Aims. The aim of this study was to longitudinally compare the clinical and radiological outcomes of anatomical total shoulder arthroplasty (aTSA) up to long-term follow-up, when using cemented keel, cemented peg, and hybrid cage peg glenoid components and the same humeral system. Methods. We retrospectively analyzed a multicentre, international clinical database of a single platform shoulder system to compare the short-, mid-, and long-term clinical outcomes associated with three designs of aTSA glenoid components: 294 cemented keel, 527 cemented peg, and 981 hybrid cage glenoids. Outcomes were evaluated at 4,746 postoperative timepoints for 1,802 primary aTSA, with a mean follow-up of 65 months (24 to 217). Results. Relative to their preoperative condition, each glenoid cohort had significant improvements in clinical outcomes from two years to ten years after surgery. Patients with cage glenoids had significantly better clinical outcomes, with higher patient-reported outcome scores and significantly increased active range of motion, compared with those with keel and peg glenoids. Those with cage glenoids also had significantly fewer complications (keel: 13.3%, peg: 13.1%, cage: 7.4%), revisions (keel: 7.1%, peg 9.7%, cage 3.5%), and aseptic glenoid loosening and failure (keel: 4.7%, peg: 5.8%, cage: 2.5%). Regarding radiological outcomes, 70 patients (11.2%) with cage glenoids had glenoid radiolucent lines (RLLs). The cage glenoid RLL rate was 3.3-times (p < 0.001) less than those with keel glenoids (37.3%) and 4.6-times (p < 0.001) less than those with peg glenoids (51.2%). Conclusion. These findings show that good long-term clinical and radiological outcomes can be achieved with each of the three aTSA designs of glenoid component analyzed in this study. However, there were some differences in clinical and radiological outcomes: generally, cage glenoids performed best, followed by cemented keel glenoids, and finally cemented peg glenoids. Cite this article: Bone Joint J 2023;105-B(6):668–678

Introduction. The clinical impact of radiolucent glenoid lines is controversial, where the presence of a radiolucent glenoid lines has been suggested to be an indicator of clinical glenoid loosening. The goal of this database analysis is to quantify and compare the pre- and post-operative outcomes of 427 patients who received a primary aTSA with one specific prosthesis and were sorted based upon the radiographic presence of a radiolucent glenoid line at latest clinical followup. Methods. 427 patients (mean age: 67.0yrs) with an average follow-up of 49.4 months was treated with aTSA for OA by 14 fellowship trained orthopaedic surgeons. Of these 427 patients, 293 had a cemented keel glenoids (avg follow-up = 50.8 months) and 134 had a cemented pegged glenoids (avg follow-up = 48.7 months). Cemented peg and keel glenoid patients were analyzed separately and also combined into 1 cohort: 288 patients (158 female, avg: 68.7 yrs; 130 male, avg: 64.9 yrs) did not have a radiolucent glenoid line (avg follow-up = 46.9 months); whereas, 139 patients (83 female, avg: 68.5 yrs; 56 male, avg: 64.6 yrs) had a radiolucent glenoid line (avg follow-up = 54.4 months). Outcomes were scored using SST, UCLA, ASES, Constant, and SPADI metrics; active ROM also measured. A two-tailed, unpaired t-test identified differences (p<0.05) in pre-operative, post-operative, and pre-to-post improvements. Results. The overall rate of radiolucent line formation in the glenoid for the combined peg and keel glenoid aTSA cohort was 32.6% (139 of 427 patients). The rate of radiolucent lines for the keel and peg glenoids was similar at nearly the same followup (∼50 months): 96 of 293 peg patients (32.8%) and 43 of 134 keel patients (32.1%) had a radiolucent glenoid line. Patients with cemented keeled glenoids had an average radiographic line score of 2.4 ± 1.3 which trended larger (p = 0.0875) than the average radiographic line score of 2.0 ± 1.1 of patients with cemented peg glenoids. Pre-operatively, no difference was noted between patients with and without radiolucent glenoid lines. (Figure 1) However post-operatively, patients with radiolucent glenoid lines were associated with significantly poorer clinical outcome scores according to all 5 clinical metrics and also had significantly less improvements in outcomes according to 4 of 5 outcome metrics; (Figure 2) poorer outcomes were observed for the combined cemented keeled and pegged glenoid cohorts and when each glenoid prosthesis type is analyzed separately. Additionally, patients with glenoid radiolucent lines were associated with significantly lower improvements in outcomes. (Figure 3) Finally, 24 complications were reported (5.6%), 14 for patients without radiolucent glenoid lines (4.9%) and 10 for patients with radiolucent glenoid lines (7.2%). Conclusions. This large-scale clinical outcome study demonstrated that aTSA patients with radiolucent glenoid lines were associated with significantly poorer clinical outcomes and a higher complication rate as compared to aTSA patients without radiolucent glenoid lines. The overall radiolucent line rate and the negative impact of radiolucent lines were no different between cemented keeled and cemented pegged glenoid components. Additional and longer-term follow-up is needed to confirm these conclusions. For figures/tables, please contact authors directly.