Factors contributing to graft matching in a patellar osteochondral allograft selection model. Issue 7 (31st July 2020)

• Record Type: Journal Article
• Authors: Huddleston, Hailey; Yanke, Adam; Inoue, Nozomu
• Is Part Of: Orthopaedic journal of sports medicine. Volume 8:Issue 7(2020)Supplement 6
• Publication Date: 2020-07-31
• Subjects: Sports medicine -- Periodicals; Orthopedics -- Periodicals; Arthroscopy -- Periodicals; Arthroplasty -- Periodicals; Knee -- Surgery -- Periodicals; 616.7
• DOI: 10.1177/2325967120S00455
• Languages: English
• ISSNs: 2325-9671

Abstract

Objectives: When performing a patellar osteochondral allograft, the patellar allograft is harvested from a similar anatomic location as the defect. This approach assumes that graft will have similar topography to the patellar defect. However, to our knowledge, no prior study has investigated the topography of the patella and what intrinsic factors of the graft and the recipient affect mismatch of the chondral and osseous layers between the graft and defect. Methods: Three-dimensional (3D) computed tomography (CT) models of the patella were created and exported into point-cloud models using a 3D reconstruction program (Mimics, Materialise Inc., Leuven, Belgium). Circular articular cartilage and subchondral bone defect models were created in each model of the recipient patella (diameter=18mm) at 3 locations: medial, distal, and lateral. Articular cartilage and subchondral bone graft models were created on all possible locations on the articular cartilage surface models of the donor patellae. 3D surface topographies of the articular cartilage surface and resulting subchondral bone surfaces were compared between graft and defect models. The graft models were virtually placed on the surface of the defect model. Least distances, defined as the shortest distance from the point in question to the corresponding point in space, where a perfect congruent match would equal a least distance of 0.00mm for given data points on the simulated articular cartilage surface, were calculated. A mean value of the least distances was calculated for each position of the graft model and for the subchondral bone surface, simultaneously. The graft model was then rotated 360° around the axis perpendicular to the articular cartilage surface in 1° increments, and the least distance of articular cartilage surface and least distance of subchondral bone surface were calculated at each rotating angle. This procedure was repeated for all points in the articular surface model of the donor patella. The 3D model creation and geometry matching were performed using a custom-written program coded by in Microsoft Visual C++ with Microsoft Foundation Class programming environment (Microsoft Corp., Redmond, WA). Multivariate linear regression analysis was conducted in SPSS (v26, IBM, Armonk, NY). Results: Chondral and osseous mismatch between the graft and defect were analyzed. ANOVA analysis on the multivariate linear regressions found significant predictors of cartilage mismatch for medial (p=0.002), lateral (p=0.022), and central (p=0.001) defects when testing 5 variables. However, no predicting variables were identified for osseous mismatch for medial (p=0.099), lateral (p=0.703), and central (p=0.641) defects. Differences in tibia width (p=0.005), bone width (p=0.004), and medial cartilage length (p=0.003) were predictive of mismatch in medial defects. When evaluating lateral defects, no variables were found to significantly effect mismatch, However, in this lateral defect group, the collinearity assumption of the regression was violated, as the VIF for bone width and lateral length were over 10. For the central group, difference in bone width (p=0.037), difference in percent of patella that was medial facet (p=0.001), and difference in tibial width (p=0.006) were predictive of mismatch. Conclusions: Differences between graft and recipient tibia width, bone width, and size of the medial or lateral facet are significant predictors of mismatch in patella allograft selection.