A Biomechanical Study of Double-row Suture-bridge and Double-screw Fixation Strength in Split-type Greater Tuberosity Fractures of the Humerus

Background: This study aimed to compare the xation strength of double-row suture-bridge (DR) and double-screw (DS) xation methods in treating split-type greater tuberosity fracture of the humerus. Methods: We established 24 fracture models of frozen pig shoulders with reserved supraspinatus, infraspinatus, and teres minor muscles. The specimens were randomly divided into two groups according to the xation methods of double-row suture-bridge (group A) and double screws (group B). The loads of 3-mm displacement, 5-mm displacement, and xation failure of each group were measured using a biomechanical device in both traction directions of the supraspinatus muscle (horizontal direction) and infraspinatus-teres minor muscles (vertical backward direction). Results: In the direction of supraspinatus traction, there was no statistical difference in the mean load values of 3-mm displacement, 5-mm displacement, and xation failure in both groups (p > 0.05). In the direction of infraspinatus-teres minor traction, the mean load values of 3-mm displacement, 5-mm displacement, and xation failure of double-screw xation were greater than those of double-row suture-bridge xation (p = 0.000). Conclusion: The xation strength of double-screw suture-bridge and double screws were equivalent in the direction of the supraspinatus. However, the xation strength of double-screw suture-bridge was less than that of double screws, indicating poor anti-rotation property. Double-screw suture-bridge xation alone is not recommended to treat split-type greater tuberosity fracture with a large fragment. of the greater tuberosity to distal humerus 15 cm, horizontal osteotomy, end in a × × 5-cm of denture

These fractures primarily occur in middle-aged men and are often accompanied by shoulder dislocation or rotator cuff injury [1][2][3]. In recent years, some scholars have proposed that operation should be performed in patients with fragment displacement over 5 mm, and in cases with fragment displacement over 3 mm, surgical treatment is essential to heavy workers and athletes with high requirements for a large amount of excise or frequent overhead activities [4,5]. Mutch et al. [6] proposed a classi cation system for isolated greater tuberosity fracture based on its characteristics, including avulsion type, compression type, and split type, and they found that the split-type fractures were more common. Several xation methods are available for split-type greater tuberosity fracture with pros and cons of each own, but they still remain highly controversial without any gold standard. In terms of mechanics, the supraspinatus, infraspinatus, and teres minor tendons are attached to the humeral greater tuberosity, pulling the fragment toward posterosuperior direction [7][8][9]. This type of fracture requires anatomic reduction and rigid xation to avoid further redisplacement. The current foreign and domestic studies mainly focus on the xation strength in the traction direction of the supraspinatus, and insu cient attention was paid to that of the infraspinatus and teres minor. Therefore, we designed this experiment to verify the biomechanical strength of two commonly used xation methods and compared their xation strength in both traction directions of the supraspinatus and infraspinatus-teres minor muscles, aiming to provide a biomechanical basis to treat isolated split-type greater tuberosity fractures of the humerus.

Materials And Methods
The porcine shoulder specimens used were purchased on the market, and no ethical review was required after consultation with the Medical Ethics Committee of Peking University People's Hospital.

Specimen preparation
We used 12 pairs of fresh frozen 6-month-old female porcine shoulders (24 in total) for this study. The bone mineral density of a 6-month-old female pig is similar to that of normal young adults. The porcine shoulder is widely used in mechanical experimental research, and the results have been widely recognized [10][11][12]. The skin, subcutaneous soft tissue, super cial fascia, and deep fascia were removed from all shoulder specimens, and 4-cm muscle-tendon stumps of the supraspinatus, infraspinatus, and teres minor at the proximal end of the greater tuberosity were preserved. The greater tuberosity was cut using a swing saw along the marked outline to establish the fracture model. The length from the vertex of the greater tuberosity to the distal humerus was 15 cm, and with horizontal osteotomy, the distal end of the humerus was embedded in a 10-× 5-× 5-cm container made of denture powder. The specimens were randomly divided into two groups with 12 specimens in each, and the corresponding methods were used for fracture xations. A single 2-mm-diameter wire rope was used for continuous hemstick suture to ensure neat edges and rm stress.
(1) Double-row suture-bridge (group A): Two medial-row suture anchors (5.0 mm, Helix, Johnson & Johnson) were implanted in the proximal part of the fragment with an angulation of 45°, and the distance between the screws was 1.5 cm. The sutures (No.2 FiberWire) were knotted after passing through the rotator cuff tendon attached to the greater tuberosity. Two lateral-row suture anchors (Versalok, Johnson & Johnson) were placed aligned with the medial-row suture anchors and 5 mm distal to the fragment. The sutures were passed through the Versalok anchors to form a mesh and then tightened to x the fragment. (Fig. 1, a). Double-screw (group B): Two 6.5-mm cancellous bone screws with washers were implanted in the central area of the fragment without penetrating the humeral head ( Fig. 1, b).

Biomechanical testing
The specimens were positioned in the electron-tensile tester (AG-IS MO; Shimadzu Corp., Tokyo, Japan). The pulling directions were set the same as the traction directions of the supraspinatus muscle and infraspinatus-teres minor muscles. The 0° position of shoulder joint was selected as the test position to more effectively simulate shoulder abduction.
Before the experiment, we applied 0-50 N at a speed of 5 mm/min to each specimen three times to eliminate soft tissue adhesion and improve the reliability and repeatability of data. After preconditioning, all specimens were loaded at a constant speed of 10 mm/min. The tensile forces of two xation methods in two traction directions were measured when the greater tuberosity fragment displaced 3 mm, 5 mm, and loaded to failure (internal xation failure, fracture or displacement of over 1 cm), respectively. The load-displacement curves of each group were recorded (Figs. 2-5).

Statistical analysis
The data were analyzed by SPSS 20.0 (IBM SPSS, Inc., Chicago, IL, USA). The mean between the two groups was compared by an Independent-Sample T-test. P < 0.05 was set as the difference level, and P < 0.001 was set as the signi cant difference level.

Comparison of two groups in the vertical backward traction direction
In the direction of supraspinatus traction, there were no signi cant differences in the xation strength between the two groups (P = 0.802, P = 0.119, and P = 0.914). All the samples in group A had displacements of over 1 cm, and in group B, 4 samples had humeral head fracture on the screw level, and 2 samples had displacements of over 1 cm (Table 1). Comparison of two groups in the horizontal traction direction In the direction of infraspinatus-teres minor traction, the loading force in group B was signi cantly greater than that in group A (P = 0.000, P = 0.000, and P = 0.000). All the samples in group A had displacements of over 1 cm, and in group B, 5 samples had displacements of over 1 cm, and 1 sample had greater tuberosity fracture on the level of the proximal screw (Table 2).

Comparison of load values in each group in two traction directions
In group A, the loading force in the supraspinatus traction direction was signi cantly greater than that in the infraspinatus-teres minor traction direction (P = 0.000, P = 0.000, and P = 0.000). In group B, the loading force in the infraspinatus-teres minor traction direction was greater than that in the supraspinatus traction direction (P = 0.005). However, there were no signi cant differences in the load to 5-mm displacements or load to failure (P = 0.215 and P = 0.183).
In group A, the loading force in the supraspinatus traction direction was signi cantly greater than that in the infraspinatus-teres minor traction direction (P = 0.000, P = 0.000, and P = 0.000). In group B, the loading force in the infraspinatus-teres minor traction direction was greater than that in the supraspinatus traction direction (P = 0.005). However, there were no signi cant differences in the load to 5-mm displacements or load to failure (P = 0.215 and P = 0.183) ( Table 3).

Discussion
We designed this biomechanical experiment based on the anatomical characteristics of the greater tuberosity and rotator cuff as well as the biomechanical characteristics of fragment displacement to compare the xation strength of two commonly used internal xation methods to treat split-type greater tuberosity fracture in two traction directions of the supraspinatus and infraspinatus-teres minor muscles. The supraspinatus plays a major role within the range of shoulder abduction less than 15°, while the infraspinatus and teres minor muscles are mainly involved in the external rotation of the shoulder joint [13]. The displacement of the greater tuberosity fragment results in the instability of the rotator cuff and dysfunction of the shoulder joint. Therefore, we should pay attention to the xation strength in both directions of shoulder abduction and external rotation. In recent years, researches on this type of fracture have focused primarily on xation strength in the direction of supraspinatus traction. A study by Seppel et al. [14] utilizing 12 cadaver bones showed no statistical difference in the xation strength of single-row and double-row anchors in the treatment of split-type greater tuberosity fracture. Lin et al. [15] compared the xation methods of double-row suture anchor, double-row suture-bridge, and double-screw xation in the direction of supraspinatus traction. When loaded to the 3-mm displacement, the xation strength of double-row suture-bridge was stronger than that of other two xation methods. When loaded to 5-mm displacement and failure, there was no statistical difference between double-row suture-bridge and double-row suture anchor, which were stronger than that of double-screw xation. However, no studies have been reported on the xation strength in the direction of infraspinatus-teres minor traction. The external rotation induced by infraspinatus and teres minor has not been paid much attention in clinic due to long-term immobilization and early limitations of external rotation of the shoulder joint after operation. To avoid shoulder joint stiffness resulting from long-term immobilization, early rehabilitation is essential but relies on the rigid xation in both directions.
The double-row suture-bridge has been widely used in clinic, which can effectively resist the traction of the supraspinatus and prevent the postoperative redisplacement of the greater tuberosity fragment [16,17]. However, this technique is more suitable for avulsion fractures of rotator cuff tendon attachment as well as rotator cuff tear [18]. For split-type greater tuberosity fracture with a large fragment, the suture anchor xation is an elastic method, which may have shortcomings such as poor strength, anchor pulling out, rotator cuff cutting, etc., and its e cacy is controversial [19,20]. In our study, the xation strength of double-row suture-bridge technique has poor xation strength to resist rotation in the direction of infraspinatus-teres minor traction. Therefore, it is not suitable for this type of fracture.
The double-screw technique has a strong advantage in the xation of cancellous bone fracture. With cancellous lag screws, the fracture was rigidly xed with compression. In this experiment, we found that the xation strength of double screws was equal to that of double-row suture anchor in the supraspinatus traction direction, but greater than that of double-row suture anchor in the infraspinatus-teres minor traction direction. Moreover, double-screw xation can be completed under arthroscopy with minimal incisions, or with a percutaneous method to reduce further damage to soft tissue and blood supply [21,22,23]. In recent years, absorbable screws have been introduced, which can reduce the burden of secondary surgery, avoid metal reaction, and reduce the phenomenon of stress shielding [24].
However, double-screw xation also has defects, and the xation strength is less than that of locking plate or tension band, and under continuous supraspinatus traction, the fracture fragment is easily displaced, making the internal xation ineffective [25,26]. Double-screw xation with washers increases the risk of acromial impingement. Furthermore, for patients with osteoporosis or comminuted fragments, it is impossible to obtain instability [27].
Clinically, various mechanical factors affect the greater tuberosity of the humerus. The major limitation of this experiment is that the in vitro biomechanical experiments can only simulate the instantaneous xation strength but not dynamic mechanical changes in the process of shoulder movement, which requires further investigation.

Conclusion
The xation strength of both double-row suture-bridge and double screws are equivalent in the direction of supraspinatus traction. However, double-screw xation provides greater xation strength compared with that of double-row suture-bridge in the direction of infraspinatus-teres minor traction. The double-row suture-bridge xation method has poor anti-rotation property, which requires long-term postoperative immobilization to prevent fragment displacement. Thus, DR xation alone is not recommended for isolated split-type greater tuberosity fracture of the humerus. In contrast, double-screw xation is more suitable for this type of fracture with advantages in mechanical capacity. Figure 1 Two xation methods: a: double row suture-bridge; b: two screws