees; p<0.001) and 1.339 mm (SD 0.712 mm; p<0.001), respectively. All displacements and rotation changes in the loose situation were larger than the reported reproducibility errors.
Conclusion
The results of this cadaveric study suggest that this new and non-invasive method is both reproducible and reliable for detection of TKA loosening. Yet, a patient study is needed, and underway, to investigate future clinical utilization.