While the outcomes of primary stabilization procedures for glenohumeral instability are highly successful, persistent instability after surgical reconstruction presents a unique challenge for surgeons to diagnose and manage. In general, the results of procedures for traumatic anterior instability are superior to those for posterior and multidirectional instability. The reasons for failure include: technical errors during the primary surgery, incorrect diagnosis, bone defects, soft-tissue defects, and patient noncompliance.

Patient evaluation should include history, physical examination, and plain radiographs. Computed tomography (CT), magnetic resonance imaging (MRI), examination under anesthesia (EUA), and intraoperative findings add useful information as well. Important factors obtained from the history include inciting event, symptoms, previous surgical procedure, and operative findings. Instability that recurs after a discrete traumatic injury may be handled differently than instability that develops gradually with minimal or no trauma. In the traumatic group, instability may result from a new lesion, whereas a gradual recurrence may indicate improper patient selection, technical errors from the previous surgery, or incorrect diagnosis and surgery. Many patients will have symptoms of continual instability, most obvious in those with dislocations requiring reduction. However, in some cases pain will supersede any sensations of instability.

In an attempt to reduce confusion in classifying instability, Kuhn et al.¹¹² recently developed the frequency, etiology, direction, and severity (FEDS) system for glenohumeral instability. Frequency was defined as solitary (1 episode), occasional (2 to 5 times/year), or frequent (>5 times/year). Etiology was defined as traumatic or atraumatic. Direction referred to the primary direction of instability (anterior, posterior, or inferior). Severity was subluxation or dislocation. The FEDS system has content validity and is highly reliable for classifying glenohumeral instability. Physical examination using provocative testing to determine the primary direction of instability produces very high levels of interrater and intrarater agreement.

It is often helpful to review past operative reports. This gives information about the procedure performed plus any pathologic lesions which were identified and addressed at the time of surgery. This knowledge can assist the revision surgeon in choosing the correct treatment option. For example, continued instability after an anterior surgery could result from an initial failure to diagnose posterior instability.

Physical examination is important in differentiating causes of recurrent instability. Positive anterior apprehension and relocation test results indicate clinical failure. Concomitant posterior or multidirectional instability should be ruled out. Generalized ligamentous laxity should be recognized and worked up so that any collagen/vascular disorders are diagnosed, as this plays significantly into the type and timing of any revision surgery. Scapular mechanics should be assessed and dyskinesia corrected before considering surgery. Using a three-dimensional (3D) kinematic analysis of scapular motion, Uhl et al.²⁰⁷ showed that side-to-side asymmetries are common in symptomatic and asymptomatic populations. However, testing in flexion showed a higher frequency of multiple-plane scapular asymmetries in the symptomatic patients compared to asymptomatic controls. Scapular dyskinesis observed in symptomatic shoulders, especially with flexion, should be considered a potential factor contributing to the dysfunction.

While primary scapular winging may be due to neurologic injury, osseous injury, or periscapular soft-tissue abnormalities,
secondary scapular winging occurs as a result of glenohumeral and subacromial conditions such as instability, labral tears, rotator cuff injury, or tendonitis. Secondary winging resolves after the primary pathologic condition has been addressed.

Plain radiographs are useful in finding bone deficiencies associated with recurrent instability and to assess placement and migration of metallic anchors. CT scans are a valuable adjunct for quantifying large glenoid rim defects or Hill-Sachs lesions. MRI, particularly with intraarticular contrast, can identify recurrent Bankart lesions, superior labral tears, medially healed Bankart lesions, capsular avulsions, capsular redundancy, and rotator cuff tears.

EUA and intraoperative findings are used to verify the diagnosis at the time of repeat surgery. EUA will differentiate between anterior, posterior, and multidirectional instability and demonstrate any hyperlaxity. Diagnostic arthroscopy may be performed to evaluate superior labral tears, medially healed Bankart lesions, capsular redundancy, humeral avulsion of the glenohumeral ligament (HAGL), and rotator cuff pathology.

Postoperative rehabilitation should be maximized before considering revision surgery. This should include correction of soft-tissue contracture, rotator cuff weakness, and periscapular dysfunction.

Both open²,⁷,²⁶,⁴⁹,⁶⁸,⁶⁹ and ⁷⁰,⁸⁴,⁸⁵,⁸⁶ and ⁸⁷,¹⁵⁵,¹⁵⁶,¹⁵⁹,¹⁶⁵,¹⁷⁷,¹⁷⁹,¹⁸¹,¹⁸⁸,¹⁹⁴,¹⁹⁹,²⁰⁸,²¹¹,²²¹,²²⁷,²³⁰ and arthroscopic⁸,¹¹,¹⁶,¹⁹,²⁸,³¹,³⁵,³⁹,⁴³,⁴⁴,⁵¹,⁶²,⁷²,⁹⁶,⁹⁸,¹⁰⁸,¹¹³,¹¹⁴,¹²⁹,¹³⁰,¹³³,¹⁴⁵,¹⁴⁷,¹⁵³,¹⁶⁴,¹⁹⁶,¹⁹⁸,²⁰¹,²⁰²,²⁰⁵,²⁰⁹,²¹⁰,²¹⁵,²¹⁶,²²⁵ approaches have been shown to be successful in the treatment of anterior shoulder instability. Open techniques are considered the “gold standard”,⁷⁰ especially in patients with bony defects of the glenoid or humeral head , some revision situations, and contact athletes.⁷⁰,¹⁵⁵,¹⁷⁷,¹⁵⁶,²⁰⁸ These methods can be divided into “anatomical” and “nonanatomical” groups. The anatomical procedures include Bankart repairs and capsular shift procedures which attempt to restore the normal capsulolabral anatomy.⁶⁹,¹⁵⁵,¹⁵⁶,¹⁵⁹,¹⁶⁵,¹⁷⁷,¹⁷⁹,¹⁸¹,¹⁹⁴,¹⁹⁹,²⁰⁸,²²¹,²³⁰ Nonanatomical stabilization attempts to compensate for the pathologic anatomy.²,⁷,²⁶,⁴⁹,⁶⁸,⁸⁴,⁸⁵,⁸⁶ and ⁸⁷,¹⁸⁸,²¹¹,²²⁷ These operations include the Bristow, Latarjet, Trillat, Magnuson-Stack, and Putti-Platt procedures.

Open Bankart repairs have been shown to have recurrence of dislocation or subluxation ranging from 0% to 10%.⁶⁹,⁷⁰,¹⁵⁵,¹⁵⁶,¹⁵⁹,¹⁶⁵,¹⁷⁷,¹⁷⁹,¹⁸¹,¹⁹⁴,¹⁹⁹,²⁰⁸,²²¹,²³⁰ In reviewing 145 procedures, Rowe et al.¹⁷⁹ reported 97% good to excellent results and only a 2% rate of redislocation. While the presence of a glenoid rim fracture did not increase the risk of recurrence, a moderate to severe Hill-Sachs lesion increased the risk to 6%. After a mean follow-up of 12 years, in 56 Bankart repairs, Gill et al.⁶⁹ showed 93% good to excellent results and only 5% recurrent dislocation due to new traumatic events more than 3 years postoperatively. In a study of long-term outcomes of open Bankart repairs in 39 patients, Pelet et al.¹⁵⁹ found 10% recurrence of dislocation and 5 patients with symptomatic arthritis requiring shoulder arthroplasty at a mean 29 years follow-up.

The Putti-Platt¹⁵⁴ procedure shortens the anterior capsule and subscapularis tendon in a pants-over-vest fashion, while the Magnuson-Stack¹²⁷ transfers the subscapularis from the lesser tuberosity more lateral to the greater tuberosity. Both operations prevent instability by restricting external rotation of the shoulder, without correcting any underlying pathologic lesions. These surgeries have recurrence rates up to 20% with decreases in external rotation up to 30 degrees,⁷⁰ Neither procedure is recommended for primary or revision instability repair.

The Latarjet-Bristow procedure involves an osteotomy of the coracoid, with its tendon attachments intact, transfer and reattachment to the anterior-inferior glenoid rim. The transfer is done through a horizontal split of the subscapularis muscle.²¹¹ This provides stability through a “triple blocking” effect.¹⁵⁸ While securing the coracoid to the anterior-inferior glenoid rim provides a “bone block” (which can manage a glenoid bone deficiency), more of the stability from this procedure is likely supplied by the conjoined tendon sling which reinforces the inferior subscapularis and anterior-inferior capsule. The third mechanism of stability comes from repair of the capsule and inferior glenohumeral ligament to the stump of the coracoacromial (CA) ligament, emulating a capsulolabral reconstruction. Similar to open Bankart repairs, the recurrence rate following Latarjet-Bristow procedures ranges from 1% to 15%.²,⁷,²⁶,⁴⁹,⁶⁸,⁸⁴,⁸⁵,⁸⁶ and ⁸⁷,¹⁸⁸,²¹¹,²²⁷ Allain et al.² reviewed 58 shoulders at an average of 14.3 years follow-up. A total of 88% of the patients had excellent or good results. There were no recurrent dislocations, 2% occasional subluxations, and 10% with apprehension on examination. In a prospective series of 118 Latarjet procedures followed for 15 years, Hovelius et al.⁸⁶ had one patient undergo revision surgery as a result of recurrent instability. There were 3 patients with recurrent dislocations and 11 with occasional subluxations. Schroder et al.¹⁸⁸ reviewed 52 Bristow procedures in midshipmen at an average follow-up of 26 years. Overall, recurrent instability occurred in 8 of 52 shoulders (15%).

Initial results of arthroscopic techniques for anterior glenohumeral stabilization have shown recurrence rates from 0% to 49%.¹¹,¹⁹,⁴³,⁴⁴,⁵¹,⁷²,⁹⁶,¹¹³,¹²⁹,¹⁴⁷,¹⁵³,¹⁹⁶,²⁰⁵,²¹⁰,²¹⁵,²¹⁶ Methods of fixation include staples,⁴⁴,¹¹³ transglenoid sutu res,¹¹,⁷²,¹⁴⁷,¹⁵³,¹⁹⁶,²⁰⁵,²¹⁰ bioabsorbable tacks,¹¹,¹⁹,³⁹,⁴³,⁵¹,⁷²,⁹⁶,⁹⁸,¹²⁹,¹⁴⁷,¹⁵³,¹⁹⁶,²⁰⁵,²¹⁰,²¹⁵,²¹⁶,²²⁵ and suture anchors.⁸,¹⁶,²⁸,³¹,³⁵,³⁹,⁶²,⁹⁸,¹⁰⁸,¹¹⁴,¹³⁰,¹³³,¹⁴⁵,¹⁶⁴,¹⁹⁸,²⁰⁹,²²⁵ The high failure rates of earlier techniques were attributed to failure to anatomically correct the pathology, poor patient selection, and application of these techniques to various types of instability.

With advancement of arthroscopic techniques and implants, and properly selected patients, the results of arthroscopic anterior instability surgery have become more comparable to those of open techniques.⁸,¹⁶,²³,²⁸,³¹,³⁵,³⁸,³⁹,⁵²,⁵⁷,⁶²,⁶⁴,⁷⁷,⁸³,⁹⁸,⁹⁹,¹⁰⁸,¹¹⁴,¹¹⁸,¹³⁰,¹³³,¹⁴⁴,¹⁴⁵,¹⁶⁰,¹⁶⁴,¹⁷⁶,¹⁹⁸,²⁰¹,²⁰²,²⁰⁴,²⁰⁹,²¹³,²²⁵ The use of suture anchors in the treatment of anterior shoulder instability was first described by Wolf²²⁵ in 1993. With shortterm follow-up, there were no complications, and only one recurrence in 50 patients treated with arthroscopic labral repair. Kim et al.⁹⁸ retrospectively evaluated outcomes of arthroscopic repair of Bankart lesions using suture anchors. With 2- to 6-year follow-up there was a 4% recurrence rate including one dislocation, two subluxations, and four patients with positive apprehension. In a prospective study of 53 patients with minimum 2 years follow-up, Gartsman et al.¹⁰⁸ had 92% good to excellent results with only 4 patients with recurrent instability, 1 requiring revision surgery. The improved results, compared
to previous arthroscopic techniques, was attributed to repairing not only the Bankart lesions, but also addressing inferior and superior lesions with selective capsulorrhaphy and rotator interval repair, when needed. Burkhart and De Beer²⁸ found 21 failures (10.8%) in 194 patients with recurrent shoulder instability treated with suture anchors. Those without significant bone defects had a failure rate of 4% versus a 67% failure rate in those with bone defects. Boileau et al.¹⁶ demonstrated 14 out of 91 patients (15%) experiencing recurrent instability at mean follow-up of 36 months after arthroscopic Bankart repair. The risk of postoperative recurrence was significantly related to the presence of a humeral or glenoid bone defect, hyperlaxity, and the use of less than four anchor points. Voos et al.²⁰⁹ prospectively followed 83 patients undergoing arthroscopic Bankart repair with suture anchors for a minimum of 2 years. Recurrent dislocation or subluxations were seen in 18% of patients. Risk factors for recurrence included patient age under 25, ligamentous laxity, and large Hill-Sachs lesions.

In order to identify which patients are at risk for recurrent instability after arthroscopic Bankart repair, Balg and Boileau¹⁰ developed the instability severity index score. A prospective study of 131 patients who underwent arthroscopic Bankart repair revealed several risk factors for recurrence. These include age less than 20 years; participation in competitive, contact, or overhead sports; the presence of a Hill-Sachs lesion on external rotation AP radiograph; and loss of the sclerotic inferior glenoid contour on true AP radiograph. Patients receiving a score of greater than six out of ten points have a 70% risk of recurrence after arthroscopic Bankart repair.

Thermal capsulorrhaphy uses a laser or monopolar radiofrequency probe to heat and denature capsular collagen, causing these molecules to unwind and shorten. This in turn reduces capsular volume.⁸²,⁸³,¹⁰⁷ While initial short-term studies were generally favorable, mid-term and long-term studies found increasing failure rates ranging from 7% to 37%.⁴,³⁴,⁴²,⁵⁵,¹²¹,¹³⁴,¹⁴¹,¹⁵²,¹⁸⁴,²²⁶ Reports of complications following thermal capsulorrhaphy include rapid onset chondrolysis, axillary nerve damage, capsular laxity, and thinning of the capsular tissue.⁷¹,⁷³,¹¹⁹,¹³²,¹⁵⁷,¹⁹³ Thermal capsulorrhaphy has been abandoned due to these complications and unacceptably high recurrence rates. In 2007, the editors of Arthroscopy concluded that the risks of thermal capsulorrhaphy outweigh any potential benefits and recommended that the procedure no longer be performed.¹²³

Recently, Wolf and colleagues¹⁷¹ described the Remplissage technique as a solution for anterior instability associated with glenoid bone loss and a large Hill-Sachs lesion. The procedure consists of an arthroscopic capsulotenodesis of the posterior capsule and infraspinatus tendon to fill the Hill-Sachs lesion, along with arthroscopic Bankart repair. In a review of 24 patients treated with this technique, only 2 (7%) had recurrent instability after significant trauma. Zhu et al.¹⁰⁶ demonstrated similar results with a minimum 2-year follow-up. They report an 8% failure rate after performing 49 remplissage procedures with arthroscopic Bankart repairs.

There are relatively few studies in the literature describing revision surgery for failure of anterior stabilization procedures.¹,¹⁵,³³,³⁶,⁴¹,⁹⁰,¹⁰⁰,¹¹⁶,¹²⁰,¹²⁴,¹²⁸,¹³¹,¹³⁹,¹⁵⁰,¹⁵⁷,¹⁸⁰,¹⁹¹,²⁰⁰,²¹⁹,²²⁸,²²⁹ The outcomes of these surgeries are less predictable than those performed primarily, with failure rates as high as 44% in patients who have had multiple prior surgeries.¹²⁰ With recurrent anterior instability, the most common revision surgery options include repeat arthroscopic reconstruction, open Bankart repair, and Latarjet-Bristow procedures. Less commonly performed procedures include bone grafts for large Hill-Sachs defects, and allograft or autograft capsular reconstruction for significant soft-tissue deficiencies as seen following failed thermal capsulorrhaphy. The optimal procedure is determined based on patient expectation, activity level, surgeon experience, and pathologic lesions leading to failure of the previous surgery.