Specimen preparation and deforming force measurement.
In this study, 12 fresh-frozen cadaveric shoulders were used; the age of the cadavers ranged from 65 to 88 years (mean age, 79 ± 8 years). There were four male and eight female cadavers. None of the specimens had a history of trauma or surgery and appeared normal upon visual inspection. They were frozen and stored at -20℃ and were defrosted overnight at room temperature prior to use. The skin, soft tissues, and muscles around the shoulder (scapula, clavicle, and humerus) were dissected except the rotator cuff muscles and the pectoralis major muscle, carefully preserving the capsular structures without any damage. The humerus and scapula remained intact. After completing specimen preparation, the scapula was fixed to a custom jig (Fig. 2) using a scapular clamp with a 20° anterior tilt in the sagittal plane[14–16].
In the SN model (PHFs involving the SN), the clavicular and sternocostal heads of the pectoralis major muscle may act as the deforming force responsible for displacement by pulling the shaft anterior and medially[6, 17–19]. In the GT model (PHFs involving GT), the supraspinatus and infraspinatus muscles may act as the deforming force responsible for displacement by pulling the GT either superiorly or posteriorly[6, 17–19]. Our design measured the tension applied on the fracture components and how it changed depending on arm position in order to investigate the optimal arm position for minimizing deforming force on the fracture components.
The direction of the suture (FiberWire; Arthrex, Naples, FL, USA) was determined along the anatomic force vectors of the representative muscles or tendons[16, 20], taking into account their origin and insertion (Fig. 2). In the SN model, considering that the pectoralis major muscle has both clavicular and sternocostal heads, it was replaced by two sutures. In the GT model, the supraspinatus and infraspinatus muscles were each replaced by a single suture. Based on the footprint of each muscle or tendon on the humerus, sutures replacing each head of the pectoralis major muscle were also transosseously secured by creating holes in the center of the lateral lip of the bicipital groove. The suture replacing either the supraspinatus or infraspinatus muscle was secured transosseously through holes in the center of the superior and middle facets of the GT, respectively. Each suture was attached to the end of a spring (Sciencelove, Goyang, Korea) with a constant elastic modulus(k) of 0.417 N/mm that was connected to a digital force gauge (AMF-30; Aliyiqi, Zhejiang, China), allowing free movement and fixation in the sagittal plane. Then, the starting point was set after pre-tensioning to 0.5 N[21].
For each fracture model, before the changes in tension were measured by increasing the deforming force according to arm position, the reference zero-point for each fracture was determined, considering the various types of shoulder immobilization braces available on the market. For the reference point of the SN model where the pectoralis major may cause medial displacement of the humerus, full adduction (0° of abduction) and full (90°) internal rotation of the arm were determined where the tension loaded by the two heads of the pectoralis major muscle was minimized. Then, tension was measured in full adduction by gradually increasing external rotation from 90° of internal rotation to 0° (neutral rotation) in decrements of 30°. For the GT model, although both the supraspinatus and infraspinatus muscles insert through the GT, tension loaded on each tendon was measured separately because the directions of the vectors were different. As the reference point for where the tension is loaded by the supraspinatus (abductor inserting on the GT), 45° of abduction along the scapular plane[14, 16] and internal rotation of 30° from the coronal plane[16, 22] was used. Tension was measured during adduction in decrements of 15°. Finally, as the reference point for where the tension is loaded by the infraspinatus (external rotator insertion on the GT), full adduction and neutral rotation (0° of rotation) of the arm were determined. Then, tension was measured while gradually increasing internal rotation from 0° (neutral rotation) to 90° of internal rotation in full adduction in decrements of 30°.
Statistical analysis including sample size calculation.
Since there has been no previous studies addressing a similar topic, the study design was determined after consulting medical statisticians. Then, a pilot study was conducted using three cadaveric specimens. The sample size was calculated using the difference in deforming force caused by the infraspinatus muscle when moving from 0° (neutral rotation) to 60° and from 0° (neutral rotation) to 90° of internal rotation. The mean ± standard deviation value for differences between each group in the pilot study was 1.3 ± 1.5 N. Based on these data, 12 specimens were needed to present 80% power at an α level of 0.05.
The Friedman test was used to identify significant differences among at least one group (an increase in deforming force due to arm movement) out of three for each muscle. The Wilcoxon signed-rank test was used for post hoc analysis for significant differences from the Friedman test with a Bonferroni correction. The Jonckheere-Terpstra test was used to determine either positive or negative trends in tension caused by each muscle across three groups of arm movement. The level of statistical significance for all tests was set at P < 0.05. All statistical analyses were conducted using SPSS Statistics for Windows (Version 27.0; IBM Corp., Armonk, NY, USA).
Ethics statement.
All cadavers used in this study were legally donated to the Surgical Anatomy Education Center, Yonsei University College of Medicine. Donors of cadavers approved the cadavers for use in research. The study was authorized by the Institutional Review Board of Yonsei University Health System, Severance Hospital (4-2023-0826). All experiments were performed in accordance with relevant guidelines and regulations.