Revision Arthroscopic Stabilization

Craig R. Bottoni, MD and Zackary Johnson, MD

The successful arthroscopic treatment of shoulder instability remains a challenge in the young athlete, most notably in those who participate in contact or overhead sports. The wide range of motion in the shoulder allows an athlete to perform many incredible feats, but it also comes at a cost. This freedom of motion can make the shoulder prone to instability, which can be disabling to the athlete. Traditionally, instability was treated using an open technique allowing for direct visualization and suture fixation of the disrupted labrum (Figure 23-1). With the introduction of arthroscopy, the standard of care for the treatment of anterior glenohumeral instability slowly transitioned to arthroscopic stabilization. Although the improvement in arthroscopic techniques has resulted in equivalency to traditional open techniques, recurrent glenohumeral instability does occur. The question of what may contribute to the failure of the primary arthroscopic procedure and subsequently, what are the options that should be considered prior to proceeding with a revision stabilization, are the focus of this chapter.

A patient who presents following a failed arthroscopic stabilization presents a challenging problem requiring thoughtful evaluation. Considerable debate exists about how to manage a failed arthroscopic shoulder stabilization. More conventional belief advised surgical treatment via an open stabilization for a failed arthroscopic stabilization. This advice has recently been challenged, and more surgeons are considering a repeat arthroscopic technique for a patient who has had a failed arthroscopic stabilization. Revision arthroscopic stabilization has a higher failure rate than primary repair, but this can be mitigated by evaluation and planning as well as proper technique. Understanding the cause of a failed stabilization is key to the revision. There are multiple reasons for failure, and each may contribute to the overall process. This chapter will review the potential etiologies of failure, recommend how to properly evaluate a failed stabilization, and then suggest options when an arthroscopic revision surgery may be indicated.

ARTHROSCOPY AND BANKART

Recurrent anterior glenohumeral instability has always been a significant problem for the athlete, especially in sports requiring overhead movements. Traditionally, instability was addressed through an open operative procedure. This slowly changed in the 1990s, when shoulder arthroscopy became more popular. Initially, arthroscopic stabilization of recurrent anterior shoulder instability resulted in much higher failure rates than what had been reported following open techniques. Most likely the higher failure rates were due to limited understanding of the pathoanatomy of the shoulder following a dislocation, limited availability of arthroscopic equipment, and minimal experience in the techniques used to perform arthroscopic Bankart repairs. When arthroscopic techniques were first attempted, only nonanatomical repairs were performed. Walch et al¹ reported on the Morgan transglenoid suture arthroscopic technique in 1995 (Figure 23-2). In their series, they had a 49% rate of poor results with the same number reporting recurrent dislocation or subluxation. Grana similarly reported a recurrence rate of 44% with the transglenoid suture technique.² In comparison to the generally accepted success rate of greater than 90% following open procedures at the time, many authors strongly opined against arthroscopic stabilization.¹ Other arthroscopic techniques employed to address anterior instability had disappointing and unacceptably high failure rates. Lane reported on a 33% recurrent instability rate using the arthroscopic staple capsulorrhaphy in 1993, and noted significant complications with this procedure.3 As arthroscopic techniques evolved and implants, primarily suture anchors, improved over time, the arthroscopic stabilization soon became the standard of care. Multiple studies since then have found the arthroscopic Bankart to have outcomes equivalent to that of open techniques. Tjoumakaris and colleagues retrospectively compared open and arthroscopic techniques in 2006 and found equivalent patient-reported outcome scores and recurrent instability.⁴ Bottoni et al looked prospectively at open vs arthroscopic stabilization and found equivalent failure rates and noted an increased loss of shoulder motion in the open group.⁵ Improved outcomes with arthroscopic techniques were likely due to a better understanding of the pathoanatomy associated with recurrent anterior shoulder instability, improved arthroscopic equipment allowing better visualization of the shoulder joint, and greater surgeon experience in treating this condition.

Figure 23-1. Open Bankart repair using deltopectoral approach, which was traditionally standard of care for instability.

Despite the advancements in arthroscopy and the benefits conferred, recurrent shoulder instability can still result following arthroscopic stabilization. The reported recurrence rate for an arthroscopic Bankart is 10% to 15% when using modern suture anchor techniques. However, specific cohorts of patients have been identified that had higher rates of failure following arthroscopic stabilization. Specifically, overhead athletes, contact athletes, those with bone loss of the humerus and/or glenoid, patients with generalized ligamentous laxity, and younger patients were all groups or findings reported to be associated with significantly higher failure rates following arthroscopic techniques employed to address their anterior shoulder instability. Despite the growing popularity of arthroscopy and improving results in primary instability, open operative techniques were usually recommended for failed arthroscopic stabilization procedures. Nevertheless, revision arthroscopic techniques have been reported with good results. In 2009, Boileau et al looked at a series of 22 patients who underwent a revision arthroscopic procedure after previous failed open stabilization and found an 85% rate of good or excellent results.⁶ Barnes and colleagues evaluated 18 shoulders that underwent a revision arthroscopic procedure after a failed open or arthroscopic stabilization and found 94% remained stable at a mean follow-up of 38 months.⁷

Figure 23-2. Early arthroscopic stabilization technique using transglenoid sutures.

A crucial factor to success in a revision shoulder stabilization is understanding why the original surgery may have failed. Carefully evaluating the probable etiology for the recurrent instability and then formulating a plan to address those issues are the keys to success. Sir Winston Churchill said, “Those who fail to learn from the past are doomed to repeat it.” This is also true of the revision arthroscopic stabilization. Understanding the reasons why an arthroscopic stabilization has failed is the first step, followed by a plan to address the new or previously unrecognized pathology. Boileau et al evaluated the factors contributing to failure of arthroscopic stabilizations. In their study of 91 consecutive arthroscopic stabilizations, they found a relatively typical rate of failure at 15%. Of these they found the strongest risk factors for failure were bone loss either on the glenoid or humeral side followed by the number of anchors used in fixation, which suggests a less than optimal original technique.⁸ Other factors implicated in the failure of shoulder stabilization are other soft-tissue injuries that were not addressed such as an anterior labroligamentous periosteal sleeve avulsion (ALPSA), glenolabral articular disruption (GLAD), a humeral avulsion of the glenohumeral ligament (HAGL) lesion, or unrecognized posterior instability.

Figure 23-3. Arthroscopic view of significant glenoid bone loss from a bony Bankart.

GLENOID BONE LOSS

Over the past decade the importance of glenohumeral bone loss in the glenoid and/or humerus has been cited as a potential etiology for failure of arthroscopic stabilization. Glenoid bone loss can occur for a variety of reasons, but is due to a single traumatic glenohumeral dislocation or respective injuries. The initial glenohumeral dislocation can often result in a bony avulsion from the anteroinferior glenoid along with the attached labrum, known as a bony Bankart (Figure 23-3). The size of this avulsion can vary, but in all cases the result is a more narrowed glenoid tract on which the humerus articulates. Bone loss can also occur because of repetitive episodes of instability. These recurrent subluxations can erode glenoid bone over time, also resulting in a narrowed glenoid tract.

Evaluating the glenoid and any bone loss is essential to planning a revision arthroscopic Bankart because it can affect technique in the revision, change the arthroscopic procedure, or disqualify the patient from an arthroscopic revision procedure. The identification of and subsequent quantification of the amount of glenoid bone loss remain a challenge. The best imaging modality and the technique to quantify glenoid bone loss remain controversial.

Imaging

Evaluation of glenoid bone loss can be accomplished through a variety of means. Traditional 3-view shoulder radiographs may raise suspicion of bone loss, but typically do not allow for any quantification of the amount of glenoid bone missing. Garth et al described the apical oblique view.⁹ This view is obtained by having the patient seated with the uninjured shoulder angled 45 degrees away from the cassette, thus putting the beam parallel to the articular surface of the glenoid and the normally anteverted shoulder. The beam is also angled 45 degrees from cephalad to caudal. When performed correctly, the coracoid should appear as a ring. The authors demonstrated the ability to see larger bony Bankart lesions as well as Hill-Sachs lesions better than standard 3-view radiographs, but smaller attritional bony defects were not well visualized. The West Point view, originally described by Rokous and colleagues, is a modified axillary view¹⁰ (Figure 23-4). The patient is positioned prone with the affected shoulder bumped approximately 3 inches off the table. The arm is abducted 90 degrees with the forearm hanging off the table. The beam is directed down 25 degrees from the horizontal and 25 degrees medially, again attempting to parallel the glenoid articular surface. Finally, the Didiée view is performed by again having the patient lie prone with the affected arm abducted and the hand on the patient’s back resting on the iliac crest. The beam is directed at a 45-degree angle from lateral to medial.¹¹ The authors found this to be an effective view for evaluating a bony Bankart but not a Hill-Sachs lesion. Again, although these views can be useful and clue a provider in to the presence of glenoid bone loss, quantifying the size of the lesion requires advanced imaging. They also can be technically difficult to obtain, especially in patients with abnormal glenoid version (Table 23-1).

Figure 23-4. West Point view showing glenoid bone loss.

Computed tomography (CT), especially with computer-rendered 3-dimensional (3D) reconstruction, has become the gold standard for evaluating and quantifying glenoid bone loss. Two-dimensional CT can also be limited in accurately quantifying glenoid bone loss because of variability in shoulder anatomy, most notably glenoid version, which will require adjustment of the imaging orientation. CT imaging of the glenoid, and subsequently the accurate quantification of bone loss, depends on the proper orientation of the patient’s shoulder and specifically, the glenoid in the CT gantry. This is solely operator dependent, and slight misalignment will dramatically change the bone loss determination. The CT images can be reconstructed via software to allow for a 3D visualization of the glenoid. Three-dimensional CT provides the best technique for evaluation of bone loss because it allows for humeral subtraction and an isolated image of the glenoid. In a patient who has failed an initial arthroscopic stabilization and bone loss is suspected, a CT should be obtained, if possible, with 3D reconstructions, to properly evaluate bone loss prior to consideration of a revision stabilization (Table 23-2; Figure 23-5).

Table 23-1. Specialized Radiograph Views for Visualizing Bone Loss

Radiographs may be used for detection but not for quantification.

Table 23-2. Methods for Calculating Bone Loss Using Computed Tomography

METHOD	DESCRIPTION
Superimposed circle method	Best-fit circle of injured and uninjured shoulder superimposed and area compared to determine % of bone loss.
Pico method	Uninjured shoulder best-fit circle (A) superimposed on injured shoulder to determine defect size. Calculated as % of bone loss (D/A × 100%).
Ratio method	Measure radius of best-fit circle (R) and the distance from center to anterior lesion (d). Use ratio d/R to determine % of bone loss based on author’s table.
Distance from bare area method	Measure distance from bare area to anterior lesion (A) and from bare area to posterior glenoid (B).
	% bone loss = ([B − A]/2B) × 100.
Surface area method	Best-fit circle of injured shoulder only. Use digital measuring means to determine % of bone loss.
Bankart length method	Best-fit circle drawn on injured shoulder. Radius of circle (R), length of osseous lesion measured (x). If x > R, then dislocation resistance is < 70% of uninjured shoulder.

Three-dimensional reconstructions are best for measurements when available.

There have been several techniques described for quantifying glenoid bone loss using CT with 3D reconstruction.12 Most techniques rely on the assumption that the inferior two-thirds of the glenoid creates a true circle with the bare spot roughly in the center.¹³ Chuang and colleagues described the superimposed circle method and the Pico method respectively, both of which rely on imaging of the contralateral shoulder and on comparing the area of fitted circles to the injured and uninjured glenoid to calculate glenoid bone loss¹⁴ (Figure 23-6). These methods can expose the patient to additional radiation because of the need to image the other shoulder. Other methods such as the ratio method described by Barchilon and the anteroposterior distance from bare area method described by Sugaya rely on distance rather than area.¹⁵ Both methods rely on the assumption that the bare spot can be roughly approximated on imaging by locating the intersection of a line drawn down the long axis of the glenoid and a line drawn horizontally through the widest point of the glenoid. The ratio method calculates bone loss by dividing the distance from the bare spot to the anterior lesion by the radius of the best-fit circle centered on the bare spot. The distance from the bare spot method is performed by measuring the distance from the bare spot to the posterior edge of the glenoid and distance from the bare spot to the anterior lesion and using these to calculate the percentage of bone loss. The surface area method as described by Sugaya et al 15 uses digital software to measure the area of the best-fit glenoid circle and the area of the defect to calculate the percentage of bone loss. Alternatively, Gerber and Nyffeler found that if one measured the length of the osseous Bankart lesion and found it to be greater than the radius of the best-fit circle, the glenoid will have 70% or less of its original resistance to dislocation.¹⁶

Figure 23-5. Bony Bankart noted on axial cuts of a shoulder CT. (Reprinted with permission from Kyong Su Min, MD.)

Classification and Treatment

Because glenoid bone loss is a common explanation for failure of arthroscopic stabilization, it should be evaluated and, if present, quantified prior to any revision surgery to consider all options, but techniques have also been described to measure bone loss arthroscopically. This can be useful in confirming a surgical plan intraoperatively. To measure bone loss during shoulder arthroscopy, a graduated probe is inserted from the posterior portal and used to measure from the bare spot to the anterior edge of the glenoid, and then to the posterior aspect of the glenoid. Assuming the bare spot is in the center of the glenoid, bone loss can then be calculated. Bone loss will be overestimated if more inferior to the horizontal axis of the inferior two-thirds of the glenoid, but this is uncommon. Also, some authors have suggested that the bare spot is typically slightly anterior to the true center of the glenoid or, at times, not present at all. Barcia and colleagues evaluated the glenoid bare spot and found that the bare spot was visible only 48% of the time and centered only 37% of the time¹⁷ (Figure 23-7). Therefore, this technique although confirmatory, is not recommended to make treatment decisions.

Figure 23-6. Best fit circle of lower 2/3 of the glenoid demonstrating bone loss.

Once the amount of bone loss has been quantified, the next step is determining the appropriate procedure to address the recurrent instability. “Critical” glenoid bone loss thresholds have been reported, but the cutoff where a bone augmentation procedure is recommended is still unclear. Traditionally, 20% to 25% bone loss has been used as a threshold for arthroscopic stabilization. Burkhart et al described the inverted-pear morphology of the glenoid, which he later specified to represent an average of 28% bone loss, had 67% recurrence rate after arthroscopic anterior stabilization.^18,19 Conversely, Porcellini and colleagues found that 92% of patients with bone loss of 25% or less who underwent arthroscopic stabilization demonstrated a stable shoulder at 2-year follow-up.²⁰ It should be noted, however, that this study looked at acute injuries. A lower threshold of glenoid bone loss may be necessary in recurrent shoulder instability because they have already demonstrated failure. Shaha et al reexamined critical bone loss and failure rates in terms of outcomes as opposed to recurrent instability. They found that bone loss between 13.5% to 20%% led to worse patient-reported outcomes.²¹ Provencher et al suggested an algorithm in their review article in 2010, that 0% to 15% bone loss may be treated arthroscopically.¹² If a small avulsion fragment of the glenoid, called a bony Bankart lesion, is present and unaddressed, it may be another source of failure. Even in small bony lesions, incorporation of the fragment in the repair, if possible, can help with healing especially in the revision setting (Figures 23-8A and 23-8B). However, in the acute injury, bone fragments are often easier to mobilize and fixate to the native glenoid. In revisions, the quality and size of the remaining bone fragment is variable. Careful preoperative evaluation of the fragment itself is important. If a large bony fragment is present that is either malunited or has a fibrous union, it must be ascertained whether it can be mobilized and incorporated into the repair. Should this be the case, careful arthroscopic preparation of the fragment should be performed to ensure a bony union when this is incorporated into the repair. In the revision setting, however, bone loss can often be attritional. Any remaining bony fragment may either be malunited so it cannot be mobilized or may have undergone significant resorption. In this case, a bone augmentation procedure may be indicated.