Results

1. Bearing Surface Damage Analysis of Coupled Total Shoulder Replacement Retrievals L. Maltio, BS1, F. Ansari, MS1, A. Mehdizadeh, BS1, J. Koller, BS1 S. Gunther, MD2, T. Norris, MD3, M. Ries, MD4, L. Pruitt, PhD1 1University of California, Berkeley, Berkeley, CA, 2 Martha Jefferson Hospital, Charlottesville, VA, 3 San Francisco Shoulder, Elbow & Hand Clinic, San Francisco, CA, 4University of California, San Francisco, San Francisco, CA Motivation Methods Six TSR bearing couples (n=5: Depuy, Warsaw, IN) were evaluated (Figure 1). Average in vivo duration was 66 ± 46 months. A grid placed on glenoid components divided the bearing surface into 17 regions of equal area. Each region was photographed at 249 dpi (16-20x) using a Nikon 1 J1 camera and Infinivar Video Microscope lens under an AmScope LED ring light. Each implant was scored by at least 2 observers. Inter-observer agreement was evaluated by calculating the average percent agreement between two observers for damage mode identification and the average percent agreement with the mean for geometric extent and severity classification. In vivo surface damage of total joint replacements (TJRs) can have clinical consequences: release of ultrahigh molecular weight polyethylene (UHMWPE) wear debris can initiate osteolysis and implant loosening [1]. Damage to metallic counterbearing surfaces can significantly increase UHMWPE wear [2]. Most retrievals studies have focused on bearing damage in hips and knee replacements [3,4]. Few studies have focused on total shoulder replacements (TSR), and these are limited to evaluating UHMWPE damage modalities [5]. This work presents a methodology that classifies the mode, geometric extent and severity of surface damage in TSR retrievals. Furthermore, trends in UHMWPE glenoid damage are correlated with damage on mated cobalt chrome (CoCr) humeral head components [6] and clinical factors. Observers note, presence, severity, and extent of 13 damage modes for each region in ImageJ(Figure 2). Damage index (DI) calculated by summing scores of each damage mode (severity x area coverage) across all 17 regions then dividing by 1683. Humeral head damage out of 3000 [6]. Glenoid damage correlated with matching humeral head damage through non-parametric Spearman’s rank correlation coefficient (ρ) in Matlab. Spearman’s coefficient (ρ) used to reveal trends between clinical data and damage mode occurrence for glenoids and humeral heads. Results Figure 3 | Summary of coverage data for each damage mode. Burnishing, wear through and subsurface crackingwere analyzed for trends but were not listed in the figure. All implants showed at least one instance of damage (Figure 3). All six glenoids demonstrated instances of scratching, striations, pits and surface deformation in at least 75% of the regions. On average, striations and pitting had a slight tendency to occur in the central 9 regions; a single instance of embedded debris was also present in the center. Abrasion, delamination, fracture and surface deformation were preferentially located in the outer eight regions of the glenoids. (Table 1). The minimum average agreement between observers on damage modes present for a given region was 84%. Table 1 | Humeral head and glenoid damage mode spearman correlations. Figure 2 | Use of ImageJ to score surface deformation by comparing the damaged area to the total area of the glenoid section. In addition examples of damage are included below. • Conclusions • Damage modes and frequency on UHMWPE glenoids are similar to those observed in hip and knee components with longer implantation times [3-7]. • To our knowledge, this is one of the few studies that directly compares damage modes on CoCr humeral heads with mated UHMWPE glenoid devices. • Statistical correlations demonstrate that certain trends may be able to reveal mechanisms of damage accumulation. For example, a positive correlation between CoCr pitting and UHMWPE mild scratching may indicate that the former damage mode may be a primary culprit for generating scratching damage on glenoid surfaces. • Trends in glenoid damage location are also revealing of damage mechanisms: Delamination and fracture occurred in the outer eight gridded regions of the glenoid, consistent with higher contact stresses that may be due to edge loading. • Damage in TSR is not only distributed over a smaller area, but also accumulates at a faster rate than hip and knee arthroplasties. This greater severity in UHMWPE damage suggests that increased metal bearing damage plays an important role in TSR wear in vivo. SCRATCHING STRIATIONS PITTING Figure 1 | Photographs of a subset of retrievals evaluated, including (a-c) humeral heads and (d-f) glenoids with average in vivo duration of 66 ± 46 months. SURFACE DEFORMATION ABRASION CRACKING Other damage modes include: delamination, embedded debris, rim Erosion, Adhesion, wear through, subsurface cracking, and burnishing. References: [1]WillertHG. J Bio Mat Res. 1977; 11:157-46; [2] SaikkoV. J. Biomed. Mater. Res. 2005; 57: 506-12; [3] Hood RW. J Bio Mat Res. 1983; 17:829-42; [4]McKellopHA. Biomat. 2007; 28: 5049-57; [5] Gunther, SB. J Arthro. 2002; 17: 95-100.; [6] Ansari F. 58th ORS Ann. Mtg. Feb 4-7, 2012; Poster # 1209; [7] Harman M. J Mat Sci: Mat in Med. 2011; 22: 1132-46 April 10-13, 2013 Boston, MassachusettsPoster #657 Acknowledgements: We would like to thank the many graduate and undergraduate students who have assisted with this project, including Kathryn Fink, Iris Kan, AJ Almaguer, Jennifer Chang, Andrew Koutsoftas, Amir Mehdizadeh, Nicole Schauser, Matthew Kury, Rachel Cheng, and Robin Parrish.