Accuracy and Precision of Volumetric Matching Micromotion Analysis (V3MA) for tibial component migration measurement in total knee arthroplasty: An experimental study and clinical proof of concept in six patients.

Abstract

Introduction: Early migration and continuous migration in the first two years is considered a predictive factor of aseptic loosening in total knee arthroplasty (TKA). The current gold standard to measure implant migration is roentgen stereophotogrammetric analysis (RSA). However, RSA is complex to implement in common practice due to the use of a calibration cage, trained radiology personnel, and insertion of tantalum markers in the bone. Therefore, computed tomography based migration analysis (CTBMA) has been proposed as an alternative method. CTBMA measures the displacement of implants relative to the host bone over time in CT images. We developed a volume-based and marker-free CTBMA called VoluMetric Matching Micromotion Analysis (V3MA). The objective of this study was to validate V3MA in tibial components by evaluating the accuracy and precision in vitro and to provide clinical proof of concept.

Methods: Three experiments (A, B, C) were conducted. In all experiments, tibial component migration was measured with both V3MA and RSA. (A) Accuracy: 21 different positions of a tibial component attached to a micromanipulator were assessed with respect to the cadaveric human tibial bone. (B) Precision: Repeated zero-migration measurements of 4 sets consisting of a cadaveric human tibial bone with cemented tibial component. (C) Clinical proof of concept: Clinical data of 6 patients. Data analysis included descriptive statistics (A, B, C), Bland-Altman analysis (A, C), and interclass correlation coefficient (ICC) (C). The primary outcome was the accuracy (A) and precision (B) of both V3MA and RSA of tibial components in vitro. Secondary outcomes included agreement of clinical tibial migration measured with V3MA and RSA (C), as well as an explorative interobserver variability (C).

Results: (A) V3MA accuracy (t*RMS) ranged between 0.02 mm and 0.15 mm for translations and was 0.03° in Y-rotation. For RSA, the accuracy ranged between 0.09 mm and 0.33 mm for translations and was 0.25° in the Y-rotation. V3MA accuracy was significantly better in anteroposterior translation (p = 0.01). (B) V3MA precision (SD) ranged from 0.01 to 0.06 mm for translations and 0.02 to 0.07° in rotations. RSA precision ranged from 0.00 to 0.06 mm for translations and 0.04 to 0.25° in rotations. V3MA precision was significantly better in all directions, except superoinferior translation. (C) The mean difference (limits of agreement) between V3MA and RSA was -0.05 mm (-0.49 to 0.39) and -0.14° (-0.79 to 0.51) for total translation and total rotation, respectively. There was a good correlation between two observers using V3MA (ICC: 0.76, 95%CI: 0.58 to 0.87).

Discussion and conclusion: The accuracy and precision of V3MA were comparable to RSA. However, results may differ for other CT resolutions and in other joints. Furthermore, V3MA was feasible and comparable to RSA in clinical practice, with good reliability between observers. Therefore, V3MA seems a promising alternative to RSA in migration measurement of tibial components in TKA.