Abstract
Introduction:
Given that factors like center of rotation (COR), neck shaft angle, glenosphere diameter and component tilt alter the biomechanics of reverse total shoulder arthroplasty (rTSA), the performance of the total rTSA system is of interest. This study compared the composite performance of two rTSA systems that were designed around a medialized or lateralized glenohumeral COR. The objective was to quantify the following outcome measures: 1) COR & humeral position; 2) range of glenohumeral abduction; 3) force to abduct; and 4) range of internal (IR)/external (ER) rotation.
Methods:
Seven pairs of shoulders were tested with a biomechanical shoulder simulator. Beads were implanted in the scapula and humerus to quantify bone positions with a fluoroscope. Spectra lines simulated the deltoid and the rotator cuff. Linear actuators simulated muscle excursion while load cells recorded applied force. Diode arrays were used to quantify arm position and calculate the humeral center of rotation. Native specimens were tested where a motion path was recorded from resting to peak glenohumeral abduction in the scapular plane. The trajectory was replayed and deltoid force vs. arm position was recorded. With the elbow flexed, the arm was articulated to maximal internal and external rotation to determine ROM limits due to impingement or soft tissue constraint. Specimens were implanted with a Tornier Aequalis Reversed Shoulder prosthesis (“A,” 36 mm glenosphere, 10° humeral retroversion, 9 mm poly insert – “medial”) or a DJO Surgical Reverse Shoulder Prosthesis (“R,” 32 mm, 30° retroversion, neutral insert/shell – “lateral”). Implants were randomized between shoulders in a pair. After implantation the test protocol was repeated. Paired-t tests (p ≤ 0.050) were adjusted with Holm's step-down correction for multiple comparisons.
Results:
Joint COR shifted inferiorly (A = 7 ± 3 mm, R = 4 ± 2 mm) and medially (A = 19 ± 4 mm, R = 12 ± 3 mm) for both systems with respect to native (p≤0.007, between systems p≤0.037). All humeri shifted inferiorly with respect to native (Fig. 1, p = 0.000, between systems p = 0.718). The RSP maintained a nearly anatomic medial/lateral humerus position, whereas the Aequalis medialized the humerus (p = 0.007). Both rTSA systems showed adduction deficit versus native arms (Fig. 2, p ≤ 0.046). Peak passive abduction, IR and ER were not significantly different between systems (p ≥ 0.113) or with respect to native (p ≥ 0.085). Deltoid force required to elevate the arm decreased ∼25% after rTSA (p ≤ 0.049), but did not differ between systems (p ≥ 0.117).
Discussion:
Understanding the implications of implant configuration is imperative to improving implant design and optimizing patient outcomes. As tested, the configurations represent over 70% of respective clinical cases. The systems varied in COR offset, humeral component version/tilt, glenosphere placement, and insert thickness, yet few kinematic differences arose. The RSP COR was more lateral than the Aequalis, yet both were medial to native. Accordingly, both systems provided a similar mechanical advantage by reducing the abduction forces. The RSP had the least adduction deficit, which could indicate increased inferior clearance around the more lateral COR. Inferior and medial humerus shift could negatively impact external rotation capability by moving the posterior cuff line of action below the COR and reducing muscle tension (Fig. 3).
