Defining deformity thresholds for the use of augmented glenoids in arthritic shoulders with posterior glenoid bone loss

Abstract

BackgroundDespite increasing utilization, little guidance is offered regarding appropriate indications for posterior augmented glenoid implants. The purpose of this study is to virtually assess the effect of implant selection on glenoid resurfacing and joint line restoration in osteoarthritic shoulders with posterior glenoid wear. MethodsThirty-three CT scans were randomly selected from a cohort of osteoarthritic shoulders with Walch B2 or B3 deformities. Imascap SAS (Wright Medical, Memphis, TN) was used to assess bony deformities and plan glenoid resurfacing. Implants simulated included: standard Pegged Cortiloc, and 15° and 25° half-wedge augments. Each component was planned for a version correction to neutral and 10° retroversion, inclination was maintained within 5° of neutral and the component was medialized until bony support reached 80%. Implantation failure was defined as: greater than 20% cancellous bone support, central peg perforation, or joint line medialization relative to the pathologic joint (accounting for polyethylene thickness). Excessive lateralization was defined as creation of a joint-line >4 mm lateral to the pathologic joint-line. ResultsThe mean retroversion deformity in this cohort was 23.7° (range: 13°-37°). When correcting to neutral, the corrective reaming alone failed in 72.7% (24/33) of cases compared to 15.2% (5/33) when correcting to 10° of implant retroversion. When correcting to neutral version failure was decreased with a 15° (27.3%; 9/33; P < .001) and 25° augment (15.2%; 5/33; P < .001). Receiver operating characteristic (ROC) analysis found that an augment is useful to resolve standard implant failure for retroversion deformities greater than 27° (area under the curve [AUC] = 0.91) and 22° (AUC = 0.77) for correcting version to neutral and 10° retroversion, respectively. When placing a 25° augment, there was high risk of overstuffing the joint for both correction to neutral (27.3%) and 10° retroversion (66.7%). Overstuffing was likely to occur when correcting deformities less than 19° (AUC = 0.96) to neutral with a 25° augment. Likewise, implanting 15° or 25° augmented glenoids in 10° of retroversion is likely to overstuff the joint for pathologic retroversion deformities smaller than 18° (AUC = 0.94) and 28° (AUC = 0.69), respectively. ConclusionThis study helps to clarify the severity of glenoid deformities more appropriately addressed with the use of a half-wedged augmented glenoid. Both the severity of glenoid deformity and the desired implant version affect appropriate indications for the use of augmented glenoids. The risk of excessive joint-line lateralization in shoulders with less severe deformities should not be ignored. These considerations, in addition to individual soft tissue tension should be considered when planning and performing glenoid resurfacing. Level of evidenceLevel III; Treatment Study