Many glenohumeral studies have demonstrated how anatomy varies across the population. Third and fourth generation shoulder prosthesis designs were developed to provide greater modularity and size ranges to better reproduce this anatomy and thus achieve better results in terms of shoulder function. This study quantifies the quality of anatomic reconstruction and compares that to long-term clinical outcomes using one fourth generation platform shoulder system.
Methodology: One hundred and forty primary total shoulder arthroplasties were performed by one experienced single surgeon between 2001 and 2009, using the same fourth generation modular prosthesis. Pre- and postoperative clinical assessments were quantified with the Constant, ASES, SPADI, SST, and UCLA scores, and active range of motion was measured. Five anatomic parameters were defined, measured, and compared pre- and postoperatively on the anterior-posterior (AP) radiographs: Humeral Head Height (HHH), Humeral Head Centering (HHC), Humeral Head Medial Offset (HHMO), Humeral Head Diameter (HHD), and Humeral Neck Angle (HNA). The differences between each of the parameters were then calculated and rated from 0 to 2 and then summed for each patient to obtain the Anatomic Reconstruction Index (ARI), which objectively quantifies and assesses the quality of the anatomic reconstruction. Patients were sorted based upon their ARI score into two groups (ARI 5 to 7 and ARI 8 to 10), and their latest follow-up outcomes were compared using the Mann-Whitney test to identify differences in preoperative and postoperative results, where p < 0.05 denoted a significant difference.
Results: Of the 140 primary prostheses performed, 78 patients were lost to follow-up, and 13 were excluded for complications that were not related to the anatomic reconstruction. Forty-nine patients (75.8 yrs., 31F/18M) were included with an average follow-up of 9.1 years. The average score for HHH was 1.9 ± 0.4, 1.8 ± 0.5 for HHC, 1.7 ± 0.5 for HHMO, 1.7 ± 0.5 for HHD, and 1.5 ± 0.7 for HNA. Thus, all reconstructions were rated good to excellent with 86% of very good/excellent reconstruction (ARI 8 to 10) and 14% good reconstruction (ARI 5 to 7). A comparison of radiographic anatomic parameters was performed for these two cohorts: HHC (< 0.0001), HNA (0.000), and ARI (<0.0001) were significantly greater for the ARI 8 to 10 cohort. Four of five postoperative clinical outcome metrics for the ARI 8 to 10 cohort were significantly greater than the mean values for the ARI 5 to 7 cohort. Additionally, mean postoperative pain on a daily basis and shoulder function for the ARI 8 to 10 cohort were significantly greater than that for the ARI 5 to 7 cohort.
Discussion: The relatively small number of good reconstructions (14%) compared to very good/excellent reconstructions (86%) and the absence of fair/poor reconstructions limited the ability for any strong linear correlations between anatomical reconstruction and clinical parameters. Despite this, patients with larger mean ARI scores were associated with significantly better outcomes for some measures. This study is limited by the use of 2D assessments from standard AP radiographs; this method can be further refined by the use of 3D quantitative assessment of each parameter.
Conclusion: This study confirmed that an improved anatomic reconstruction results in better postoperative clinica outcomes. The fourth generation prosthesis used for this study allows continuous in-situ adjustment of the humeral head orientation through use of the spherical taper on the replicator plate and also a double adjustment of offset through the use of the offset humeral head and offset taper on the replicator plate.