Fig. 5.1
Axial CT view of patient with posterior boy Bankart lesion following traumatic posterior instability event (arrow). Hemarthrosis secondary to trauma is also visible intra-articularly
Fig. 5.2
Axial view of CT scan with engaging reverse Hill-Sachs lesion on the anterior humeral head due to recurrent posterior instability
Magnetic resonance arthrogram (MRA) is more sensitive for diagnosis of chondrolabral pathology than standard MRI without contrast administration [39]. Pathology consistent with posterior instability injury, particularly when caused by internal impingement, seen on MRA includes posterior undersurface rotator cuff abnormalities, abnormal labral signal, and cystic changes at the posterosuperior humeral head [40–42]. Additionally, MRA reveals static posterior translation of humeral head on the glenoid, discrete tears in the posterior capsule (Fig. 5.3), labral tears/splits or labrocapsular lesions, posterior HAGL lesions, posterior labral periosteal sleeve avulsion (PLPSA), and subscapularis tendon avulsions (Fig. 5.4) [42, 43]. A Bennet lesion, enthesopathy, or heterotopic ossification at the origin of the posterior IGHL is also commonly seen in overhead throwers [44]. In a radiographic study of overhead athletes with GIRD and internal impingement symptoms, throwers tended to have a thicker labrum (6.4 vs. 2.9 mm), longer capsule-labrum length (8.8 mm vs. 5.4 mm), and shallower capsular recess in the posterior inferior shoulder (94° vs. 65°) than non-throwing controls [42].
Fig. 5.3
Axial MRI imaging showing posterior capsulolabral injury with capsular separation (arrow)
Fig. 5.4
Axial cut of an MRA revealing a glenolabral articular disruption (GLAD) lesion. The displaced flap of articular cartilage is seen in the posteroinferior capsule
In a review of 28 shoulders with previously diagnosed posterior HAGL lesions, Rebolledo et al. found that the most common concurrent injuries were reverse Hill-Sachs lesions (36 %), anterior Bankart lesions (29 %) and posterosuperior rotator cuff tears (25 %), and bony HAGL avulsion (7 %). The posterior HAGL pathology was a complete tear in 71 %, a partial tear in 25 %, and a floating lesion in 4 % of patients [18].
The posterior labrum has a number of anatomic variants. Using CT arthrogram, Nourissat et al. defined four varieties of posterior labrum insertion: Type 1, the most common, was a posterior labrum fully inserted flush with cartilage, Type 2 was a medialized insertion of superior segment, Type 3 was a medialization of superior and medial segment, and Type 4 was a medialized insertion of the entire posterior labrum [45].
Kim et al. defined a classification for tears of the posterior labrum. Type I is an incomplete detachment, Type II the nominal “Kim lesion,” Type III a chondrolabral erosion, and Type IV a flap tear of the posteroinferior labrum (Fig. 5.5) [46]. A “Kim lesion” is a concealed but complete detachment that can masquerade on arthroscopy as a superficial crack in the junction of the posteroinferior chondrolabral junction, but hides a complete detachment of the labrum from the glenoid rim.
Fig. 5.5
(a) Axial cut, proton density MRI images showing posteroinferior labrum degeneration and tearing (arrow), and (b) inferior blunting of the posteroinferior margin of the labrum (arrowhead)
Radiographic studies have consistently shown that shoulders with atraumatic posterior instability have alterations in the glenoid and chondrolabral version than age-matched controlled shoulders. Excessive retroversion of the glenoid has been defined as greater than −7° in the sagittal plane [47]. The increased retroversion is most frequently seen at the inferior aspect of the glenoid [48]. Kim et al. [49] examined 33 shoulders with atraumatic posterior instability and observed that the glenoid was more shallow and there was increased retroversion of the bony glenoid and chondrolabral complex at the middle and inferior glenoid in the unstable patients. Hurley et al. reported an average glenoid retroversion of −10° in patients with posterior instability as compared to −4° in uninjured controls [50]. In a series by Bradley et al., mean chondrolabral retroversion was 10.7° (controls 5.5°) and mean glenoid version 7.1° vs. 3.5°.
Rarely performed, but potentially useful in diagnosis, is an MRA performed with the shoulder in the abduction external rotation position. This arm position can reveal a “peel-back” lesion of posterosuperior labrum which may reduce to anatomic position in neutral arm position. This position can also help expose a Kim lesion, with the arm movement and position forcing intra-articular contrast to track into the Kim lesion under the labrum and reveal the extent of the damage. In a retrospective series of 34 MR in both the ABER and standard position with known labral peel-back seen on the ensuring arthroscopic repairs, this MR technique was 73 % sensitive and 100 % specific for the diagnosis; positive predictive value was 100 %, and negative predictive value was 78 %. Of these 34 patients, five had labral tears that would not be visible and went undiagnosed in MR in the standard neutral position [51].
MRA is also useful in the postoperative setting at a mean of just 6 months following surgery; 40 patients with recurrent instability underwent repeat MRA prior to undergoing revision shoulder stabilization. Finding was compared with pathology observed at revision arthroscopy. MRA was 91.9 % accurate for diagnosis of labral tears with a 96.2 % sensitivity. Accuracy for detecting rotator cuff lesions was 87.2 % (sensitivity 94.1 %) and for biceps injury was 95.7 % (sensitivity 85.7 %) [52].
5.9 Nonoperative Treatment
Conservative treatment for a minimum of 6 months is almost universally recommended. Activity modification and physical therapy, with particular focus on dynamic muscular stabilizers of the shoulder, have been shown to result in significant improvement in many patients [12, 32, 53, 54]. Elements of scapular dyskinesia or protraction can be addressed with a periscapular strengthening regimen. For patients with symptoms due to GIRD, the “sleeper stretch” has been shown to be extremely effective in avoiding progression of symptoms or surgery. To perform this stretch, the patient lies on the injured side with the staple stabilized against a wall, the shoulder and elbow each flexed to 90°. The contralateral, nondominant arm applies passive internal rotation force to the wrist [22]. Nevertheless, some patients will not improve with therapy alone. Burkhead et al. showed that only 16 % of patients with traumatic posterior instability improved with therapy, versus nearly 80% resolution in patients with atraumatic instability [8, 53].
5.10 Surgical Treatment
Options for operative treatment of posterior instability are divided into two categories, based on the underlying type of lesion causing the instability. Procedures to correct or supplement osseous deficiencies of the glenoid or humerus address engaging reverse Hill-Sachs lesions, excessive glenoid retroversion, glenoid hypoplasia, or posterior glenoid deficiency. The other subgroup is procedures to correct soft tissue injuries of the rotator cuff or posterior capsulolabral complex.
5.11 Arthroscopic Treatment
Arthroscopic procedures for posterior instability have become the standard of care in the absence of severe bone deficiency. Surgical techniques address a spectrum of pathology in posteriorly unstable shoulders (Figs. 5.6, 5.7, and 5.8). The authors prefer a lateral decubitus position for arthroscopic labral repair and posterior capsular imbrication. Necessary portals include a high posterolateral portal and two anterior rotator interval portals, one high and one low. The angle of the glenoid makes anchor placement challenging, so placement of the posterior portal as lateral as possible facilitates the correct angle. An accessory portal in the far lateral position for anchor placement is often helpful.
Fig. 5.6
Arthroscopic view of an isolated posterior labral tear
Fig. 5.7
An arthroscopic view of a posterior labrum periosteal sleeve avulsion (POLPSA) lesion prior to repair. (a) Shows the periosteal sleeve layer separation along the posterior labrum. (b) Shows the posterior displacement of the capsulolabral complex along and posteroinferior glenoid neck, indicative of periosteal sleeve avulsion-type of injury that has retracted medially
Fig. 5.8
Arthroscopic view of the GLAD lesion (glenohumeral articular disruption) seen on MRI in Figure 4. (a) Shows the lesions initiation within the chondral layer of the glenoid surface. (b) Shows the complete separation of the posterior cartilage from the glenoid, along with the attached labrum. This correlates with the MRI image in Figure 4 showing contrast tracking underneath the chondral surface
For the repair, the high anterior portal is used for viewing and the lower rotator interval portal, just superior to the subscapularis, is used for shuttling. A large cannula is placed in the posterior portal. In the case of a labral tear, the labrum is mobilized with an arthroscopic elevator analogous to that for anterior labral repair. Anchors are placed in the posterior glenoid rim, sutures retrieved through the anterior portal. Sutures are subsequently shuttled through the soft tissue, a variety of devices are available to perform this. Capsular plication is dependent on the pathology – shoulders with traumatic labral tears can be treated with labral repair only; posterior instability resulting from laxity requires a more substantial imbrication of the posterior capsule. Posterior instability always has an inferior component, so a few plication sutures through the inferior and sometimes anteroinferior capsule are performed.
Postoperatively, patients should be immobilized with the arm in neutral rotation after posterior instability surgery to keep the tension off the posterior repair. Four weeks of immobilization is usually sufficient with elbow, wrist, and hand motion during this early recovery; however, this is adjusted for patients with atraumatic etiology. The authors generally avoid internal rotation for at least 6 weeks.
5.12 Open Procedures for Bone Deficiency
Mclaughlin presented a technique to prevent the edges of a reverse Hill-Sachs lesion from engaging on the posterior glenoid rim by limiting maximal internal rotation. The subscapularis tendon is detached from the lesser tuberosity and retracted medially and then reattached into the defect with sutures [1]. The Neer modification of the McLaughlin procedure transfers the subscapularis tendon with the lesser tuberosity still attached via an osteotomy [8, 17, 55, 56]. Iliac crest bone graft has been used to fill small lesions of <25 % of the humeral articular surface, as well as allograft osteochondral bone plugs for larger defects of up to 40–50% of the articular surface [8, 17, 56–58]. Recently, a more common variant on this procedure for reverse Hill-Sachs is a reverse remplissage, which can be done arthroscopically [59]. The subscapularis muscle is sutured into the defect with suture anchors without detaching it from the lesser tuberosity. Duey et al. also described a technique to arthroscopically imbricate the middle glenohumeral ligament into the defect [60].
The deficient or retroverted glenoid can be corrected with a bone block procedure or a posterior glenoid neck opening wedge osteotomy [61–64]. A posterior bone block procedure was performed in the lateral position via an open, posterior deltoid-splitting approach, transferring a bicortical graft from posterior iliac crest to an extra-articular position on the posterior glenoid; unlike an anterior bone block, the capsule is intentionally left to interpose between the humeral head and the bone block [65]. Postoperatively, patients are able to regain full external rotation and return to sports at their prior level [64–66]. Millet et al. descried a technique to reconstruct posterior glenoid deficiency with a distal tibial osteoarticular allograft with good outcomes in two patients [67]. A systematic review of bone block procedures confirmed that bone grafting was a reliable procedure since significant improvement in outcomes scores was regularly reported. However, long-term studies showed a clear deterioration in outcomes over time, and the majority of studies reported high frequency of radiographic graft lysis, humeral head osteonecrosis, and advanced osteoarthritis [68].
Lafosse et al. [69] described a posterior bone block procedure analogous to the anterior arthroscopic Latarjet, and this may be appropriate in experienced hands.
5.13 Outcomes
Mauro et al. recently examined the effect of glenoid version and width on outcomes following arthroscopic posterior stabilization. In 118 MRAs on athletes with recurrent posterior unidirectional instability, patients with wider and more retroverted glenoid (mean glenoid version 10.8°, mean glenoid width 28.9 mm) had better mean preoperative pain and ASES than those with narrower and more anteverted glenoids. Postoperatively, the wider glenoids continued to have better pain and ASES scores and decreased risk of failure. No correlation was seen between chondral and labral width or version with any pre- or postoperative outcome measured. Thirteen patients who failed capsulolabral repair had 3.0 mm smaller labral width and 3° less labral retroversion, but no bony version differences [70].
In a large study of 200 shoulders (183 patients), Bradley et al. demonstrated excellent outcomes in athletes with unidirectional posterior instability. All patients underwent arthroscopic posterior capsulolabral repair with anchored or anchorless repair. At 36 months, mean ASES score improved 40 points, to 85.1. Stability, pain, and function scores were also improved. Sixty-four percent returned to their prior level of play, but 10 % were unable to return to their sport due to their shoulder injury. Twelve patients were failures by the ASES criteria of score <60, and 14 had a stability score >5; seven patients were failures by both scoring rubrics [71].
Arner et al. [72] showed excellent results of arthroscopic posterior capsulolabral repair in American football players with unidirectional posterior instability. In a group of 56 athletes, 93% returned to play, with 79% returning to the same level of competition at a mean of 44.7 months postoperatively. Significant improvements were seen in pain and ASES scores and 96% were satisfied. No patient was re-dislocated and no revision procedures were required. Baseball players have also shown excellent outcomes. Wanich et al. treated a series of baseball players with posterior labral tears of the lead shoulder. Eleven of 12 treated with surgery returned to previous level of play at 5.9 months postoperatively, and all patients regained full preoperative internal and external range of motion [73]. Wrestlers are also able to return to high levels of participation following arthroscopic posterior capsulolabral complex repair and do so without recurrence over the ensuring season [74].
In a comparison of overhead throwing athletes and non-throwing athletes who underwent capsulolabral repair, McClincy et al. showed no statistical differences in ASES scores, stability, strength, or range of motion at 37 months postoperatively. However, only 60% of throwers were able to return to preoperative level of play. Surgical repair that included discrete suture anchor placement showed a tenfold increase in likelihood of returning to play when compared with anchorless repairs. No variation between repair constructs was seen in return of non-throwing athletes. Additionally, true pitchers had worse return to play rates (50%) when compared with other types of throwers [75].
Pediatric athletes also have excellent results following arthroscopic repair when nonoperative treatment fails. In 25 cases under 18 years of age with unidirectional posterior instability, 92% of shoulders were stable at final follow-up, with two recurrent instability episodes. Equivalent level of play was achieved in 67 %. Outcomes were improved in males, contact athletes, and those with history of traumatic injury as cause of instability [76].
Voluntary dislocators commonly do poorly with surgical interventions [77, 78], and rehabilitation only is encouraged for a more prolonged period than the 6 months recommended for other patient groups. In a series of patients with multidirectional instability and ability to voluntarily dislocate posteriorly, DASH scores improved significantly at all time points with a three-phase rehabilitation program. Therapy focused on correction of abnormal muscle patterns, restoration of correct scapular kinetics, and strengthening [79].
Conclusion
Posterior shoulder instability, whether due to traumatic events, accumulative microtrauma, or overhead throwing mechanics, can lead to pain, instability, mechanical symptoms, and inability to perform at a prior level of athletic performance. Historically, it was a commonly missed diagnosis, but can be strongly suspected on proper clinical exam and confirmed with focused imaging studies revealing specific spectrum of pathology. Nonoperative treatment is recommended first, especially in overhead athletes, but arthroscopic capsulolabral repair has shown excellent resolution of symptoms and return to play.