Michael J. Pagnani, MD and Jason A. Jones, MD
Open stabilization techniques have had a long history of effectiveness in reducing anterior shoulder instability. The results of such procedures have been almost uniformly excellent, with postoperative recurrence rates generally reported between 0% and 5% in nonselected populations.1–8 Yet, owing largely to the popularity of arthroscopic methods of stabilization,9 open capsular repairs have been largely neglected in the training of younger orthopedic surgeons. Although arthroscopic repairs have advantages (including smaller incisions, less trauma to the subscapularis tendon, less perioperative pain, easier rehabilitation, and more predictable return of motion), they have been plagued by unacceptably high recurrence rates in high-risk groups such as contact athletes, patients with capsular laxity, and those with bone loss of the humeral head or glenoid fossa.10–19 Balg and Boileau10 devised an Instability Severity Index Score that recommended that arthroscopic methods of stabilization be avoided in most of these high-risk patients. It is now commonly recommended that such at-risk patients be treated with a bone-augmentation technique such as the Latarjet procedure. In our opinion, such an algorithm is essentially going from “point A to point C.” These guidelines ignore “point B”—open capsular repair.
RATIONALE FOR OPEN CAPSULAR REPAIR
The hope that the results of “modern” techniques for arthroscopic stabilization for anterior shoulder instability would approximate those reported for open capsular repair has not been realized; recurrence rates for arthroscopic repair have been shown in recent meta-analyses to continue to exceed the historical rates for open stabilization.20,21 Hohmann et al, in a recent systematic review, noted that results reported in the literature for arthroscopic stabilization between 2005 and 2015 had not improved statistically when compared to results reported between 1995 and 2004.21 Alkaduhimi and colleagues,20 in another systematic review published in 2016, concluded, “Despite advances in surgical techniques and devices during the last 20 years, … the recurrence rate for arthroscopic shoulder stabilization has only marginally decreased.”
These higher failure rates have persisted despite careful patient selection by which many arthroscopic studies exclude high-risk groups such as contact athletes, patients with bony defects of the humeral head and glenoid, and patients with capsular laxity.11,15,22–24 The preponderance of highly selected patients in reports of the results of arthroscopic stabilization makes comparison with the outcomes of open stabilization without such exclusions problematic. In addition, the dearth of case series on the results of open stabilization from experienced surgeons at leading centers of shoulder surgery over the past 20 years does not permit adequate assessment of “modern” open techniques for such comparisons. However, in this era of evidence-based medicine, no fewer than 7 separate meta-analyses have concluded that the results of open stabilization have been superior to those of arthroscopic stabilization.15,20,21,25–28 In fact, Hohmann et al21 found a 37% higher risk of recurrent instability with arthroscopic techniques compared to open methods.
Open stabilization has several advantages over arthroscopic repair that may well explain the differences in recurrence rates: 1) Open methods allow the surgeon to completely free the capsule from the subscapularis tendon to precisely tension the capsule without adherence to the subscapularis; 2) the rotator interval can be better visualized and tensioned via an open technique; open rotator interval closure does not have the same effect on shoulder translation or rotation as arthroscopic closure29,30; 3) the thickness of the capsule can be doubled by overlapping the capsule during open repair; 4) the arm can be optimally positioned for open repair; 5) open techniques allow for repair of the capsule and labrum with knots tied extra-articularly, eliminating concern about suture impingement on the articular surfaces of the shoulder.
The general indications for surgical treatment of recurrent anterior shoulder instability are highly subjective. They include a desire of the patient to avoid recurrent episodes of instability (including the necessity of reporting to the emergency room on a frequent basis to have the shoulder reduced), problems with recurrent pain, an inability to perform certain activities because of apprehension regarding the shoulder, and the desire to improve athletic performance with improved shoulder stability. Failure of a thorough trial of nonoperative treatment is also an indication for surgical treatment.
Indications for open stabilization over arthroscopic stabilization include 1) participation in a contact or collision sport, 2) male patients younger than 20 years with instability, 3) small to moderate bony defects of the humeral head or glenoid, 4) humeral avulsion of the glenohumeral ligaments, 5) failed arthroscopic repair, and 6) atraumatic instability. Essentially patients with an Instability Severity Index Score of 6 or greater (in which arthroscopic methods have a reported failure rate of 70%),10 are candidates for open capsular repair. In our practice, such patients represent approximately 80% of the patients who fail conservative treatment and require surgical treatment for anterior instability. We will consider 2 of these indications in the following sections.
The results of arthroscopic stabilization in contact athletes have been generally disappointing, with failures rates ranging from 14% to 60%.12,14–16,19,20,24 Recently, a report from the Hospital for Special Surgery revealed a 26% recurrence rate after arthroscopic repair.19 Contemporary systematic reviews have noted an 18% failure rate in contact/collision athletes15 and an 8 times greater absolute risk for recurrence in collision athletes compared to other patients after arthroscopic repair.20 Okoroha and colleagues31 reported a 26% recurrence rate in National Football League players after operative repair.
In contrast, our series of open stabilization in 58 American football players yielded a 3% recurrence rate, with 2 athletes sustaining postoperative subluxation and no postoperative dislocations.4
Bony Defects of Humeral Head/Glenoid
High recurrence rates have been reported after arthroscopic Bankart repair in patients with bony defects of the glenoid and humeral head.11,12 Burkhart and De Beer12 reported that contact athletes who had an “engaging” Hill-Sachs lesion or “inverted pear” glenoid had a recurrence rate of 89%. Their subdivision of Hill-Sachs lesions into “engaging” and “non-engaging” types is commonly used in contemporary lectures on shoulder instability to determine the risk of postoperative instability after arthroscopic repair. More recently, the description of “on-track” and “off-track” lesions has come into common usage. The use of these terms (and the impression of high rates of failure after operative treatment) has been erroneously assigned to open capsular repair as well. In fact, a review of the published results of traditional open capsular repair in the face of bony deficiency consistently reveals outcomes that have been more than acceptable. Rowe et al,5 in their historic 1978 end-result study of open Bankart repairs, found that postoperative recurrence actually decreased, from 3.5% to 2%, in patients with defects of the glenoid rim. Bigliani and colleagues32 reported a 12% recurrence rate after open capsular shift in patients with glenoid bone loss. Rowe et al5 found only a slight increase in recurrence after open Bankart repair in patients with moderate or severe Hill-Sachs lesions (5% vs 3.5%). Gill et al1,2 found a recurrence rate of 6% in the presence of a large Hill-Sachs lesion.
We have assessed recurrence rates with a contemporary method of open anterior stabilization without bony augmentation in patients in our practice with defects of the glenoid and/or humeral head.3 The overall recurrence rate was 2% (2/103), with one patient having a postoperative dislocation and the other experiencing subluxation. There were no recurrences in the 14 patients with glenoid rim deficiency. Patients with engaging Hill-Sachs lesions had a 4% recurrence (1/28), but this was not statistically significant. One of the 9 patients (11%) with large defects of the humeral head had a recurrence—again, not statistically significant. These results suggest that very large bony defects of the humeral head may be more problematic for open capsular repairs than glenoid defects. Overall, however, the results in patients with bony defects are comparable to those reported for bone-block procedures such as the Latarjet.
The resurgence of interest in bone-block procedures such as the Latarjet has, in our opinion, significant pitfalls. The popularity of bony augmentation procedures of this type, which were largely abandoned in North America until relatively recently, raises several concerns: 1) unless performed in a modified form, they do not address capsular laxity or capsulolabral separation, 2) there is a high risk of complications from hardware loosening or nonunion, 3) revision surgery is difficult, and 4) there is a high incidence of postoperative arthrosis. Although there is little doubt that such procedures are effective at restoring stability, complication rates after the Latarjet can be extremely high. In an observational review of patients referred to our office for continued shoulder dysfunction after a Latarjet procedure at another institution,33 we noted that 21 of 27 patients had a nonunion of the transferred coracoid. Nineteen of the 21 also had screw breakage. (The fact that only 5 of the 21 patients with coracoid nonunion had instability complaints leads one to wonder what clinical role, if any, the bone block actually plays in restoring stability.) More than half the patients had radiographic evidence of osteoarthritis, and nerve injuries occurred in 6 patients with 2 axillary nerve injuries causing permanent disability. In addition, 3 patients had biceps pain and asymmetry and 1 had subscapularis insufficiency. Although this was a selected population that sought our assistance after problems with the procedure, it is clear that significant complications are not uncommon—especially when the Latarjet is performed by surgeons who are inexperienced with the technique.
In another study,3 we concluded that large defects of the humeral head and/or glenoid are uncommon—even in a tertiary referral shoulder practice with a large percentage of contact athletes. In a 6-year period, we encountered only 9 patients with large (> 4 cm long and 0.5 cm deep) humeral head defects and only 4 with large (> 20% of the glenoid diameter) defects of the glenoid.
Based on the low rates of recurrence, motion loss that was equal to or better than that reported for bone-block procedures, and the seemingly self-evident premise that the complication rate of capsular repair alone should be lower than that of capsular repair combined with bone augmentation, it appears that bone-block or grafting procedures are not necessary in the majority of patients with bony defects of the glenoid and/or humeral head if they are treated with contemporary techniques of open stabilization.
Contraindications to the open technique include voluntary instability and concomitant psychological disease. Large Hill-Sachs lesions or glenoid lesions may (in our opinion, rarely) require supplemental bone grafting to compensate for the defects.34 In our practice, such procedures are generally reserved for patients who have failed an attempt an attempt at open capsular repair—an uncommon situation.
We prefer to use arthroscopic methods of stabilization in athletes whose primary focus is throwing and in other overhead athletes who cannot accept any restrictions in postoperative motion. If an arthroscopic method is used in an overhead athlete, we strongly encourage the surgeon to have a frank discussion with the athlete and his or her family about the high risk of recurrent instability should the athlete return to contact sports. We subtly encourage the multisport athlete to prioritize the importance of his or her various athletic pursuits. If a contact sport is the priority, we recommend an open repair. In such a case, we do not hesitate to delay operative intervention until the postseason if the athlete has infrequent episodes of instability and can continue to play at a level that he or she finds acceptable. On the other hand, if an overhead sport is deemed to be primary, we encourage early arthroscopic repair. In the unusual case that an open method is used in this group (ie, after failed arthroscopic repair in an athlete who continues to prioritize an overhead sport), we use the technique described by Jobe and colleagues35 in which the subscapularis tendon is split rather than detached.
PATHOANATOMY OF ANTERIOR INSTABILITY
The shoulder has the greatest range of motion of any joint in the human body. Because bony restraints to motion are minimal, the surrounding soft tissue maintains the humeral head on the glenoid—yet the shoulder capsule is large, loose, and redundant. There are 3 main ligaments in the anterior capsule that help prevent subluxation or dislocation. These ligaments are known as the superior glenohumeral ligament (SGHL), the middle glenohumeral ligament (MGHL), and the inferior glenohumeral ligament complex (IGHLC). Damage to the IGHLC, which supports the inferior part of the shoulder capsule like a hammock, is related to most cases of anterior instability. The Bankart lesion, involving detachment of the IGHLC insertion on the glenoid, is the most common pathologic lesion associated with traumatic anterior instability. Defects or injuries to the SGHL and MGHL may also contribute to instability.36
The primary goals of the surgical treatment of shoulder instability should be to restore stability and to provide the patient with near-full, pain-free motion. Older techniques of open shoulder stabilization tended to limit shoulder range of motion in exchange for providing stability. Limiting shoulder motion often led to osteoarthritis; we now understand that it is probably more important to preserve motion than it is to stabilize the shoulder. As a result, any method of open stabilization should be designed to provide close to full range of motion of the shoulder as well as stability.
The diagnosis of an anterior shoulder dislocation is usually obvious. The patient typically gives a history of a traumatic injury in which the shoulder “popped out” and had to be reduced. Often, the arm is positioned in abduction and external rotation at the time of the episode. In some cases, however, dislocation can occur with no history of significant trauma. These latter patients are frequently noted to have generalized ligamentous laxity and multidirectional instability and are less likely to demonstrate a Bankart lesion. Such patients typically have enlargement of the rotator interval and loose capsular tissue.
The diagnosis of anterior subluxation can be more subtle. The chief complaint may be a sense of movement, pain, or clicking with certain activities. Pain, rather than instability, may be the predominant complaint. The pain commonly localizes to the bicipital groove area and to the infraspinatus fossa, possibility because of compensatory overload of the dynamic stabilizers of the shoulder.
Apprehension tests are designed to induce anxiety and protective muscular contraction as the shoulder is brought into a position of instability. The anterior apprehension test is performed with the arm abducted and externally rotated. As the examiner progressively increases the degree of external rotation, the patient develops apprehension that the shoulder will “slip out.” This test is uniformly positive in patients with anterior instability.
During the relocation test, the examiner’s hand is placed over the anterior shoulder of the supine patient. A posteriorly directed force is applied with the hand to prevent anterior translation of the humeral head. The shoulder is then abducted and externally rotated as it is in the apprehension test. A positive result is obtained when this anterior pressure allows increased external rotation and diminishes associated pain and apprehension. The relocation test seems to be more reliable in overhead athletes and may not be positive in all cases of anterior instability.
The belly press and lift off tests should also be performed to confirm the integrity of the subscapularis tendon.
Routine radiographic examination of the unstable includes an anteroposterior (AP) view (deviated 30 to 45 degrees from the sagittal plane to parallel the glenohumeral joint), a trans-scapular (Y) view, and an axillary view. West Point and Stryker Notch views are helpful in demonstrating bony lesions of the humeral head and glenoid.
Magnetic resonance imaging (MRI) is useful to determine whether a Bankart lesion is present and also to assess patients for evidence of concomitant rotator cuff or superior labral pathology. The accuracy of MRI in determining labral pathology is, in our experience, increased with arthrography. Because of the possibility of concomitant rotator cuff injury, MRI should always be considered in older patients with instability—especially if strength and motion are slow to recover after a traumatic episode.
Computed tomography (CT) scans may be indicated if bony deficiency is suspected on plain films. However, the surgeon should be cautioned that that CT tends to overestimate the size of larger glenoid lesions and that CT measurement of smaller lesions is not superior to arthroscopic measurement. These phenomena have been previously noted by others.37,38
TIMING OF OPEN SURGICAL REPAIR
As mentioned earlier, we do not hesitate to delay open surgical treatment of anterior instability until an athlete completes his or her competitive season. Although there is increasing evidence that the results of arthroscopic stabilization are less satisfactory when surgery is delayed or after multiple recurrences,39,40 delaying open repair has had no such negative impact on recurrence rates in our experience.
Our basic procedure for the open surgical treatment of recurrent anterior instability is a modification of the Bankart procedure and involves repair of the anterior capsule and labrum to the glenoid. In most cases, the capsular ligaments are stretched as well as detached, and the procedure is also designed to remove any abnormal laxity.
The procedure is performed after preoperative placement of an interscalene block and an inter-scalene catheter for postoperative analgesia. In most cases, the block is supplemented with general laryngeal mask airway anesthesia. In properly selected patients, the procedure may be performed using regional anesthesia alone.
We routinely perform a complete arthroscopic examination before proceeding with the open procedure. The patient is placed in the beach-chair position with the head elevated to 60 degrees. Before elevation of the head, an armboard is attached to the operative side of the table and folded sheets are taped to the armboard. The armboard is then rotated against the side of the table to provide arthroscopic access to the shoulder until conversion to the open procedure. A mechanical arm holder is used to control shoulder position during arthroscopy. The shoulder is examined through standard anterior and posterior arthroscopic portals. Supplementary portals are created if indicated based on the findings of labral or rotator cuff pathology. Any loose bodies are removed during arthroscopy. When superior or (less commonly) posterior labral tears accompany anterior instability, they are repaired arthroscopically because access to these parts of the labrum is not possible with an open anterior approach. Similarly, supraspinatus or infraspinatus pathology is addressed with arthroscopic treatment before the open procedure. When sizable Hill-Sachs lesions are noted in patients at high risk for recurrent instability, we now commonly perform arthroscopic “remplissage” to cover the defect prior to the open repair.41 In addition, the arthroscopic examination is helpful in determining the specific anterior pathology prior to the open repair.