Retroglenoid osteotomy with capsular shift for posterior shoulder instability: technical note and short-term results

Abstract

Purpose

Posterior shoulder instability, often related to glenoid dysplasia, can also be associated with inferior capsuloligamentous instability. This study presents the technical procedure of the retroglenoid osteotomy with posterior capsular shift, supported by clinical and radiological evaluation. Our hypothesis was that the re-orientation of the glenoid and the capsular retensioning could achieve shoulder stability.

Methods

Eleven shoulders of nine patients with posterior shoulder instability and glenoid dysplasia underwent retroglenoid osteotomy with posterior capsular shift. A graft was inserted into the posterior opening, with full respect of the anterior wall of the scapular neck. The technique is detailed. Clinical, radiological, and functional outcomes were assessed over an average follow-up period of 20 months.

Results

This study showed clinical improvements in all patients, with the complete resolution of instability symptoms and radiological correction of glenoid retroversion. No complications or re-operations were observed. Range of motion in the shoulder joints was preserved, and patients reported satisfaction with their outcomes. The total Constant Score increased from 58.2 to 76 points, the mean post-operative Walch-Duplay score was to 92 points (10–100).

Conclusion

In the context of posterior shoulder instability with glenoid dysplasia and inferior laxity, the retroglenoid osteotomy associated with the posterior capsular shift offers substantial clinical improvement in the short term, leading to symptoms resolution, stable shoulder and radiological correction of glenoid retroversion.

Level of evidence

level 4 Current clinical practice.

Introduction

Posterior glenohumeral instability represents 3% of all shoulder instabilities [1]. The underlying etiologies are numerous and multifactorial, including glenoid retroversion, traumas and sequelae, posterior labral lesions, hyperlaxity (voluntary and secondary involuntary dislocations), and muscle deficits [2–4].

The excessive displacement of the humeral head toward the back of the glenoid cavity characterizes the posterior instability, which can lead to chronic pain, joint effusions, reduced muscle strength, and overall shoulder instability. This condition is often challenging to diagnose because its symptoms are generally vague and can be confused with other shoulder conditions, such as subacromial impingement or glenoid labrum injuries.

Clinically, it can manifest as mild shoulder discomfort to true recurrent dislocation [5–8].

The initial treatment of posterior shoulder instability is typically non-surgical [9]. Functional rehabilitation aims to strengthen the muscles to stabilize the humeral head within the glenoid cavity. This treatment yields satisfactory results in 89% of cases [2]. In case of medical treatment failure, surgical intervention may be necessary [7] like capsuloplasties, posterior Bankart procedure, posterior bone abutment, posterior glenoid osteotomy.

We consider the glenoid retroversion as the main osseous cause of posterior instability, and our concern was to correct this glenoid position. One surgical option to address posterior shoulder instability associated with glenoid retroversion is the posterior opening-wedge osteotomy. Scott [10] first proposed this technique in 1967 when he published an operative technique of posterior glenoid opening osteotomy used in a series of three cases in the context of chronic glenoid retroversion. This initial technique described the use of an iliac bone graft. We have developed the use of an in-situ graft harvested from the scapular spine.

We also consider the posterior capsule distension as a part of the causes of instability. This has guided our choice for treating the posterior shoulder instability with both posterior glenoid osteotomy and capsular retensioning.

The different etiologies of the posterior instability of the shoulder are often consider separately, and the association of both the bony glenoid insufficiency and the posterior and inferior capsuloligamentous distension is not commonly considered. The recent publication of the Delphi-round consensus of a panel of experts has provided new insights about this complex pathology [11, 12], and made an interesting updating of the etiologies and treatments.

The objective of our study was to evaluate the clinical and morphological outcomes of posterior glenoid opening osteotomy with capsular shift in the treatment of posterior scapulo-humeral instability with an in-situ spinal graft.

Materials and methods

We conducted a single-center retrospective study from 2010 to 2019, including all adult patients undergoing posterior glenoid opening osteotomy for the treatment of posterior glenohumeral instability associated with a posterior glenoid retroversion exceeding 12°, completed by a posterior capsular shift. Informed consent to participate was obtained from all the participants, and they all allowed the using of their clinical data and imaging pictures. The research was conducted in accordance with the Declaration of Helsinki. This study received the agreement of the Ethical Committee for Researches on Existing Data of the Rouen-University Hospital (N° E2024-21).

A subjective sensation of subluxation or true posterior shoulder dislocation associated with at least one positive posterior instability test during clinical examination defined the posterior shoulder instability. The pre-operative planning assessed the glenoid retroversion by using various imaging modalities, including standard shoulder computed tomography, shoulder CT arthrography, and shoulder magnetic resonance imaging.

Patients

Our cohort consisted of eleven shoulders from ten patients (seven males and three females). One patient underwent surgery on both shoulders, with a three-year interval between surgeries. The average age of our cohort was 30 years (from 21 to 43 years). Among the eleven shoulders studied, five were on the dominant upper limb, and six were on the non-dominant upper limb. The minimum follow-up in our current practice was one year.

The patients who underwent this procedure presented with different etiologies for their posterior shoulder instability: for five operated shoulders, it was a post-traumatic instability (three dislocations during physical activities, two dislocations during a professional activity), while for six shoulders, a previous voluntary childhood posterior shoulder instability became involuntary in adulthood. They could be considered as Type B in the recent classification by Paksoy et al. [13].

Exclusion criteria

None of the patients had undergone previous shoulder surgery on the affected shoulder, none had a previous fracture, none presented any neurological disorder. The first surgical procedure performed was the posterior glenoid opening osteotomy (Table 1). No age limit was considered, this short series represents the complete series of our department. No osteoarthritic changes were observed on the pre-operative imaging.

Table 1.

Demographic data

Cases	Sex	Age (Years)	Operated upper-limb	Dominant upper-limb	Etiology	Graft harvesting site	Associated surgery	Follow-up (months)
1	M	28	L	L	Post-traumatic (professional)	Iliac	Posterior capsular shift	12
2	M	28	R	R	Voluntary became involuntary	Scapular	Posterior capsular shift	39
3	F	23	R	R	Post-traumatic (sport)	Scapular	Posterior capsular shift	14
4	M	32	L	R	Post-traumatic (sport)	Scapular	Posterior capsular shift	15
5	M	30	L	R	Voluntary became involuntary	Scapular	Posterior capsular shift	12
6	F	29	L	R	Voluntary became involuntary	Scapular	Posterior capsular shift	13
7	M	29	R	L	Post-traumatic (sport)	Scapular	Posterior capsular shift	12
8	F	39	L	L	Voluntary became involuntary	Scapular	Posterior capsular shift	72
9	M	43	R	L	Voluntary became involuntary	Scapular	Posterior capsular shift	12
10	M	29	L	R	Voluntary became involuntary	Scapular	Posterior capsular shift	14
11	M	21	R	R	Post-traumatic	Scapular	Posterior capsular shift	13

Clinical evaluation

Shoulder mobility

We studied active joint range of motion in abduction, flexion, external rotation one (elbow to the body), external rotation two (abduction at 90°), external rotation three (flexion at 90°), internal rotation, pre-operatively, post-operatively at three months and at the last follow-up.

Subjective instability and objective laxity

Clinical parameters assessed pre-operatively included the presence of posterior instability as described by the patient (subluxation, true dislocation, posterior apprehension test), and confirmed the posterior instability through the posterior relocation test.

Subluxation was defined as the patient’s sensation of posterior displacement immediately followed by spontaneous repositioning of the glenohumeral joint.

True posterior dislocation was described by the patient as a sensation of dislocation with posterior movement of the humeral head beyond the posterior border of the glenoid, resulting in functional impairment requiring reduction by a third party or by the patients themselves.

The relocation test corresponds to the sensation of instability during a maneuver in which the patient lies supine on the examination table with the examined shoulder extending beyond the table. The examiner performs it by blocking the scapulothoracic sliding plane and exerting posterior mobilization of the humeral head alone. The test is positive when the patient describes a sensation of instability during posterior mobilization and a disappearance of this sensation upon cephalic recentering.

The tests of posterior laxity confirmed a posterior drawer and/or a catch during mobilization The tests of inferior laxity associated a positive sulcus test and a positive Gagey passive hyperabduction test [14].

Gagey’s test [14] involves immobilizing the scapulothoracic sliding plane and passively abducting the shoulder. This test is positive when passive abduction is measured superior to 105°, indicating inferior instability describing capsular laxity contributing to glenohumeral joint instability.

Radiological evaluation

Radiological parameters studied pre-operatively included glenoid cavity retroversion on three-dimensional imaging (computed tomography, MRI) combined with standard 2D radiography [15]. Post-operatively, our standard care protocol for monitoring shoulders operated on for instability included reviewing standard radiographs with anterior and axillary views to assess the new version of the glenoid cavity, graft placement, osseous integration (consolidation), and the presence of complications.

The pre-operative glenoid retroversion was measured in eight shoulders using standard computed tomography, in two shoulders using shoulder CT arthrography, and in one shoulder using shoulder magnetic resonance imaging (Figs. 1 and 2). In this series, the posterior edge of the glenoid was respected.

Fig. 1.

Axial slice: pre-operative CT scan, from patient 3 with glenoid dysplasia (15°)

Fig. 2.

Axial slice: pre-operative CT-arthrography scan, from patient 1 with glenoid retroversion (22°)

Surgical technique

We relied on the surgical technique proposed by Scott in 1967 [10] with some modifications. In our series of eleven shoulders, all patients underwent a posterior capsular shift in addition to the osteotomy and graft. During the three first operations, we used an iliac bone graft, while in the remaining six operations, we used an in-situ graft harvested from the scapular spine.

The patients were placed in a prone position. After identifying the scapular spine, acromion, and the posterior border of the deltoid muscle, the incision was vertical along the posterior border of the deltoid muscle. Partial deltoid muscle detachment by periosteal elevation along the posterior border of the spine (donor site preparation) exposed the space between the infraspinatus and teres minor muscles, allowing visualization of the scapular neck and the posterior surface of the glenohumeral joint capsule. The capsulotomy followed a “T” capsular incision, releasing both upper and lower capsular valves at the scapular base.

We marked the osteotomy line seven millimeters medial to the posterior edge of the glenoid cavity (Fig. 3). Then, we prepared the posterior osteotomy line by fine drilling, creating a “postage stamp” osteotomy. The osteotomy used an oscillating saw, taking care to preserve the anterior wall of the scapular neck.

Fig. 3.

Posterior view: retroglenoid osteotomy, dry bone. The mean distance of the suprascapular nerve (coursing around the lateral edge of the scapular spine) to the posterior edge of the glenoid cavity is 17 mm. The dotted line corresponds to the osteotomy

Regarding graft harvesting, two options were available:

• Iliac graft with harvesting and shaping to the desired dimensions ipsilateral to the operated side.

• Scapular graft: by taking advantage of the previously performed periosteal elevation (Fig. 4), allowing us to immediately access the posterosuperior surface of the scapular spine, offering a bone graft of 25 to 30 mm in length and five to seven millimeters in width (Fig. 5a and b), depending on the required anteversion correction (Figs. 6a and b and 7).

Fig. 4.

Subperiosteal exposure, posterior view

Fig. 5.

a Superior view: bone graft harvesting line, along the scapular spine. b Postero-superior view: bone graft harvesting line, corresponding to the required surface of the graft

Fig. 6.

a Superior view: The junction between the spine and the acromion has been sectioned on the drawing for showing the line of the retroglenoid osteotomy. b Postero-superior view: The graft is harvested from the posterior edge of the spine, with the thickness corresponding to the desired retroglenoid angular correction, and from the upper surface of the scapular spine. Two holes have been made in the cortical bone on each side of the retroglenoid osteotomy line, for passing sutures

Fig. 7.

Posterior view: the graft has been impacted in the retroglenoid osteotomy. A trans-osseous suture can reinforce the fixation. The posterior capsular shift closes and increases the thickness of the capsule

The trapezius muscle is raised upwards with the periosteum on its upper surface, and the deltoid is spread downwards, to expose the posterior edge and upper surface of the scapular spine, close to the junction with the acromion. At the end of the procedure, the periosteal suture covers the harvested area.

The postero-anterior osteotomy respects the anterior cortical bone of the scapular neck.

The impaction of the graft on the osteotomy site maintained the desired glenoid version and did not require stabilization with osteosynthesis materials.

Finally, the purse-string suture of the two capsule valves constituted a posterior capsular shift effect [Fig. 7], and we reinserted the periosteal-faced deltoid and trapezius muscles on the scapular spine before skin closure.

In the post-operative period, the immobilization used an orthopedic vest during one month, in neutral rotation position. Patients were immobilized for one to three weeks while avoiding active anterior elevation, adduction, and internal rotation, for protecting the posterior and inferior capsular repair and retensioning. Then, the internal rotation mobilization was initiated. After two months, the mobilization was unrestricted. The return to work occurred between six weeks and three months, while return to sports training was allowed at three months.

Post-operative clinical evaluation

The clinical controls at three months and one year or maximal follow-up performed the measurements of the range of motion of the operated shoulder in passive and active joint motion in abduction, flexion, external rotation one (elbow to the body), external rotation two (abduction at 90°), external rotation three (flexion at 90°), and internal rotation.

The clinical examination tests were the same as in pre-operative period:

• Posterior instability: described by the patient (subluxation, true dislocation, posterior apprehension test) and confirmed during clinical examination (posterior instability: relocation test and inferior instability).

• Posterior laxity: searching for a posterior drawer and/or a catch during mobilization.

• Inferior and Posterior laxity: Sulcus test and Gagey’s passive hyperabduction test.

Radiological evaluation

Radiological parameters studied pre-operatively included glenoid cavity retroversion on three-dimensional imaging (computed tomography, MRI) combined with standard 2D radiography [15]. Post-operatively, our standard care protocol for monitoring shoulders operated on for instability included reviewing standard radiographs with anterior and axillary views to assess the new version of the glenoid cavity, graft placement, osseous integration (bone graft healing), and the presence of complications.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Results

Complications and reintervention

Out of the eleven operated shoulders, none presented immediate post-operative complications such as fractures, hematomas, surgical site infections, or nerve injuries (suprascapular nerve). No complications were identified during the follow-up. None of the patients required further surgery on the operated shoulder. No patients experienced chronic pain during the follow-up period.

Clinical results

The clinical results of this study showed satisfaction from all patients with complete relief of sensations of instability and catching in early post-operative and one-year follow-up, with no residual apprehension.

Seven out of eleven operated shoulders were declared pain-free (mean score of pain 11.8 points/15; 10 to 15).

The activities of daily living showed no limitations in nine shoulders (mean score 19.2/20 points; 16 to 20).

The range of motion was moderately improved, as no stiffness was present pre-operatively. The average pre-operative abduction was 163.3° (90° − 180°), and 164.4° (120° − 180°) post-operatively.

The average pre-operative active flexion was 153.3° (90° − 180°), compared to an average of 170° (150° − 180°) post-operatively at the longest follow-up.

External rotation one showed improvement with an average pre-operative measurement of 68° (40° − 90°) and an average post-operative measurement of 87° (60° − 90°). External rotations two and three increased from 69° (40–90) in pre-operative, to post-operative measurements (90°).

The vertebral level of hand excursion also showed improvement in internal rotation, with an average pre-operative level of T8 (T4-T12) compared to an average level of T7 (T4-T12) post-operatively at one year (Table 2).

Table 2.

Clinicals results about the active anterior elevation (flexion), abduction, external rotation (in degrees°), internal rotation (vertebral level), laxity/drawer and (sub)luxation preoperative (preop) and postoperative (postop). external rotation 1 (vertical position along the trunk), external rotation 2 (in abduction 90°), external rotation 3 (in flexion 90°)

Cases	Flexion (°)		Abduction (°)		External Rotation (°)				Internal Rotation Levels		Laxity/drawer Preoperative/ Postoperative		Sub+ Luxation (+/++) None (-)
	Pre op	Post op	Pre op	Post op	Pre op		Post op		Pre op	Post op	Preop	Post op	Pre op	Post op
					ER 1	ER 2&3	ER 1	ER 2&3
1	90	120	90	150	60	90	80	90	T12	T12	Ant/Post	-	-	-
2	160	160	120	160	80	90	90	90	T8	T9	Inf/Ant/Post	-	++	-
3	180	160	180	160	90	90	90	90	T8	T8	Inf/Ant/Post	-	+	-
4	180	180	180	180	60	90	80	90	T4	T4	Post	-	-	-
5	160	180	180	180	60	90	60	90	T6	T5	Inf/Post	-	+	-
6	160	160	90	160	80	90	60	90	T10	T8	Inf/Ant/Post	-	+	-
7	180	160	180	180	60	90	45	90	T10	T6	Inf/Post	-	+	-
8	180	180	180	180	80	90	80	90	T6	T6	Inf/Post	-	++	-
9	180	180	180	180	80	90	80	90	T9	T8	Inf/Post	-	++	-
10	180	180	180	180	60	90	60	90	T10	T10	Inf/Post	-	++	-
11	90	140	90	140	40	60	40	90	T12	T12	Inf/Post	-	++	-

The total Constant Score [16] was improved for all the patients, from 58.2 points (36–43) pre-operatively to 76 points (57–84) (Table 3).

Table 3.

Clinical results according to the Constant score[16] parameters

Cases	Pain (/15 points)		Activities of daily living (/20 points)		Range of movement (/40 points)		Power (/25 points)		Total Constant Score (100 points)
	Pre-Op	Post-Op	Pre-Op	Post-Op	Pre-Op	Post-Op	Pre-Op	Post-Op	Pre-Op	Post-Op
1	5	10	13	16	22	26	3	5	43	57
2	10	15	14	20	24	38	2	8	50	81
3	10	10	15	20	36	36	4	6	65	67
4	10	10	14	20	40	40	8	10	72	80
5	5	15	14	20	38	40	6	9	63	84
6	10	15	13	20	38	38	8	9	69	82
7	5	15	14	20	36	36	8	10	63	77
8	5	15	11	20	38	38	4	5	56	78
9	10	15	14	20	36	36	8	10	68	81
10	10	15	14	20	36	38	6	10	56	83
11	5	10	9	16	18	32	4	8	36	66

The functional results were also assessed through the Walch-Duplay score for instability of the shoulder [17], and the mean score was 92 points (80–100) (Table 4). Only five patients had a sports practice with leisure level (table tennis (2), badminton, martial arts (Policeman) and danse), and achieved a complete return to sports. None presented any instability.

Table 4.

The post-operative results about eleven shoulders according to the Walch-Duplay score of instability of the shoulder [17]

Patients	Level of sports practice CLN	Type of sport 0 1 2 3 4 5	Side RL Dd	Daily activity (/25 points)	Stability (/25 points)	Pain (/25 points)	Mobility (/25 points)	Total (100 points)
1	N	0	Rd	25	25	15	15	80
2	L	1	RD	25	25	25	25	100
3	L	1	RD	15	25	15	25	80
4	L	2	Ld	25	25	15	25	90
5	L	1	Ld	25	25	25	25	100
6	N	0	Ld	25	25	15	25	100
7	N	0	Rd	25	25	15	25	90
8	L	1	LD	25	25	25	25	100
9	N	0	Rd	25	25	25	25	100
10	N	0	Ld	25	25	25	25	100
11	N	0	RD	25	25	15	15	80

Due to the small sample size and similarity of observed values, statistical analysis was not considered meaningful or reliable.

Radiological results

All standard radiographs taken at the one-year follow-up confirmed the presence of a fully consolidated graft. The X-Rays follow-up did not show any osteoarticular complication. No complication was observed, neither loss of the grafts nor fractures of the donor site.

Radiological results demonstrated a reduction in glenoid retroversion after surgery. The average pre-operative retroversion was 18.6 degrees (12° − 25°), while that measured post-operatively at one year was 6.4 degrees (3° − 8°), resulting in an average correction of 11.7 degrees (9° − 22°) (Table 5).

Table 5.

Radiological results of the included patients

Cases	Preoperative measurement technique	Postoperative measurement technique	Graft consolidation (1 year)	Postoperative complication	Preoperative glenoid retroversion (°)	Post-operative glenoid retroversion (°)	Glenoid retroversion delta (°)
1	CT	Radiography (axillary view)	Fully	0	22	8	14
2	CT-arthrography	Radiography (axillary view)	Fully	0	16	7	9
3	CT	Radiography (axillary view)	Fully	0	15	6	9
4	CT	Radiography (axillary view)	Fully	0	12	7	5
5	CT-arthrography	Radiography (axillary view)	Fully	0	18	5	13
6	CT	Radiography (axillary view)	Fully	0	25	3	22
7	arthro-MRI	Radiography (axillary view)	Fully	0	17	5	12
8	CT	Radiography (axillary view)	Fully	0	19	4	15
9	CT	Radiography (axillary view)	Fully	0	16	6	10
10	CT	Radiography (axillary view)	Fully	0	20	10	10
11	CT	Radiography (axillary view)	Fully	0	25	10	10

Pre-operative measurements were obtained from 3D imaging on scans; while post-operative controls and radiological measurements were performed, using 2D standard radiographs (Figs. 8a and b and 9a and b).

Fig. 8.

a Anteroposterior view: post-operative radiography from patient 8 with scapula spine graft (G) incorporated in the posterior glenoid osteotomy. The anterior cortical hinge has been fully respected. b Axillary view: post-operative radiography from patient 8 with scapula spine graft (G) incorporated in the posterior glenoid osteotomy. The anterior aspect of the scapular neck has been carefully respected (A)

Fig. 9.

a Antero-posterior view: 1-year follow-up radiography from patient 8 with scapular spine graft (G) integrated in the posterior osteotomy. b Axillary view: 1-year follow-up radiography from patient 8 with scapular spine graft (G) integrated in the posterior osteotomy

Standard radiography was less irradiating, more accessible, cost-effective, and included in our routine follow-up. The radiological check-up was composed of a frontal view and an axillary view.

Edwards et al. showed the importance of the axillary profile in shoulder instability (anterior instability in this case), enabling lesions of the glenoid cavity and glenoid neck to be identified [18].

Discussion

The sequence of the treatment is important, but the field of posterior instability of the shoulder is complex and includes post-traumatic instability, and voluntary dislocation in young people that became unvoluntary in adults. These conditions related to the patient’s history led to separate traumatic and atraumatic posterior instability.

Conservative surgical treatments for young individuals with posterior shoulder instability associated with glenoid dysplasia miss consensual indications: physiotherapy, modifications to professional and sports activities, and even immobilization are the primary medical treatment options for posterior shoulder instability [19, 20]. Immobilization for one to three weeks, but with avoiding active anterior elevation, adduction, and internal rotation movements. Secondarily, the aim of the physiotherapy is the muscle strengthening of the rotator cuff muscles, posterior deltoid muscle, and scapular stabilizing muscles. All our patients had undergone physiotherapy and proprioceptive rehabilitation.

The surgical treatment takes place when a patient continues to experience clear instability despite well-executed medical treatment and functional rehabilitation for a sufficient duration of at least three to six months [21]. The proposed operation techniques are related to the osseous and capsulolabral pathologies. The osseous lesions are glenoid dysplasia and retroversion, that are constitutional features, but also posterior glenoid loss and bony Bankart avulsion, that are post-traumatic lesions. The capsule distension can be a part of a global hyperlaxity, but also the sequel of a previous posterior dislocation. The labrum avulsion can be related to a previous posterior dislocation.

According to these types of lesions, the proposed techniques can be posterior bone blocks, posterior labrum repair, retroglenoid osteotomy, posterior capsular shift, but also remplissage of MacLaughlin impacted fracture.

The association of an increased retroversion of the glenoid with a posterior and inferior distension of the capsule leads to consider both together these etiologic parameters of the posterior instability. Our series is limited to patients operated for posterior laxity with respect of the posterior edge of the glenoid, and postero-inferior capsuloligamentous distension.

The patients in our series did not experience major complications. In their systematic review of 356 operated shoulders out of nine studies with different follow-up durations, Sardar et al. reported 20% persistent instability and 4% fractures (glenoid neck and acromion) [22]. Patient-reported pain varies in different studies: Brewer et al. [23] did not find post-operative pain, while Hawkins et al. observed the presence of persistent discomfort without associated chronic pain in three out of 12 shoulders [24]. Inui and Nobuhara demonstrated a significant reduction in post-operative pain compared to pre-operative levels [25].

Different surgical treatments have been described in cases without glenoid dysplasia, such as open or arthroscopic posterior capsular shift, reverse Bankart or filling reverse Hill-Sachs lesion with favorable functional outcomes [26–29]. The decision to pursue these treatments should take into consideration the presence of capsulolabral involvement alone or combined with bony abnormalities. Analysis of glenoid version is a crucial factor in the surgical decision-making process, and is essential for tailoring surgical treatment to the root cause of posterior shoulder instability. Privitera et al. conducted a study measuring normal glenoid version in 410 healthy and willing subjects, finding an average version of -1° to + 3° [30]. In their systematic literature review, Sandar et al. reported an average pre-operative retroversion of 15° in operated glenoids, compared to six degrees post-operatively, demonstrating the morphological effectiveness of retroglenoid osteotomy in glenoid retroversion [22]. Several surgeries involving a bony procedure have been described, such as posterior glenoid opening osteotomy, humeral derotation, posterior open bony block or posterior arthroscopic bony block [31, 32]. Kouvalchouk’s acromial pediculated bone-block has the advantage of combining mechanical action through the bone block and a dynamic stabilization by the deltoid muscle.

The procedure therefore treats the symptom rather than the cause of posterior instability with glenoid dysplasia [33] as the retroversion is not corrected. In 2017, Métais et al. redescribed this surgical technique under arthroscopy adding a posterior capsular shift [34].

Strict capsulolabral repair in a patient with glenoid retroversion > 10° is associated with a less favorable functional prognosis [29]. We consider regrettable that the inferior laxity is not considered as a major parameter.

Godenèche et al. showed no difference in outcomes of surgical treatment for traumatic or atraumatic posterior shoulder instability, in their study about 117 patients with 48 months (12–159)-follow up, and different procedures (bone blocks, Bankart repair, Capsular retensioning) [35].

Our study suggested that posterior opening osteotomy yielded good short-term results for instability, but glenohumeral dislocations could recurred. With an average follow-up of 15 years, Waltenspül et al. described new signs of instability in 86% of patients from a cohort of seven shoulders, appearing on average six years after surgery [36]. All patients in this cohort exhibited arthritic changes in the operated shoulder at the last radiological check-up [36]. Sardar et al. reported the occurrence of only 6% arthrosis post-operatively [20]. This data seems to vary widely among studies, ranging from 25% in Graichen et al.‘s study [37] to 7.7% in Bessems and Vegter’s study [38]. Another study from Malik et al. found new signs of instability in 22%, 18% of complications (7% arthrosis post-operatively and 5% of fractures) [39].

Further studies with long follow-up should help to establish a reliable value about the development of arthrosis in patients who have undergone retroglenoid osteotomy. This aspect remains difficult to assess, because the instability is a risk factor of osteoarthritis.

The predominant use of spinal grafts (90%) in our study allowed for high-quality bone graft harvesting, appropriately sized, close to the surgical site, while minimizing post-operative iliac pain and, to a lesser extent, scarring. Other studies, such as Brewer et al., demonstrated the benefits of iliac grafts, which have higher bone density, greater stability, and easier preparation than spinal grafts [23].

All patients in this study underwent a posterior capsular shift procedure to combine capsulolabral treatment with the bony treatment of osteotomy. Adjusting capsular tension, along with correcting glenoid retroversion, showed better results [40]. Neer et al. described an inferior capsular shift procedure for inferior and multidirectional instability. This surgery can correct instability without osseous modification procedure [41].

Regarding our results, we consider that both the bone etiology and the capsular discrepancy are involved in the posterior instability. This leads to correct these two causes, with an osseous re-orientation and a capsular retensioning.

Limitations

Limitations The primary limitations of our study are the small series and the short follow-up duration, as other scientific studies have shown the recurrence of instability symptoms over time and the development of glenohumeral arthrosis. A single experienced surgeon performed all the operations, which was the primary strength of this study.

Our series of patients is reduced, but this short sample corresponds to our surgical activity in posterior shoulder instability, that is limited to rare cases. This is why it remains difficult to accumulate a large experience in this field of shoulder surgery Excepted the cases 1 and 11 who showed a light limitation of the shoulder flexion at the longest outcome, the recovery of the operated shoulders looked complete.

The post-operative imaging did not use a CT-Scan in all the cases, as the authors considered axillary views sufficient for evaluating graft integration and humeral head centering, and the glenoid re-orientation. This could induce limitation in the accuracy of the evaluation of the glenoid version, but the axillary views were sufficient for assessing the grafts integration and the version, but also the centering of the humeral head.

Conclusion

Over the one-year term, retroglenoid osteotomy with postero-inferior capsular shift performed in cases of posterior shoulder instability with glenoid dysplasia achieved an apparent clinical improvement, with recovery of the shoulder posterior stability and the glenoid radiological version correction. This short series emphasizes the importance of taking into account the inferior laxity associated with the posterior laxity. 

Abbreviations

Operated upper-limb

L: Left; R: Right

Dominant upper limb

L: Left; R: Right

CT-arthrography

Computed Tomography arthrography

MRI

Magnetic Resonance Imaging

Arthro- MRI

Arthrography- Magnetic Resonance Imaging

2D radiography

two-dimensional radiography

ER

External rotation

Pre-op

Pre-operative

Post-op

Post-operative

Level of sports practice CLN

Competition, Leisure, No sports practice

Authors’ contributions

MP: data collection, manuscript editing. AH: data collection, manuscript editing. FD: single surgeon, data collection, manuscript editing, figures drawing. All authors reviewed the manuscript and agreed for the submission of the second revised version.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The research was conducted in accordance with the Declaration of Helsinki. This study received the agreement of the Ethical Committee for Researches on Existing Data of the Rouen-University Hospital (N° E2024-21). Informed consent to participate was obtained from all the participants.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests