- Email: [email protected]
Effect of rotator interval closure on glenohumeral stability and motion: A cadaveric study
Effect of rotator interval closure on glenohumeral stability and motion: A cadaveric study Nobuyuki Yamamoto, MD, Eiji Itoi, MD, Yilihamu Tuoheti, MD, Nobutoshi Seki, MD, Hidekazu Abe, MD, Hiroshi Minagawa, MD, Yoichi Shimada, MD, Kyoji Okada, MD Akita, Japan
The effect of rotator interval closure, which is performed as an adjunct to arthroscopic stabilization of the shoulder, has not been clarified. Fourteen freshfrozen cadaveric shoulders were used. The position of the humeral head was measured using an electromagnetic tracking device with the capsule intact, sectioned, and imbricated between the superior glenohumeral ligament and the subscapularis tendon (SGHL/ SSC closure) or between the superior and middle glenohumeral ligaments (SGHL/MGHL closure). The direction of translational loads (10, 20, and 30 N) and arm positions were (1) anterior, posterior, and inferior loads in adduction; (2) anterior load in abduction/external rotation in the scapular plane; and (3) anterior load in abduction/external rotation in the coronal plane. The range of motion was measured using a goniometer under a constant force. Both methods reduced anterior translation in adduction. Only SGHL/MGHL closure reduced anterior translation in abduction/external rotation in the scapular plane and posterior translation in adduction. Both methods reduced the range of external rotation and horizontal abduction. Rotator interval closure is expected to reduce remnant anterior/posterior instability and thereby improve the clinical outcomes of arthroscopic stabilization procedures. (J Shoulder Elbow Surg 2006;15: 750-758.)
P revious studies showed that results of the arthro-
scopic stabilization of the glenohumeral joint were inferior to those of open procedures in terms of recurrence rate.4,11,19,20,25,29,34 Recently, however, the outcome of arthroscopic stabilization has improved dramatically.2,9,30,32 This may be due to advances in From the Division of Orthopedic Surgery, Department of Neuro and Locomotor Science, Akita University School of Medicine. Reprint requests to: Eiji Itoi, MD, Department of Orthopedic Surgery, Tohoku University School of Medicine, 1-1 Seiryo-machi, Aoba-ku, Sendai 980-8574, Japan (E-mail: itoi-eiji@mail. tains.tohoku.ac.jp). Copyright © 2006 by Journal of Shoulder and Elbow Surgery Board of Trustees. 1058-2746/2006/$32.00 doi:10.1016/j.jse.2005.12.009
750
techniques and instrumentation. This may also be due to adjunct procedures performed during stabilization procedures such as suture plication or thermal capsulorrhaphy. Rotator interval (RI) closure is one of these adjunct procedures, which is thought to correct pathoanatomy of shoulder instability, and as a result, to improve the outcome of arthroscopic instability repairs.2,10,15,30 –32 Gartsman et al10 found that repair of the RI was a critical factor in 14 of 53 shoulders treated arthroscopically for anteroinferior glenohumeral instability, and that RI repair contributed to the improved clinical outcomes observed in their study. The recent investigations have supported the concept that even anterior-inferior instability is associated with multiple lesions.9,30 Capsular and ligamentous laxity or elongation are present in various degrees in most patients with shoulder instability, and RI lesions are one of a host of potential lesions.9,15,30,32 The significance of an RI lesion in glenohumeral instability has become increasingly apparent, as this anatomic structure has been shown to contribute to shoulder stability in clinical and biomechanical studies.8,14,22 We have found only one biomechanical study, by Harryman et al12, that investigated quantitatively the effect of operative sectioning and imbrication of the RI. They reported the effect of vertical sectioning and medial-lateral closure of the RI capsule on motion and stability of the shoulder. Their conclusion was that imbrication of the RI capsule increased the resistance to inferior and posterior translations. However, it is not the medial-lateral closure, but the superior-inferior closure of the RI that is commonly performed clinically. To our knowledge, there has been no study to clarify the effect of superior-inferior imbrication of the RI on shoulder biomechanics. We hypothesized that closure of the RI capsule would increase resistance to anterior translation of the humeral head in normal shoulders, as well as in shoulders with anterior instability. The purpose of the current study was to determine the effects of closure of the RI with respect to glenohumeral motion and stability.
J Shoulder Elbow Surg Volume 15, Number 6
Figure 1 Specimen and the electromagnetic sensors. A pair of acrylic plates was fixed to the medial border of the scapula by means of plastic screws (1 arrow). Another pair of acrylic plates was fixed to the scapular body by means of Kirschner wires (2 arrows). One electromagnetic sensor (2 arrowheads) was attached to the humeral shaft and the other (1 arrowhead) was attached to the scapular spine. The source was attached to the positioning device.
MATERIALS AND METHODS Preparation of specimens
We used 14 fresh-frozen cadaveric shoulders from donors who were 25 to 88 years old (mean, 64 years old) at the time of death. The frozen shoulders were thawed overnight at room temperature for experimentation. None of the shoulders had a rotator cuff tear or radiologic evidence of glenohumeral osteoarthritis. Each specimen had been disarticulated at the scapulothoracic joint proximally and dissected at the midpart of the humerus distal to the deltoid attachment distally. The skin, subcutaneous tissue, and all of the muscles were removed except for those of the rotator cuff, which were elevated from the scapula. The lateral 1/3 of the scapular spine, including the acromion and the distal half of the coracoid process, was removed. The coracohumeral ligament was preserved. A pair of acrylic plates (240 mm in length ⫻ 40 mm in width) was fixed to the medial border of the scapula by means of plastic screws, 4 mm in diameter (Figure 1). Another pair of acrylic plates (220 mm in length ⫻ 130 mm in width) was fixed to the scapular body by means of the Kirschner wires (1.8 mm in diameter). The scapula was sandwiched between 2 pair of plates, allowing free motion of the cables in between, which were attached to the rotator cuff tendons. An intramedullary rod (10 mm in diameter)
Yamamoto et al
751
Figure 2 Custom-designed shoulder-positioning device.
was inserted into the proximal humeral shaft and fixed in place with polymethyl methacrylate. The specimen was then attached to a custom-designed shoulder-positioning device (Figure 2). The device allowed the humerus to be placed in a given plane of elevation (such as the scapular or coronal plane), a given angle of glenohumeral elevation (0 to 100°), and a given angle of humeral rotation (external or internal). The device had double frames by which the abduction angle was measured. The distal tip of the intramedullary rod was passed through a slit of the medial frame of the device and a force transducer was passed through a slit of the lateral frame. Neutral rotation was defined relative to the trunk, which was equivalent to 30° of external rotation relative to the scapular plane. The position of the humerus relative to the scapula was determined using a goniometer attached to the shoulder positioning device. A 22-N force35 was applied to the humeral head against the glenoid fossa through the cables attached to the subscapularis (10-N force), supraspinatus (3.5-N force), and infraspinatus/teres minor tendons (8.5-N force) with pulleys and weights to keep the humeral head centered in the glenoid fossa during the test. We divided that 22-N compression force according to the physiological cross-sectional area3 of each muscle. A screw was inserted perpendicular to the humeral
752
Yamamoto et al
shaft, 10° internally rotated from the plane including the humeral axis and the bicipital groove.18 The screw was used as a reference to indicate the anterior/posterior direction of the humerus. Abduction angles were 0, 30, 45, and 60° relative to the scapula, simulating 0, 30, 60, and 90° of abduction of the arm relative to the trunk.26 The specimen was kept moist with a spray of saline solution applied every 5 to 10 minutes during the test, which was performed at room temperature (24°C). Measurement devices
The 3-dimensional kinematics of the humerus relative to the scapula were monitored with an electromagnetic tracking device (3Space Tracker System, Polhemus Navigation Sciences Division, McDonnell Douglas Electronics Company, Colchester, VT). This system allowed for measurement of the 3-dimensional position and orientation of a sensor in relation to a source. In this experiment, one sensor was attached to the humeral shaft and the other to the scapular spine. The source was fixed to the positioning device (Figure 1). The accuracy of this system has been reported.1,12,28 Because this system had magnetic sensors, we needed to know whether metals, such as the weights and the Kirschner wires used in this experiment, had any influence on the measurements. Our pilot study showed that the magnetic sensors were accurate within 0.1 mm of translation, if the sensor was 5 cm away from the weights. The Kirschner wires did not show any influence on the measurements within the range of our experiment. The range of motion was measured using a goniometer, while a force applied to the distal end of the intramedullary rod in each direction was monitored using a force transducer (Digital force gauge, IMADA, Toyohashi, Japan). Measurement frequency of this force transducer was 20 times per second, the resolution was 0.1 N, and the capacity of measurement was 196 N. The transducer was connected through an amplifier (Polygraph366, NEC, Tokyo, Japan) to a recorder (OMNICORDER 8M15, Sanei, Tokyo, Japan) to record the force data during the test. Experimental conditions
The experiments were performed under the following conditions: with the RI capsule (1) intact, (2) sectioned horizontally, and (3) imbricated either between the superior glenohumeral ligament (SGHL) and the middle glenohumeral ligament (MGHL) (SGHL/MGHL closure group) or between the SGHL and the subscapularis (SSC) (SGHL/SSC closure group), with the translational force applied (1) in the anterior, posterior, and inferior directions with the arm at 0° of abduction (in neutral rotation), (2) in the anterior direction with the arm at 60° of abduction in
J Shoulder Elbow Surg November/December 2006
the scapular plane and 60° of external rotation, (3) in the anterior direction with the arm at 60° of abduction in the coronal plane and maximum external rotation (this position is similar to the one used for an anterior apprehension test) and with a load of (1) 0 N, (2) 10 N, (3) 20 N, and (4) 30 N applied to the proximal humerus. The coronal plane was defined as a plane that was 30° horizontally abducted relative to the scapular plane.21 These loads were selected because a previous study12,27 had indicated that such loads yielded normal ranges of anterior and posterior translation. Measurement under each condition of the RI capsule was performed in the order of (1) with the RI capsule intact, (2) sectioned, and (3) imbricated between the SGHL and SSC (SGHL/SSC closure group) or between the SGHL and the MGHL (SGHL/MGHL closure group). After the experiment, the glenohumeral joint was disarticulated, and the types of MGHL were recorded using DePalma’s classification.6 Preparation of the RI capsule
Before the measurement of intact RI, we punctured the capsule anteriorly with an 18-gauge needle to vent the capsule to eliminate the effect of intraarticular pressure on shoulder stability. The RI capsule was horizontally sectioned using a sharp blade approximately 2 cm in length just below the coracohumeral ligament (Figure 3). The appearance of the intraarticular structures, such as the glenoid rim and the articular surface of the humeral head, through the sectioned capsular opening ensured a complete release of the RI capsule. The imbrication of the RI capsule was performed as follows. In the SGHL/SSC closure group (Figure 4), the superior limb of the first suture (#2 Ethibond, Ethicon, Somerville, NJ) was passed through the superior RI capsule between the anterior edge of the supraspinatus tendon and the SGHL 1 cm medial to the humeral attachment of the RI capsule. The inferior suture limb was then passed through the SSC tendon at the corresponding site of the superior limb placement and 2 to 3 mm inferior to the superior margin of the SSC tendon. The second suture limbs were passed through the same structures but 1 cm medial to the first suture limbs. The knots were tied one by one over the RI capsule. In the SGHL/MGHL closure group, the inferior suture limb was passed through the MGHL where the MGHL crosses the SSC tendon. Although the direction of the sutures in both groups were different, we used these imbrications because these were the procedures performed clinically.30,32 During the imbrication procedures, the shoulder was kept at 0° of abduction and 30° of external rotation to simulate the clinical procedures.
Yamamoto et al
J Shoulder Elbow Surg Volume 15, Number 6
Figure 3 Horizontal sectioning (approximately 2 cm in length) of the mid-portion of the RI capsule. The coracohumeral ligament (CHL) was preserved. SSP, supraspinatus tendon; SSC, subscapularis tendon.
753
Figure 4 RI closure. This photograph shows SGHL/SSC closure. The superior suture limb (1 arrow) was passed through the superior RI capsule between the anterior margin of the supraspinatus tendon (SSP) and the superior glenohumeral ligament (SGHL) 1 cm medial to the humeral attachment of the RI capsule. The inferior suture limb (2 arrows) was passed through the subscapularis tendon (SSC) at the corresponding site of the superior limb placement and 2 to 3 mm inferior to the superior margin of the subscapularis tendon.
Preparation for translation measurement
A translational force was applied to the proximal humerus through nylon cables, which were passed through drill holes in the proximal humerus. The direction of the translational load was perpendicular to the scapular plane when the arm was at 0° of abduction and at 60° of abduction in the scapular plane, whereas it was perpendicular to the coronal plane when the arm was at 60° of abduction in the coronal plane. A rod holder with ball-bearing slide to allow lowfriction anterior/posterior translations of the humerus was fixed at the distal tip of the intramedullary rod when translation was measured (Figure 5). Because of this device, the humeral head was translated smoothly in the anterior/posterior directions when translational forces were applied. Motion measurements
The range of motion was measured: (1) abduction in the scapular plane, (2) internal rotation with the arm at 0° of abduction, (3) external rotation with the arm at 0° of abduction, (4) internal rotation with the arm at 60° of abduction in the scapular plane, (5) external rotation with the arm at 60° of abduction in the scapular plane, and (6) horizontal abduction with
the arm at 60° of abduction and 60° of external rotation. In our pilot study, we determined a set of torque to be applied to the glenohumeral joint during motion measurements. Torque was calculated by multiplying the length of the arm (length from the humeral head to the tip of the intramedullary rod) and applied force. We chose the following set of torque, which seemed to be appropriate to make the glenohumeral joint come to the limit of motion without causing too much tension on the joint capsule: 800 N-mm for abduction, 250 N-mm for internal and external rotations, and 700 N-mm for horizontal abduction. With this torque applied in each direction of measurement, we measured the range of motion using a goniometer attached to the shoulder positioning device. The motion measurements were performed first with the RI capsule intact and, then, with the RI capsule imbricated. Definition of axes
All glenohumeral motions were defined with respect to the scapula. These scapula-referenced motions were not the same as motions that were referenced to the trunk, a commonly used clinical method.
754
Yamamoto et al
J Shoulder Elbow Surg November/December 2006
Figure 6 Anterior translation of the humeral head at 0° of abduction. After the RI capsule was sectioned, the anterior translation increased. After imbrication of the RI capsule, the anterior translation significantly decreased compared with sectioned RI capsule. Values are the mean ⫾ standard deviation. Intact RI, intact rotator interval capsule; sectioned RI, sectioned rotator interval capsule; imbricated RI (SGHL/SSC), imbricated rotator interval capsule between the superior glenohumeral ligament and the subscapularis tendon; imbricated RI (SGHL/MGHL), imbricated rotator interval capsule between the superior glenohumeral ligament and the middle glenohumeral ligament.
Figure 5 A rod holder with ball-bearing slide (arrows) to allow low-friction anterior/posterior translations of the humerus was fixed at the distal tip of the intramedullary rod when translation was measured. Because of this device, the humeral head was translated smoothly in the anterior/posterior directions when translational forces were applied.
The axes of the source of the electromagnetic tracking device on the shoulder positioning device were used as the axes in our experiment. The y-axis was defined as a line passing through the origin (the source) and parallel to the vertical line (the superior/inferior axis). The x-axis was defined as a line passing through the origin, parallel to the horizontal line, and perpendicular to the plane in which the acrylic plates were placed (the anterior/posterior axis). The z-axis was defined as a line passing through the origin and perpendicular to both the x- and y-axes (the medial/ lateral axis). Then, each axis value of the sensor on the humerus with the RI capsule intact and the humerus unloaded, was used as a baseline, and all the changes in the y-axis and x-axis values from the baseline under various conditions were used for analyses of the superior/inferior translations and anterior/posterior translations, respectively. Data analysis
The translation data were compared using a 2-way repeated measures analysis of variance to see the
effects of both capsular conditions and loading. However, there was a significant interaction between these parameters. Therefore, the effects of these parameters were separately analyzed using a 1-way repeated measures analysis of variance. This was also used to compare the range of motion with different capsular conditions. When there was a significant effect, the Scheffé multiple comparisons procedure was used to determine which individual values were different from one another. Statistical analysis was performed with use of statistical software packages (Statview, SAS Institute, Inc., Cary, NC). The statistical significance was set at the 0.05 level. With 14 cadaveric shoulders in each group, there was 80% power to detect a difference in means between the conditions that is equal to 2 standard deviations (alpha ⫽ 0.05, beta ⫽ 0.2). RESULTS Both the SGHL and the MGHL were present in all the specimens. The MGHL was a well-defined, distinct structure in 8 specimens and poorly defined in 6 specimens according to DePalma’s classification.6 Measurement of translation
At 0° of abduction. Sectioning of the RI capsule increased the anterior translation compared with the intact RI capsule but not to a statistically significant level, except for 20 N of load (Figure 6). On the other hand, both imbricated RI groups showed significant reduction in anterior translation com-
Yamamoto et al
J Shoulder Elbow Surg Volume 15, Number 6
Figure 7 Posterior translation of the humeral head at 0° of abduction. SGHL/MGHL closure significantly decreased the posterior translation with 30 N of load. Values are the mean ⫾ standard deviation. Intact RI, intact rotator interval capsule; sectioned RI, sectioned rotator interval capsule; imbricated RI (SGHL/ SSC), imbricated rotator interval capsule between the superior glenohumeral ligament and the subscapularis tendon; imbricated RI (SGHL/MGHL), imbricated rotator interval capsule between the superior glenohumeral ligament and the middle glenohumeral ligament.
Figure 8 Inferior translation of the humeral head at 0° of abduction. The inferior translation did not decrease after RI closure. Values are the mean ⫾ standard deviation. Intact RI, intact rotator interval capsule; sectioned RI, sectioned rotator interval capsule; imbricated RI (SGHL/SSC), imbricated rotator interval capsule between the superior glenohumeral ligament and the subscapularis tendon; imbricated RI (SGHL/MGHL), imbricated rotator interval capsule between the superior glenohumeral ligament and the middle glenohumeral ligament.
pared with sectioned RI, but not with intact RI except for SGHL/MGHL closure with 10 N of load. There was no significant difference between 2 imbricated groups. Posterior translation was significantly decreased only after SGHL/MGHL closure compared with sectioned RI with 30 N of load (P ⫽ .0279) (Figure 7). Although not significant, SGHL/SSC closure showed a decrease in the posterior translation with 30 N of load compared with sectioned RI (P ⫽ .0979). Inferior translation did not show any significant changes even after imbrication (Figure 8).
755
Figure 9 Anterior translation of the humeral head in abduction/ external rotation (in the scapular plane). SGHL/MGHL closure significantly decreased the anterior translation compared with sectioned RI capsule. Values are the mean ⫾ standard deviation. Intact RI, intact rotator interval capsule; sectioned RI, sectioned rotator interval capsule; imbricated RI (SGHL/SSC), imbricated rotator interval capsule between the superior glenohumeral ligament and the subscapularis tendon; imbricated RI (SGHL/MGHL), imbricated rotator interval capsule between the superior glenohumeral ligament and the middle glenohumeral ligament.
At 60° of abduction in the scapular plane/60° of external rotation. SGHL/MGHL closure significantly reduced anterior translation with 20 N and 30 N of load compared with sectioned RI (P ⫽ .0152, P ⫽ .0087) (Figure 9). Furthermore, there was a significant difference between SGHL/MGHL closure and intact RI. SGHL/SSC closure also showed a decrease, but not statistically significant, in anterior translation with 30 N of load compared with sectioned RI (P ⫽ .0972). There was no significant difference between the 2 methods of RI imbrication. At 60° of abduction in the coronal plane/maximum external rotation. There were no significant differences among every capsular condition, except for comparison of SGHL/SSC closure with sectioned RI with 10 N of load (Figure 10). Measurement of glenohumeral motion (Table I)
Both SGHL/SSC closure group and SGHL/MGHL closure group significantly reduced the ranges of external rotation and horizontal abduction compared with the intact RI capsule. Limitation of external rotation at 60° of abduction in the SGHL/MGHL closure group was significantly greater than that in the SGHL/SSC closure group (P ⫽ .0268). External rotation at 0° of abduction showed a difference between 2 imbricated groups, but not at the significant level (P ⫽ .0629). DISCUSSION The RI is a space between the supraspinatus and subscapularis that is covered by the joint capsule and reinforced by the coracohumeral ligament in its
756
Yamamoto et al
Figure 10 Anterior translation of the humeral head in the apprehension position. There were no significant differences among the capsular conditions except for SGHL/SSC closure with 10 N of load. Values are the mean ⫾ standard deviation. Intact RI, intact rotator interval capsule; sectioned RI, sectioned rotator interval capsule; imbricated RI (SGHL/SSC), imbricated rotator interval capsule between the superior glenohumeral ligament and the subscapularis tendon; imbricated RI (SGHL/MGHL), imbricated rotator interval capsule between the superior glenohumeral ligament and the middle glenohumeral ligament.
superficial part and by the SGHL in its deepest part.5 Recently, RI closure has become one of the adjunct procedures to arthroscopic stabilization of the glenohumeral joint. Good clinical results have been reported.2,9,15,30 –32 Most reports were based on the evidence of Harryman’s study,12 and most surgeons expected RI closure to stabilize the glenohumeral joint inferiorly and posteriorly. However, what we do clinically is the superior-inferior closure of the RI, which is different from the medial-lateral closure, which Harryman et al did in their study. There has been no evidence of the effects of RI closure on stability and kinematics of the glenohumeral joint. There are several methods of RI closure. In this study, we selected and compared two imbrication methods, imbrication between the SGHL and subscapularis tendon, and between the SGHL and MGHL, because these two methods are commonly used.2,10,15,30 –32 Our data showed that both imbrication methods decreased the anterior translation at 0° of abduction. In addition, SGHL/MGHL closure decreased posterior translation at 0° of abduction. SGHL/MGHL closure also decreased anterior translation in abduction/external rotation in the scapular plane. Therefore, in cases with an anterior instability in abduction/external rotation in the scapular plane or posterior instability in adduction, SGHL/MGHL closure may be beneficial. However, the reduction of external rotation at 60° of abduction in SGHL/MGHL closure was greater than that in SGHL/SSC closure. From the view point of preserving the range of motion, SGHL/SSC closure may be better than SGHL/MGHL closure, especially in overhead throwing athletes.
J Shoulder Elbow Surg November/December 2006
There are three possible mechanisms of anterior stabilization at 0° of abduction by SGHL/SSC closure: (1) tensioning of the SGHL, (2) the barrier effect of the supraspinatus and subscapularis tendon, and (3) the sealed capsular effect. O’Connell et al,23 in a cadaveric study, reported that the SGHL developed the most strain in adduction and external rotation. In the present study, the RI capsule was imbricated with the arm at 0° of abduction and 30° of external rotation. Therefore, the SGHL may be taut even in the neutral position after imbrication, because the SGHL is pulled down onto the subscapularis tendon (mechanism 1). RI closure between the SGHL and subscapularis also brings the anterior border of the supraspinatus tendon closer to the subscapularis tendon. This makes a barrier composed of the supraspinatus and subscapularis tendons, which is expected to prevent the anterior displacement of the humeral head (mechanism 2). Mechanism 3 is not relevant in this experimental setup, because we had already vented the capsule. Because the MGHL attaches to the capsule covering the surface of the subscapularis tendon, closing the space between the SGHL and MGHL pulls the subscapularis tendon up, closer to the SGHL. Therefore, the above-mentioned three mechanisms also seem to be applicable to the stabilization observed after SGHL/MGHL closure. In addition, a tight MGHL, which is the primary anterior stabilizer with the arm in adduction and external rotation,23 may function as an anterior stabilizer in neutral rotation, because it may become tight even in neutral rotation after RI closure. Kuhn et al16 demonstrated that the contributions of the SGHL and MGHL were equal to that of the inferior glenohumeral ligament (IGHL) with the arm in abduction external rotation in the scapular plane. Therefore, a tight MGHL is likely to increase anterior stability in abduction/external rotation in the scapular plane. SGHL/MGHL closure also contributed to the posterior stability. The tight anterior structures, such as the anterior capsule, SGHL, and MGHL, may have prevented the posterior translation of the humeral head. According to Dempster,7 shoulder stability is provided by both the anterior and posterior structures, because both are observed to force the articular surface of the humeral head against the glenoid. Translation in one direction causes tension in the ligament in the opposite side of the joint, and thus, posterior translation causes increased tension in the anterior structures. This is known as Dempster’s global concept of stability. Our study showed that RI closure added stability in adduction and in abduction/external rotation in the scapular plane but not in abduction/external rotation in the coronal plane. This was probably due to the function of the IGHL.33 With the arm in abduction in
Yamamoto et al
J Shoulder Elbow Surg Volume 15, Number 6
757
Table I The range of motion of SGHL/SSC and SGHL/MGHL closure P-value
Angular range of motion*† (in degrees)
SGHL/SSC closure
SGHL/MGHL closure
SGHL/SSC closure
SGHL/MGHL closure
Comparison of SGHL/SSC and SGHL/MGHL closure
89 ⫾ 3
81 ⫾ 12
83 ⫾ 9
NS
NS
NS
32 ⫾ 11
29 ⫾ 3
26 ⫾ 6
NS
0.0354
NS
51 ⫾ 6
45 ⫾ 6
39 ⫾ 8
NS
0.0270
NS
30 ⫾ 10
27 ⫾ 3
26 ⫾ 5
NS
NS
NS
58 ⫾ 5 28 ⫾ 4
46 ⫾ 8 23 ⫾ 4
41 ⫾ 9 20 ⫾ 7
0.0020 0.0017
0.0105 0.0114
0.0268 NS
Comparison with intact RI‡
Imbricated RI§
Motion Abduction Internal rotation at 0° of abduction External rotation at 0° of abduction Internal rotation at 60° of abduction External rotation at 60° of abduction Horizontal abduction
Intact RI‡
NS ⫽ not significant at the P ⬍ .05 level. *The results are given as the mean and standard deviation. †Achieved by the application of each torque set in advance. ‡Intact RI: intact rotator interval capsule. §Imbricated RI: imbricated rotator interval capsule.
the coronal plane and in maximum external rotation, the IGHL becomes tight and prevents the anterior translation of the humeral head. In this position of the anterior apprehension test, the already tight IGHL prevents anterior translation of the humeral head regardless of RI capsular conditions. In fact, there were no significant differences in translation among all the capsular conditions of RI in this apprehension position. RI closure exercises its stabilizing function where the IGHL does not work because of elongation or because of position. By adding RI closure, greater stability would be obtained and the better outcomes of arthroscopic stabilization would be expected. In the present study, inferior translation in adduction did not show any significant changes even after imbrication. This could be explained by two reasons. First, before the measurement of translation, we punctured the capsule with a needle to create a vented condition to eliminate the effect of intraarticular pressure on shoulder stability. With the arm in adduction, we know that intraarticular pressure is an important inferior stabilizer.13,17 Therefore, intraarticular pressure may affect inferior translation in actual patients. However, this is not relevant in this experimental setup because we already vented the capsule. Second, the coracohumeral ligament was preserved to the extent of our ability in the present study. When the RI capsule was sectioned below the coracohumeral ligament, care was taken not to damage the ligament, because it was an inferior stabilizer.14,24 On the other hand, a cadaveric study by Harryman et al12 revealed that the RI capsule was sectioned and tight-
ened together with the coracohumeral ligament. They reported that imbrication of the RI capsule increased the resistance to inferior translation. This difference between the present study and the Harryman’s study might cause different results. Most recent investigations10,15,30 have supported the concept that even anterior-inferior instability is associated with multiple lesions, in addition to the Bankart lesion. Stokes et al30 reported that the RI was almost always widened in recurrent instability. As the humeral head dislocates anteriorly and the anterior capsulolabral structures are torn, a secondary lesion may occur following the dislocation. In their experience, the secondary lesions most often involve the RI. Gartsman et al10 also reported that a tear or redundancy of the RI capsule was 1 of the lesions identified in patients with glenohumeral instability. Furthermore, Karas et al15 reported that primary interval pathology was indicated by a patulous interval capsule. In the literature, we found no report describing its true incidence among patients with anterior glenohumeral instability. If the RI lesions are present in patients with glenohumeral instability, RI closure is expected to provide further stability as our results demonstrated. The present study had several limitations. First, simulated RI closure is different from actual RI closure because the RI capsule was cut and then imbricated. In addition, the capsular condition of the cadaveric specimens was different from that of patients with glenohumeral instability. We also applied 10 N, 20 N, and 30 N of load to the humeral head. Appling a uniform load to the humeral head was not physio-
758
Yamamoto et al
logic, although it made it easier to compare the stabilizing function of the RI closure. In conclusion, RI closure reduces anterior/posterior translation in adduction and anterior translation at 60° of abduction in the scapular plane. Based on the results of the present study, RI closure is expected to reduce remnant anterior/posterior instability and, thereby, improve the clinical outcomes of arthroscopic stabilization procedures. However, RI closure should be carefully considered in overhead throwing athletes, because it decreases the range of external rotation and horizontal abduction.
J Shoulder Elbow Surg November/December 2006
15. 16. 17. 18. 19. 20.
REFERENCES
1. An KN, Jacobsen MC, Berglund LJ, Chao EYS. Application of a magnetic tracking device to kinesiologic studies. J Biomech 1988; 21:613-20. 2. Angelo RL. Controversies in arthroscopic shoulder surgery: arthroscopic versus open Bankart repair, thermal treatment of capsular tissue, acromioplasties—are they necessary? Arthroscopy 2003; 19:224-8. 3. Bassett R, Browne A, An KN, Morrey BF. A biomechanical analysis of shoulder muscles in the position of instability. J Biomech 1990;23:405-7. 4. Benedetto KP, Glotzer W. Arthroscopic Bankart procedure by suture technique: indications, technique, and results. Arthroscopy 1992;8:111-5. 5. Clark JM, Harryman DT II. Tendons, ligaments, and capsule of the rotator cuff. Gross and microscopic anatomy. J Bone Joint Surg Am 1992;74:713-25. 6. DePalma AF, Callery G, and Bennett GA. Shoulder joint: variational anatomy and degenerative regions of the shoulder joint. Instr Course Lect 1949;6:255-81. 7. Dempster WT. Mechanism of shoulder movement. Arch Phys Med Rehablil 1965;46:49-70. 8. Field LD, Warren RF, O’Brien SJ, Altchek DW, Wickiewicz TL. Isolated closure of rotator interval defects for shoulder instability. Am J Sport Med 1995;23:557-63. 9. Gartsman GM, Roddey TS, Hammerman SM. Arthroscopic treatment of antero-inferior glenohumeral instability. J Bone Joint Surg Am 2000;82:991-1003. 10. Gartsman GM, Taverna E, Hammerman SM. Arthroscopic rotator interval repair in glenohumeral instability: description of an operative technique. Arthroscopy 1999;15:330-2. 11. Grana WA, Buckley PD, Yates CK. Arthroscopic Bankart suture repair. Am J Sports Med 1993;21:348-53. 12. Harryman DT II, Sides JA, Harris SL, Matsen FA III. The role of the rotator interval capsule in passive motion and stability of the shoulder. J Bone Joint Surg Am 1992;74:53-66. 13. Itoi E, Motzkin NE, Browne AO, Hoffmeyer P, Morrey BF, An KN. Intraarticular pressure of the shoulder. Arthroscopy 1993;9: 406-13. 14. Itoi E, Berglund LJ, Grabowski JJ, Naggar L, Morrey BF, An KN. Superior-inferior stability of the shoulder: role of the coracohu-
21. 22. 23. 24. 25.
26. 27. 28. 29. 30. 31. 32. 33. 34. 35.
meral ligament and the rotator interval capsule. Mayo Clin Proc 1998;73:508-15. Karas SG. Arthroscopic rotator interval repair and anterior portal closure an alternative technique. Arthroscopy 2002;18:436-9. Kuhn JE, Bey MJ, Huston LJ, Blasier RB, Soslowsky LJ. Ligamentous restraints to external rotation of the humerus in the late-cocking phase of throwing. Am J Sports Med 2000;28:200-5. Kumar VP, Balasubramanian P. The role of atmospheric pressure in stabilizing the shoulder. An experimental study. J Bone Joint Surg Br 1985;67:719-21. Matsen FA III, Kirby RM. Office evaluation and management of shoulder pain. Orthop Clin North Am 1982;13:453-75. Matthews LS, Vetter WL, Oweida SJ, J Spearman, DL. Arthroscopic staple capsulorrhaphy for recurrent anterior shoulder instability. Arthroscopy 1988;4:106-11. Morgan CD, Bodenstab AB. Arthroscopic Bankart suture repair: technique and early results. Arthroscopy 1987;3:111-22. Morrey BF, Itoi E, An K-N. Biomechanics of the shoulder. In: Rockwood CA Jr, Matsen FA III, editors. The shoulder. 2nd ed. Philadelphia: WB Saunders; 1998, p. 233-76. Nobuhara K, Ikeda H. Rotator interval lesion. Clin Orthop 1987; 223: 44-50. O’Connell PW, Nuber GW, Mileski RA, Lautenschlager E. The contribution of the glenohumeral ligaments to anterior stability of the shoulder. Am J Sport Med 1990;18:579-84. Ovesen J, Nielsen S. Experimental distal subluxation in the glenohumeral joint. Arch Orthop Trauma Surg 1985;104:78-81. Pagnani MJ, Warren RF, Altchek DW, Wickiewicz TL, Anderson AF. Arthroscopic shoulder stabilization using transglenoid sutures: a four-year minimum follow-up. Am J Sports Med 1996;24: 459-67. Poppen NK, Walker PS. Normal and abnormal motion of the shoulder. J Bone Joint Surg Am 1976;58:195-201. Selecky MT, Tibone JE, Yang BY, McMahon PJ, Lee TQ. Glenohumeral joint translation after arthroscopic thermal capsuloplasty of the rotator interval. J Shoulder Elbow Surg 2003;12:139-43. Sidles JA, Larson RV, Garbini JL, Downey DJ, Matsen FA III. Ligament length relationships in the moving knee. J Orthop Res 1988;6:593-610. Speer KP, Warren RF, Pagnani M, Warner JJ. An arthroscopic technique for anterior stabilization of the shoulder with a bioabsorbable tack. J Bone Joint Surg Am 1996;78:1801-7. Stokes DA, Savoie FH, Field LD. Arthroscopic repair of anterior glenohumeral instability and rotator interval lesions. Orthop Clin North Am 2003;34:529-39. Taverna E, Sansone V, Basttistella F. Arthroscopic rotator interval repair: the three-step all-inside technique. Arthroscopy 2004;20: 105-9. Treacy SH, Field LD, Savoie FH. Rotator interval capsule closure. An arthroscopic technique. Arthroscopy 1997;13:103-6. Turkel SL, Panio MW, Marshall JL, Girgis FG. Stabilizing mechanisms preventing anterior dislocation of the glenohumeral joint. J Bone Joint Surg Am 1981;63:1208-17. Walch G, Boileau P, Levigne C, Mandino A, Neyret P, Donell S. Arthroscopic stabilization for recurrent anterior shoulder dislocation: results of 59 cases. Arthroscopy 1995;11:173-9. Warner JJ, Deng XH, Warren RF, Trozilli PA. Static capsuloligamentous restraints to superior-inferior translation of the glenohumeral joint. Am J Sport Med 1992;20:675-85.
The effect of rotator interval closure, which is performed as an adjunct to arthroscopic stabilization of the shoulder, has not been clarified. Fourteen freshfrozen cadaveric shoulders were used. The position of the humeral head was measured using an electromagnetic tracking device with the capsule intact, sectioned, and imbricated between the superior glenohumeral ligament and the subscapularis tendon (SGHL/ SSC closure) or between the superior and middle glenohumeral ligaments (SGHL/MGHL closure). The direction of translational loads (10, 20, and 30 N) and arm positions were (1) anterior, posterior, and inferior loads in adduction; (2) anterior load in abduction/external rotation in the scapular plane; and (3) anterior load in abduction/external rotation in the coronal plane. The range of motion was measured using a goniometer under a constant force. Both methods reduced anterior translation in adduction. Only SGHL/MGHL closure reduced anterior translation in abduction/external rotation in the scapular plane and posterior translation in adduction. Both methods reduced the range of external rotation and horizontal abduction. Rotator interval closure is expected to reduce remnant anterior/posterior instability and thereby improve the clinical outcomes of arthroscopic stabilization procedures. (J Shoulder Elbow Surg 2006;15: 750-758.)
P revious studies showed that results of the arthro-
scopic stabilization of the glenohumeral joint were inferior to those of open procedures in terms of recurrence rate.4,11,19,20,25,29,34 Recently, however, the outcome of arthroscopic stabilization has improved dramatically.2,9,30,32 This may be due to advances in From the Division of Orthopedic Surgery, Department of Neuro and Locomotor Science, Akita University School of Medicine. Reprint requests to: Eiji Itoi, MD, Department of Orthopedic Surgery, Tohoku University School of Medicine, 1-1 Seiryo-machi, Aoba-ku, Sendai 980-8574, Japan (E-mail: itoi-eiji@mail. tains.tohoku.ac.jp). Copyright © 2006 by Journal of Shoulder and Elbow Surgery Board of Trustees. 1058-2746/2006/$32.00 doi:10.1016/j.jse.2005.12.009
750
techniques and instrumentation. This may also be due to adjunct procedures performed during stabilization procedures such as suture plication or thermal capsulorrhaphy. Rotator interval (RI) closure is one of these adjunct procedures, which is thought to correct pathoanatomy of shoulder instability, and as a result, to improve the outcome of arthroscopic instability repairs.2,10,15,30 –32 Gartsman et al10 found that repair of the RI was a critical factor in 14 of 53 shoulders treated arthroscopically for anteroinferior glenohumeral instability, and that RI repair contributed to the improved clinical outcomes observed in their study. The recent investigations have supported the concept that even anterior-inferior instability is associated with multiple lesions.9,30 Capsular and ligamentous laxity or elongation are present in various degrees in most patients with shoulder instability, and RI lesions are one of a host of potential lesions.9,15,30,32 The significance of an RI lesion in glenohumeral instability has become increasingly apparent, as this anatomic structure has been shown to contribute to shoulder stability in clinical and biomechanical studies.8,14,22 We have found only one biomechanical study, by Harryman et al12, that investigated quantitatively the effect of operative sectioning and imbrication of the RI. They reported the effect of vertical sectioning and medial-lateral closure of the RI capsule on motion and stability of the shoulder. Their conclusion was that imbrication of the RI capsule increased the resistance to inferior and posterior translations. However, it is not the medial-lateral closure, but the superior-inferior closure of the RI that is commonly performed clinically. To our knowledge, there has been no study to clarify the effect of superior-inferior imbrication of the RI on shoulder biomechanics. We hypothesized that closure of the RI capsule would increase resistance to anterior translation of the humeral head in normal shoulders, as well as in shoulders with anterior instability. The purpose of the current study was to determine the effects of closure of the RI with respect to glenohumeral motion and stability.
J Shoulder Elbow Surg Volume 15, Number 6
Figure 1 Specimen and the electromagnetic sensors. A pair of acrylic plates was fixed to the medial border of the scapula by means of plastic screws (1 arrow). Another pair of acrylic plates was fixed to the scapular body by means of Kirschner wires (2 arrows). One electromagnetic sensor (2 arrowheads) was attached to the humeral shaft and the other (1 arrowhead) was attached to the scapular spine. The source was attached to the positioning device.
MATERIALS AND METHODS Preparation of specimens
We used 14 fresh-frozen cadaveric shoulders from donors who were 25 to 88 years old (mean, 64 years old) at the time of death. The frozen shoulders were thawed overnight at room temperature for experimentation. None of the shoulders had a rotator cuff tear or radiologic evidence of glenohumeral osteoarthritis. Each specimen had been disarticulated at the scapulothoracic joint proximally and dissected at the midpart of the humerus distal to the deltoid attachment distally. The skin, subcutaneous tissue, and all of the muscles were removed except for those of the rotator cuff, which were elevated from the scapula. The lateral 1/3 of the scapular spine, including the acromion and the distal half of the coracoid process, was removed. The coracohumeral ligament was preserved. A pair of acrylic plates (240 mm in length ⫻ 40 mm in width) was fixed to the medial border of the scapula by means of plastic screws, 4 mm in diameter (Figure 1). Another pair of acrylic plates (220 mm in length ⫻ 130 mm in width) was fixed to the scapular body by means of the Kirschner wires (1.8 mm in diameter). The scapula was sandwiched between 2 pair of plates, allowing free motion of the cables in between, which were attached to the rotator cuff tendons. An intramedullary rod (10 mm in diameter)
Yamamoto et al
751
Figure 2 Custom-designed shoulder-positioning device.
was inserted into the proximal humeral shaft and fixed in place with polymethyl methacrylate. The specimen was then attached to a custom-designed shoulder-positioning device (Figure 2). The device allowed the humerus to be placed in a given plane of elevation (such as the scapular or coronal plane), a given angle of glenohumeral elevation (0 to 100°), and a given angle of humeral rotation (external or internal). The device had double frames by which the abduction angle was measured. The distal tip of the intramedullary rod was passed through a slit of the medial frame of the device and a force transducer was passed through a slit of the lateral frame. Neutral rotation was defined relative to the trunk, which was equivalent to 30° of external rotation relative to the scapular plane. The position of the humerus relative to the scapula was determined using a goniometer attached to the shoulder positioning device. A 22-N force35 was applied to the humeral head against the glenoid fossa through the cables attached to the subscapularis (10-N force), supraspinatus (3.5-N force), and infraspinatus/teres minor tendons (8.5-N force) with pulleys and weights to keep the humeral head centered in the glenoid fossa during the test. We divided that 22-N compression force according to the physiological cross-sectional area3 of each muscle. A screw was inserted perpendicular to the humeral
752
Yamamoto et al
shaft, 10° internally rotated from the plane including the humeral axis and the bicipital groove.18 The screw was used as a reference to indicate the anterior/posterior direction of the humerus. Abduction angles were 0, 30, 45, and 60° relative to the scapula, simulating 0, 30, 60, and 90° of abduction of the arm relative to the trunk.26 The specimen was kept moist with a spray of saline solution applied every 5 to 10 minutes during the test, which was performed at room temperature (24°C). Measurement devices
The 3-dimensional kinematics of the humerus relative to the scapula were monitored with an electromagnetic tracking device (3Space Tracker System, Polhemus Navigation Sciences Division, McDonnell Douglas Electronics Company, Colchester, VT). This system allowed for measurement of the 3-dimensional position and orientation of a sensor in relation to a source. In this experiment, one sensor was attached to the humeral shaft and the other to the scapular spine. The source was fixed to the positioning device (Figure 1). The accuracy of this system has been reported.1,12,28 Because this system had magnetic sensors, we needed to know whether metals, such as the weights and the Kirschner wires used in this experiment, had any influence on the measurements. Our pilot study showed that the magnetic sensors were accurate within 0.1 mm of translation, if the sensor was 5 cm away from the weights. The Kirschner wires did not show any influence on the measurements within the range of our experiment. The range of motion was measured using a goniometer, while a force applied to the distal end of the intramedullary rod in each direction was monitored using a force transducer (Digital force gauge, IMADA, Toyohashi, Japan). Measurement frequency of this force transducer was 20 times per second, the resolution was 0.1 N, and the capacity of measurement was 196 N. The transducer was connected through an amplifier (Polygraph366, NEC, Tokyo, Japan) to a recorder (OMNICORDER 8M15, Sanei, Tokyo, Japan) to record the force data during the test. Experimental conditions
The experiments were performed under the following conditions: with the RI capsule (1) intact, (2) sectioned horizontally, and (3) imbricated either between the superior glenohumeral ligament (SGHL) and the middle glenohumeral ligament (MGHL) (SGHL/MGHL closure group) or between the SGHL and the subscapularis (SSC) (SGHL/SSC closure group), with the translational force applied (1) in the anterior, posterior, and inferior directions with the arm at 0° of abduction (in neutral rotation), (2) in the anterior direction with the arm at 60° of abduction in
J Shoulder Elbow Surg November/December 2006
the scapular plane and 60° of external rotation, (3) in the anterior direction with the arm at 60° of abduction in the coronal plane and maximum external rotation (this position is similar to the one used for an anterior apprehension test) and with a load of (1) 0 N, (2) 10 N, (3) 20 N, and (4) 30 N applied to the proximal humerus. The coronal plane was defined as a plane that was 30° horizontally abducted relative to the scapular plane.21 These loads were selected because a previous study12,27 had indicated that such loads yielded normal ranges of anterior and posterior translation. Measurement under each condition of the RI capsule was performed in the order of (1) with the RI capsule intact, (2) sectioned, and (3) imbricated between the SGHL and SSC (SGHL/SSC closure group) or between the SGHL and the MGHL (SGHL/MGHL closure group). After the experiment, the glenohumeral joint was disarticulated, and the types of MGHL were recorded using DePalma’s classification.6 Preparation of the RI capsule
Before the measurement of intact RI, we punctured the capsule anteriorly with an 18-gauge needle to vent the capsule to eliminate the effect of intraarticular pressure on shoulder stability. The RI capsule was horizontally sectioned using a sharp blade approximately 2 cm in length just below the coracohumeral ligament (Figure 3). The appearance of the intraarticular structures, such as the glenoid rim and the articular surface of the humeral head, through the sectioned capsular opening ensured a complete release of the RI capsule. The imbrication of the RI capsule was performed as follows. In the SGHL/SSC closure group (Figure 4), the superior limb of the first suture (#2 Ethibond, Ethicon, Somerville, NJ) was passed through the superior RI capsule between the anterior edge of the supraspinatus tendon and the SGHL 1 cm medial to the humeral attachment of the RI capsule. The inferior suture limb was then passed through the SSC tendon at the corresponding site of the superior limb placement and 2 to 3 mm inferior to the superior margin of the SSC tendon. The second suture limbs were passed through the same structures but 1 cm medial to the first suture limbs. The knots were tied one by one over the RI capsule. In the SGHL/MGHL closure group, the inferior suture limb was passed through the MGHL where the MGHL crosses the SSC tendon. Although the direction of the sutures in both groups were different, we used these imbrications because these were the procedures performed clinically.30,32 During the imbrication procedures, the shoulder was kept at 0° of abduction and 30° of external rotation to simulate the clinical procedures.
Yamamoto et al
J Shoulder Elbow Surg Volume 15, Number 6
Figure 3 Horizontal sectioning (approximately 2 cm in length) of the mid-portion of the RI capsule. The coracohumeral ligament (CHL) was preserved. SSP, supraspinatus tendon; SSC, subscapularis tendon.
753
Figure 4 RI closure. This photograph shows SGHL/SSC closure. The superior suture limb (1 arrow) was passed through the superior RI capsule between the anterior margin of the supraspinatus tendon (SSP) and the superior glenohumeral ligament (SGHL) 1 cm medial to the humeral attachment of the RI capsule. The inferior suture limb (2 arrows) was passed through the subscapularis tendon (SSC) at the corresponding site of the superior limb placement and 2 to 3 mm inferior to the superior margin of the subscapularis tendon.
Preparation for translation measurement
A translational force was applied to the proximal humerus through nylon cables, which were passed through drill holes in the proximal humerus. The direction of the translational load was perpendicular to the scapular plane when the arm was at 0° of abduction and at 60° of abduction in the scapular plane, whereas it was perpendicular to the coronal plane when the arm was at 60° of abduction in the coronal plane. A rod holder with ball-bearing slide to allow lowfriction anterior/posterior translations of the humerus was fixed at the distal tip of the intramedullary rod when translation was measured (Figure 5). Because of this device, the humeral head was translated smoothly in the anterior/posterior directions when translational forces were applied. Motion measurements
The range of motion was measured: (1) abduction in the scapular plane, (2) internal rotation with the arm at 0° of abduction, (3) external rotation with the arm at 0° of abduction, (4) internal rotation with the arm at 60° of abduction in the scapular plane, (5) external rotation with the arm at 60° of abduction in the scapular plane, and (6) horizontal abduction with
the arm at 60° of abduction and 60° of external rotation. In our pilot study, we determined a set of torque to be applied to the glenohumeral joint during motion measurements. Torque was calculated by multiplying the length of the arm (length from the humeral head to the tip of the intramedullary rod) and applied force. We chose the following set of torque, which seemed to be appropriate to make the glenohumeral joint come to the limit of motion without causing too much tension on the joint capsule: 800 N-mm for abduction, 250 N-mm for internal and external rotations, and 700 N-mm for horizontal abduction. With this torque applied in each direction of measurement, we measured the range of motion using a goniometer attached to the shoulder positioning device. The motion measurements were performed first with the RI capsule intact and, then, with the RI capsule imbricated. Definition of axes
All glenohumeral motions were defined with respect to the scapula. These scapula-referenced motions were not the same as motions that were referenced to the trunk, a commonly used clinical method.
754
Yamamoto et al
J Shoulder Elbow Surg November/December 2006
Figure 6 Anterior translation of the humeral head at 0° of abduction. After the RI capsule was sectioned, the anterior translation increased. After imbrication of the RI capsule, the anterior translation significantly decreased compared with sectioned RI capsule. Values are the mean ⫾ standard deviation. Intact RI, intact rotator interval capsule; sectioned RI, sectioned rotator interval capsule; imbricated RI (SGHL/SSC), imbricated rotator interval capsule between the superior glenohumeral ligament and the subscapularis tendon; imbricated RI (SGHL/MGHL), imbricated rotator interval capsule between the superior glenohumeral ligament and the middle glenohumeral ligament.
Figure 5 A rod holder with ball-bearing slide (arrows) to allow low-friction anterior/posterior translations of the humerus was fixed at the distal tip of the intramedullary rod when translation was measured. Because of this device, the humeral head was translated smoothly in the anterior/posterior directions when translational forces were applied.
The axes of the source of the electromagnetic tracking device on the shoulder positioning device were used as the axes in our experiment. The y-axis was defined as a line passing through the origin (the source) and parallel to the vertical line (the superior/inferior axis). The x-axis was defined as a line passing through the origin, parallel to the horizontal line, and perpendicular to the plane in which the acrylic plates were placed (the anterior/posterior axis). The z-axis was defined as a line passing through the origin and perpendicular to both the x- and y-axes (the medial/ lateral axis). Then, each axis value of the sensor on the humerus with the RI capsule intact and the humerus unloaded, was used as a baseline, and all the changes in the y-axis and x-axis values from the baseline under various conditions were used for analyses of the superior/inferior translations and anterior/posterior translations, respectively. Data analysis
The translation data were compared using a 2-way repeated measures analysis of variance to see the
effects of both capsular conditions and loading. However, there was a significant interaction between these parameters. Therefore, the effects of these parameters were separately analyzed using a 1-way repeated measures analysis of variance. This was also used to compare the range of motion with different capsular conditions. When there was a significant effect, the Scheffé multiple comparisons procedure was used to determine which individual values were different from one another. Statistical analysis was performed with use of statistical software packages (Statview, SAS Institute, Inc., Cary, NC). The statistical significance was set at the 0.05 level. With 14 cadaveric shoulders in each group, there was 80% power to detect a difference in means between the conditions that is equal to 2 standard deviations (alpha ⫽ 0.05, beta ⫽ 0.2). RESULTS Both the SGHL and the MGHL were present in all the specimens. The MGHL was a well-defined, distinct structure in 8 specimens and poorly defined in 6 specimens according to DePalma’s classification.6 Measurement of translation
At 0° of abduction. Sectioning of the RI capsule increased the anterior translation compared with the intact RI capsule but not to a statistically significant level, except for 20 N of load (Figure 6). On the other hand, both imbricated RI groups showed significant reduction in anterior translation com-
Yamamoto et al
J Shoulder Elbow Surg Volume 15, Number 6
Figure 7 Posterior translation of the humeral head at 0° of abduction. SGHL/MGHL closure significantly decreased the posterior translation with 30 N of load. Values are the mean ⫾ standard deviation. Intact RI, intact rotator interval capsule; sectioned RI, sectioned rotator interval capsule; imbricated RI (SGHL/ SSC), imbricated rotator interval capsule between the superior glenohumeral ligament and the subscapularis tendon; imbricated RI (SGHL/MGHL), imbricated rotator interval capsule between the superior glenohumeral ligament and the middle glenohumeral ligament.
Figure 8 Inferior translation of the humeral head at 0° of abduction. The inferior translation did not decrease after RI closure. Values are the mean ⫾ standard deviation. Intact RI, intact rotator interval capsule; sectioned RI, sectioned rotator interval capsule; imbricated RI (SGHL/SSC), imbricated rotator interval capsule between the superior glenohumeral ligament and the subscapularis tendon; imbricated RI (SGHL/MGHL), imbricated rotator interval capsule between the superior glenohumeral ligament and the middle glenohumeral ligament.
pared with sectioned RI, but not with intact RI except for SGHL/MGHL closure with 10 N of load. There was no significant difference between 2 imbricated groups. Posterior translation was significantly decreased only after SGHL/MGHL closure compared with sectioned RI with 30 N of load (P ⫽ .0279) (Figure 7). Although not significant, SGHL/SSC closure showed a decrease in the posterior translation with 30 N of load compared with sectioned RI (P ⫽ .0979). Inferior translation did not show any significant changes even after imbrication (Figure 8).
755
Figure 9 Anterior translation of the humeral head in abduction/ external rotation (in the scapular plane). SGHL/MGHL closure significantly decreased the anterior translation compared with sectioned RI capsule. Values are the mean ⫾ standard deviation. Intact RI, intact rotator interval capsule; sectioned RI, sectioned rotator interval capsule; imbricated RI (SGHL/SSC), imbricated rotator interval capsule between the superior glenohumeral ligament and the subscapularis tendon; imbricated RI (SGHL/MGHL), imbricated rotator interval capsule between the superior glenohumeral ligament and the middle glenohumeral ligament.
At 60° of abduction in the scapular plane/60° of external rotation. SGHL/MGHL closure significantly reduced anterior translation with 20 N and 30 N of load compared with sectioned RI (P ⫽ .0152, P ⫽ .0087) (Figure 9). Furthermore, there was a significant difference between SGHL/MGHL closure and intact RI. SGHL/SSC closure also showed a decrease, but not statistically significant, in anterior translation with 30 N of load compared with sectioned RI (P ⫽ .0972). There was no significant difference between the 2 methods of RI imbrication. At 60° of abduction in the coronal plane/maximum external rotation. There were no significant differences among every capsular condition, except for comparison of SGHL/SSC closure with sectioned RI with 10 N of load (Figure 10). Measurement of glenohumeral motion (Table I)
Both SGHL/SSC closure group and SGHL/MGHL closure group significantly reduced the ranges of external rotation and horizontal abduction compared with the intact RI capsule. Limitation of external rotation at 60° of abduction in the SGHL/MGHL closure group was significantly greater than that in the SGHL/SSC closure group (P ⫽ .0268). External rotation at 0° of abduction showed a difference between 2 imbricated groups, but not at the significant level (P ⫽ .0629). DISCUSSION The RI is a space between the supraspinatus and subscapularis that is covered by the joint capsule and reinforced by the coracohumeral ligament in its
756
Yamamoto et al
Figure 10 Anterior translation of the humeral head in the apprehension position. There were no significant differences among the capsular conditions except for SGHL/SSC closure with 10 N of load. Values are the mean ⫾ standard deviation. Intact RI, intact rotator interval capsule; sectioned RI, sectioned rotator interval capsule; imbricated RI (SGHL/SSC), imbricated rotator interval capsule between the superior glenohumeral ligament and the subscapularis tendon; imbricated RI (SGHL/MGHL), imbricated rotator interval capsule between the superior glenohumeral ligament and the middle glenohumeral ligament.
superficial part and by the SGHL in its deepest part.5 Recently, RI closure has become one of the adjunct procedures to arthroscopic stabilization of the glenohumeral joint. Good clinical results have been reported.2,9,15,30 –32 Most reports were based on the evidence of Harryman’s study,12 and most surgeons expected RI closure to stabilize the glenohumeral joint inferiorly and posteriorly. However, what we do clinically is the superior-inferior closure of the RI, which is different from the medial-lateral closure, which Harryman et al did in their study. There has been no evidence of the effects of RI closure on stability and kinematics of the glenohumeral joint. There are several methods of RI closure. In this study, we selected and compared two imbrication methods, imbrication between the SGHL and subscapularis tendon, and between the SGHL and MGHL, because these two methods are commonly used.2,10,15,30 –32 Our data showed that both imbrication methods decreased the anterior translation at 0° of abduction. In addition, SGHL/MGHL closure decreased posterior translation at 0° of abduction. SGHL/MGHL closure also decreased anterior translation in abduction/external rotation in the scapular plane. Therefore, in cases with an anterior instability in abduction/external rotation in the scapular plane or posterior instability in adduction, SGHL/MGHL closure may be beneficial. However, the reduction of external rotation at 60° of abduction in SGHL/MGHL closure was greater than that in SGHL/SSC closure. From the view point of preserving the range of motion, SGHL/SSC closure may be better than SGHL/MGHL closure, especially in overhead throwing athletes.
J Shoulder Elbow Surg November/December 2006
There are three possible mechanisms of anterior stabilization at 0° of abduction by SGHL/SSC closure: (1) tensioning of the SGHL, (2) the barrier effect of the supraspinatus and subscapularis tendon, and (3) the sealed capsular effect. O’Connell et al,23 in a cadaveric study, reported that the SGHL developed the most strain in adduction and external rotation. In the present study, the RI capsule was imbricated with the arm at 0° of abduction and 30° of external rotation. Therefore, the SGHL may be taut even in the neutral position after imbrication, because the SGHL is pulled down onto the subscapularis tendon (mechanism 1). RI closure between the SGHL and subscapularis also brings the anterior border of the supraspinatus tendon closer to the subscapularis tendon. This makes a barrier composed of the supraspinatus and subscapularis tendons, which is expected to prevent the anterior displacement of the humeral head (mechanism 2). Mechanism 3 is not relevant in this experimental setup, because we had already vented the capsule. Because the MGHL attaches to the capsule covering the surface of the subscapularis tendon, closing the space between the SGHL and MGHL pulls the subscapularis tendon up, closer to the SGHL. Therefore, the above-mentioned three mechanisms also seem to be applicable to the stabilization observed after SGHL/MGHL closure. In addition, a tight MGHL, which is the primary anterior stabilizer with the arm in adduction and external rotation,23 may function as an anterior stabilizer in neutral rotation, because it may become tight even in neutral rotation after RI closure. Kuhn et al16 demonstrated that the contributions of the SGHL and MGHL were equal to that of the inferior glenohumeral ligament (IGHL) with the arm in abduction external rotation in the scapular plane. Therefore, a tight MGHL is likely to increase anterior stability in abduction/external rotation in the scapular plane. SGHL/MGHL closure also contributed to the posterior stability. The tight anterior structures, such as the anterior capsule, SGHL, and MGHL, may have prevented the posterior translation of the humeral head. According to Dempster,7 shoulder stability is provided by both the anterior and posterior structures, because both are observed to force the articular surface of the humeral head against the glenoid. Translation in one direction causes tension in the ligament in the opposite side of the joint, and thus, posterior translation causes increased tension in the anterior structures. This is known as Dempster’s global concept of stability. Our study showed that RI closure added stability in adduction and in abduction/external rotation in the scapular plane but not in abduction/external rotation in the coronal plane. This was probably due to the function of the IGHL.33 With the arm in abduction in
Yamamoto et al
J Shoulder Elbow Surg Volume 15, Number 6
757
Table I The range of motion of SGHL/SSC and SGHL/MGHL closure P-value
Angular range of motion*† (in degrees)
SGHL/SSC closure
SGHL/MGHL closure
SGHL/SSC closure
SGHL/MGHL closure
Comparison of SGHL/SSC and SGHL/MGHL closure
89 ⫾ 3
81 ⫾ 12
83 ⫾ 9
NS
NS
NS
32 ⫾ 11
29 ⫾ 3
26 ⫾ 6
NS
0.0354
NS
51 ⫾ 6
45 ⫾ 6
39 ⫾ 8
NS
0.0270
NS
30 ⫾ 10
27 ⫾ 3
26 ⫾ 5
NS
NS
NS
58 ⫾ 5 28 ⫾ 4
46 ⫾ 8 23 ⫾ 4
41 ⫾ 9 20 ⫾ 7
0.0020 0.0017
0.0105 0.0114
0.0268 NS
Comparison with intact RI‡
Imbricated RI§
Motion Abduction Internal rotation at 0° of abduction External rotation at 0° of abduction Internal rotation at 60° of abduction External rotation at 60° of abduction Horizontal abduction
Intact RI‡
NS ⫽ not significant at the P ⬍ .05 level. *The results are given as the mean and standard deviation. †Achieved by the application of each torque set in advance. ‡Intact RI: intact rotator interval capsule. §Imbricated RI: imbricated rotator interval capsule.
the coronal plane and in maximum external rotation, the IGHL becomes tight and prevents the anterior translation of the humeral head. In this position of the anterior apprehension test, the already tight IGHL prevents anterior translation of the humeral head regardless of RI capsular conditions. In fact, there were no significant differences in translation among all the capsular conditions of RI in this apprehension position. RI closure exercises its stabilizing function where the IGHL does not work because of elongation or because of position. By adding RI closure, greater stability would be obtained and the better outcomes of arthroscopic stabilization would be expected. In the present study, inferior translation in adduction did not show any significant changes even after imbrication. This could be explained by two reasons. First, before the measurement of translation, we punctured the capsule with a needle to create a vented condition to eliminate the effect of intraarticular pressure on shoulder stability. With the arm in adduction, we know that intraarticular pressure is an important inferior stabilizer.13,17 Therefore, intraarticular pressure may affect inferior translation in actual patients. However, this is not relevant in this experimental setup because we already vented the capsule. Second, the coracohumeral ligament was preserved to the extent of our ability in the present study. When the RI capsule was sectioned below the coracohumeral ligament, care was taken not to damage the ligament, because it was an inferior stabilizer.14,24 On the other hand, a cadaveric study by Harryman et al12 revealed that the RI capsule was sectioned and tight-
ened together with the coracohumeral ligament. They reported that imbrication of the RI capsule increased the resistance to inferior translation. This difference between the present study and the Harryman’s study might cause different results. Most recent investigations10,15,30 have supported the concept that even anterior-inferior instability is associated with multiple lesions, in addition to the Bankart lesion. Stokes et al30 reported that the RI was almost always widened in recurrent instability. As the humeral head dislocates anteriorly and the anterior capsulolabral structures are torn, a secondary lesion may occur following the dislocation. In their experience, the secondary lesions most often involve the RI. Gartsman et al10 also reported that a tear or redundancy of the RI capsule was 1 of the lesions identified in patients with glenohumeral instability. Furthermore, Karas et al15 reported that primary interval pathology was indicated by a patulous interval capsule. In the literature, we found no report describing its true incidence among patients with anterior glenohumeral instability. If the RI lesions are present in patients with glenohumeral instability, RI closure is expected to provide further stability as our results demonstrated. The present study had several limitations. First, simulated RI closure is different from actual RI closure because the RI capsule was cut and then imbricated. In addition, the capsular condition of the cadaveric specimens was different from that of patients with glenohumeral instability. We also applied 10 N, 20 N, and 30 N of load to the humeral head. Appling a uniform load to the humeral head was not physio-
758
Yamamoto et al
logic, although it made it easier to compare the stabilizing function of the RI closure. In conclusion, RI closure reduces anterior/posterior translation in adduction and anterior translation at 60° of abduction in the scapular plane. Based on the results of the present study, RI closure is expected to reduce remnant anterior/posterior instability and, thereby, improve the clinical outcomes of arthroscopic stabilization procedures. However, RI closure should be carefully considered in overhead throwing athletes, because it decreases the range of external rotation and horizontal abduction.
J Shoulder Elbow Surg November/December 2006
15. 16. 17. 18. 19. 20.
REFERENCES
1. An KN, Jacobsen MC, Berglund LJ, Chao EYS. Application of a magnetic tracking device to kinesiologic studies. J Biomech 1988; 21:613-20. 2. Angelo RL. Controversies in arthroscopic shoulder surgery: arthroscopic versus open Bankart repair, thermal treatment of capsular tissue, acromioplasties—are they necessary? Arthroscopy 2003; 19:224-8. 3. Bassett R, Browne A, An KN, Morrey BF. A biomechanical analysis of shoulder muscles in the position of instability. J Biomech 1990;23:405-7. 4. Benedetto KP, Glotzer W. Arthroscopic Bankart procedure by suture technique: indications, technique, and results. Arthroscopy 1992;8:111-5. 5. Clark JM, Harryman DT II. Tendons, ligaments, and capsule of the rotator cuff. Gross and microscopic anatomy. J Bone Joint Surg Am 1992;74:713-25. 6. DePalma AF, Callery G, and Bennett GA. Shoulder joint: variational anatomy and degenerative regions of the shoulder joint. Instr Course Lect 1949;6:255-81. 7. Dempster WT. Mechanism of shoulder movement. Arch Phys Med Rehablil 1965;46:49-70. 8. Field LD, Warren RF, O’Brien SJ, Altchek DW, Wickiewicz TL. Isolated closure of rotator interval defects for shoulder instability. Am J Sport Med 1995;23:557-63. 9. Gartsman GM, Roddey TS, Hammerman SM. Arthroscopic treatment of antero-inferior glenohumeral instability. J Bone Joint Surg Am 2000;82:991-1003. 10. Gartsman GM, Taverna E, Hammerman SM. Arthroscopic rotator interval repair in glenohumeral instability: description of an operative technique. Arthroscopy 1999;15:330-2. 11. Grana WA, Buckley PD, Yates CK. Arthroscopic Bankart suture repair. Am J Sports Med 1993;21:348-53. 12. Harryman DT II, Sides JA, Harris SL, Matsen FA III. The role of the rotator interval capsule in passive motion and stability of the shoulder. J Bone Joint Surg Am 1992;74:53-66. 13. Itoi E, Motzkin NE, Browne AO, Hoffmeyer P, Morrey BF, An KN. Intraarticular pressure of the shoulder. Arthroscopy 1993;9: 406-13. 14. Itoi E, Berglund LJ, Grabowski JJ, Naggar L, Morrey BF, An KN. Superior-inferior stability of the shoulder: role of the coracohu-
21. 22. 23. 24. 25.
26. 27. 28. 29. 30. 31. 32. 33. 34. 35.
meral ligament and the rotator interval capsule. Mayo Clin Proc 1998;73:508-15. Karas SG. Arthroscopic rotator interval repair and anterior portal closure an alternative technique. Arthroscopy 2002;18:436-9. Kuhn JE, Bey MJ, Huston LJ, Blasier RB, Soslowsky LJ. Ligamentous restraints to external rotation of the humerus in the late-cocking phase of throwing. Am J Sports Med 2000;28:200-5. Kumar VP, Balasubramanian P. The role of atmospheric pressure in stabilizing the shoulder. An experimental study. J Bone Joint Surg Br 1985;67:719-21. Matsen FA III, Kirby RM. Office evaluation and management of shoulder pain. Orthop Clin North Am 1982;13:453-75. Matthews LS, Vetter WL, Oweida SJ, J Spearman, DL. Arthroscopic staple capsulorrhaphy for recurrent anterior shoulder instability. Arthroscopy 1988;4:106-11. Morgan CD, Bodenstab AB. Arthroscopic Bankart suture repair: technique and early results. Arthroscopy 1987;3:111-22. Morrey BF, Itoi E, An K-N. Biomechanics of the shoulder. In: Rockwood CA Jr, Matsen FA III, editors. The shoulder. 2nd ed. Philadelphia: WB Saunders; 1998, p. 233-76. Nobuhara K, Ikeda H. Rotator interval lesion. Clin Orthop 1987; 223: 44-50. O’Connell PW, Nuber GW, Mileski RA, Lautenschlager E. The contribution of the glenohumeral ligaments to anterior stability of the shoulder. Am J Sport Med 1990;18:579-84. Ovesen J, Nielsen S. Experimental distal subluxation in the glenohumeral joint. Arch Orthop Trauma Surg 1985;104:78-81. Pagnani MJ, Warren RF, Altchek DW, Wickiewicz TL, Anderson AF. Arthroscopic shoulder stabilization using transglenoid sutures: a four-year minimum follow-up. Am J Sports Med 1996;24: 459-67. Poppen NK, Walker PS. Normal and abnormal motion of the shoulder. J Bone Joint Surg Am 1976;58:195-201. Selecky MT, Tibone JE, Yang BY, McMahon PJ, Lee TQ. Glenohumeral joint translation after arthroscopic thermal capsuloplasty of the rotator interval. J Shoulder Elbow Surg 2003;12:139-43. Sidles JA, Larson RV, Garbini JL, Downey DJ, Matsen FA III. Ligament length relationships in the moving knee. J Orthop Res 1988;6:593-610. Speer KP, Warren RF, Pagnani M, Warner JJ. An arthroscopic technique for anterior stabilization of the shoulder with a bioabsorbable tack. J Bone Joint Surg Am 1996;78:1801-7. Stokes DA, Savoie FH, Field LD. Arthroscopic repair of anterior glenohumeral instability and rotator interval lesions. Orthop Clin North Am 2003;34:529-39. Taverna E, Sansone V, Basttistella F. Arthroscopic rotator interval repair: the three-step all-inside technique. Arthroscopy 2004;20: 105-9. Treacy SH, Field LD, Savoie FH. Rotator interval capsule closure. An arthroscopic technique. Arthroscopy 1997;13:103-6. Turkel SL, Panio MW, Marshall JL, Girgis FG. Stabilizing mechanisms preventing anterior dislocation of the glenohumeral joint. J Bone Joint Surg Am 1981;63:1208-17. Walch G, Boileau P, Levigne C, Mandino A, Neyret P, Donell S. Arthroscopic stabilization for recurrent anterior shoulder dislocation: results of 59 cases. Arthroscopy 1995;11:173-9. Warner JJ, Deng XH, Warren RF, Trozilli PA. Static capsuloligamentous restraints to superior-inferior translation of the glenohumeral joint. Am J Sport Med 1992;20:675-85.