Cartilage microstructure assessed by high resolution micro-computed tomography in non OA knees

Abstract

Introduction and aims: In osteoarthritis (OA), loss of the articular cartilage, changes of the cortical subchondral plate and alterations in subchondral bone architecture are observed. At the microscopic level, no established 3D method exists to measure cartilage volume and thickness. The objective was to develop an easy-to-implement method for precise and accurate measurements of volume and thickness of cartilage from micro-CT images and to apply the new method to a larger amount of data obtained from human cadavers. Methods: Twenty five left cadaveric knees without OA as determined from radiographs using an adapted Kellgren-Lawrence grading scale (grade<2) from 13 women and 9 men (mean age 80.9AE10.0 years) were included in the study. After dissection, vertical cores (7 mm in diameter) were extracted in the lateral region of the lateral tibial plateau. The native cores were imaged with micro-CT (Skyscan 1172 â , voxel size 10.2mm) and the following parameters were measured: cartilage volume (Cart.Vol, mm 3 ) after manual contouring of the cartilage followed by interpolation with the software from Skyscan: CtAnâ. Cartilage thickness (Cart.Th, mm) was determined manually (man) and by using the thickness plugin of BoneJ (plug), which was originally developed to measure trabecular thickness. In a subgroup of 20 cores, short term reproducibility (RMSCV%) was determined from 3 acquisitions with intermediate repositioning. One scan per specimen was analyzed 3 times by the same observer (intra-observer: IA) or by three different observers (inter-observer: IE). In order to assess accuracy in a subgroup of 10 cores, holes with diameters of 2mm, 3mm, and 4mm were artificially generated with a dermatological punch. The accuracy error was determined by comparing nominal and measured hole sizes (Wilcoxon test and Bland and Altman plots). Results: Mean Cart.Vol was 69.6AE10.9mm 3 . Mean Cart.Th was 1.75AE0.29mm (man) and 1.78AE0.26mm (plug). The thickness variation was 0.13AE0.06 mm and the maximum thickness was 1.94AE0.26mm (plug). The correlation between the two Cart.Th analysis techniques was rZ0.94 (p<10 -4 ). For the measurement chain, the Cart.Vol_RMSCV% was 1.35%. For the image analysis, the Cart.Vol_RMSCV% were 0.42% and 1.80% for IA and IE, respectively. MeanAESD of the nominal hole volumes were 7.2AE2.0mm 3 (2mm), 16.7AE4.1mm 3 (3mm), 27.9AE8.6mm 3 (4mm), and the measured volumes were 7.7AE2.4 mm 3 (2mm), 17.1AE5.75 mm 3 (3mm) and 27.6AE8.6mm 3 (4mm), with no significant differences. The mean biases were +0.49AE1.6mm 3 (2mm), +0.41AE4.2mm 3 (3 mm) and -0.34AE4.4mm 3 (4 mm). Conclusion: In conclusion, with micro-CT, it was possible to measure the cartilage volume and thickness in 3D in humans with good precision and accuracy.

Three‐dimensional mapping of the joint space for the diagnosis of knee osteoarthritis based on high resolution computed tomography: Comparison with radiographic, outerbridge, and meniscal classifications

“…25 The meniscal classification has been previously described. 26 Briefly, it is as follows: Gr1, normal intact menisci attached with sharp inner borders; Gr2, fraying at inner borders, surface fibrillation, and no tears; Gr 3, partial substance tears, fraying, and fibrillations; Gr4, complete substance tear and loss of tissue. Three bone cores (3 cm height and 7 mm diameter) were harvested with a circular diamond saw (BROT 1 , Argenteuil, France) in the medial tibial plateau in three different positions: (i) peripheral cores located midway between the anterior and posterior edges of the tibial plateau in an external position totally covered by meniscus, (ii) medial anterior cores located in an central position compared with the lateral core never covered by meniscus, and (iii) medial posterior cores located in an internal position compared with the lateral core and in a more posterior position (POST) and partially covered.…”