Arthroscopic distal clavicle autograft with suture button fixation for shoulder instability with glenoid bone loss: a case series

Abstract

Background

Distal clavicle autograft (DCA) has garnered increasing popularity as a cost-effective and safe autograft for glenoid reconstruction in shoulder instability. The purpose of this study is to report the clinical outcomes of a single-surgeon case series of patients managed with arthroscopic DCA for glenoid bone loss with suture button fixation at minimum follow-up of 1 year.

Methods

All cases of recurrent anterior glenohumeral instability treated with DCA by a single surgeon between January 2020 and December 2023 were reviewed. Patients were included if DCA was performed arthroscopically with 2 suture buttons for fixation, with minimum clinical follow-up of 1 year. Patients were evaluated post-operatively for recurrent instability or complications. Glenoid bone loss was measured using 3-dimensional reconstruction and best-fit circle methods on pre-operative computed tomography scans. Coracoclavicular (CC) distances were measured pre-operatively and post-operatively to assess for acromioclavicular joint instability and change in preoperative to post-operative CC distance was assessed with dependent t-test.

Results

Eight patients were included. Average age was 32.6 years. There were 7 males and 1 female. At surgery, mean glenoid bone loss was 23.3% (range: 17.8%-30.8%). At harvest, the average length, thickness, and width of the DCA were 22.8 mm, 11.6 mm, and 11.3 mm, respectively. After preparation, the graft averaged 22.8 mm in length with a width and depth of 10 mm, respectively. At a mean follow-up of 2.6 years (range: 1.5-4.0 years), no patients had recurrent subluxation or dislocation. No evidence of CC ligament damage was noted; average preoperative and post-operative CC distances were 11.5 mm and 11.3 mm, respectively, with mean change of −0.7 mm (95% confidence interval: −4.8 to +3.4 mm, P = .69). One patient was noted to have one loose posterior extracapsular button without clinical complication. No other complications were noted.

Conclusion

In this case series of patients who underwent arthroscopic glenoid reconstruction with DCA, we noted low complication rates and no evidence of recurrent instability.

Glenoid bone loss remains one of the most significant risk factors for recurrent shoulder instability following arthroscopic management of glenohumeral instability. Although the amount of bone loss necessitating restoration remains controversial, the literature suggests that glenoid bone defects exceeding 20% should be reconstructed with a bone block.^4,7,20 Lo et al²⁶ described the “inverted pear” glenoid, which represents a significant amount of bone loss–at least 25%-27% of the width of the inferior glenoid–with high risk of recurrent instability with soft-tissue Bankart repair only.

Several techniques to reconstruct glenoid bone loss have been described, some of which include the Bristow and Latarjet coracoid transfer, iliac crest bone grafting, and distal tibia allograft (DTA).^29,30,39 Coracoid-based transfers have been considered the standard of care; however, they remain technically challenging, particularly when done arthroscopically, and carry an underappreciated complication rate that can be as high as 30%, including nonunion, nerve, and vascular injuries.^8,12,15,18 A recent systematic review found a composite short-term complication rate after open Latarjet of 6.1% and after arthroscopic Latarjet of 6.8%, including a 3.2% rate of graft complications (fracture or nonunion).¹⁹ For DTA, an overall complication rate of 7.1% has been reported in recent systematic reviews,^27,37 with hardware complications being the most common (3.8%).

Distal clavicle autograft (DCA) has garnered increasing popularity due to its autograft nature, low donor-site morbidity, articular surface, availability, lower cost, and versatility.^29,34 Several recent anatomic and cadaveric studies have shown supportive evidence for DCA, demonstrating adequate mean articular surface coverage area similar to what is afforded with the Latarjet procedure,²³ lower mean contact pressures for DCA when compared to coracoid grafts,³² and the ability to restore large glenoid defects exceeding 25% while restoring more native glenoid articular surface when compared to coracoid transfers.²¹ While these studies have shown promising preliminary data, there remains a paucity of literature investigating the clinical outcomes of glenoid augmentation with DCA.

Fixation options for bony augmentation of the glenoid typically include suture buttons or screws. Suture buttons have been compared to screw fixation for DTA and Latarjet procedures, with mixed data. Pancura et al³¹ found superiority of screw fixation in DCA, whereas a large systematic review comparing outcomes following Latarjet found lower rates of graft resorption and hardware-related complications with suture button fixation compared to screws.¹⁶ Suture button fixation is increasingly being used for bony reconstruction procedures in cases of complex shoulder instability and may provide advantages such as flexibility to retension graft fixation and less hardware-related irritation, which may be advantageous in the case that revision surgery is later necessary.³⁸

The purpose of this study is to report the clinical outcomes of a single-surgeon case series of patients managed with arthroscopic DCA for glenoid bone loss with suture button fixation. We hypothesized that patients would not have any further dislocation events following this procedure at minimum 1-year follow-up.

Methods

A retrospective review of all cases of recurrent anterior glenohumeral instability treated with DCA performed by the senior author (OA), a fellowship-trained orthopedic shoulder and elbow surgeon, between January 2020 and December 2023 were reviewed. Distal clavicle graft choice was based upon surgeon preference. Indications for DCA were recurrent instability with glenoid bone loss exceeding subcritical threshold of 13.5%.³⁶ Patients were included if DCA was secured to the glenoid with 2 suture buttons for fixation and patients had a minimum clinical follow-up of 1 year. Patients were excluded if they had full-thickness rotator cuff tears at the time of injury, concomitant proximal humerus fractures other than a Hill–Sachs defect, or neuromuscular disorders.

Electronic medical records were reviewed to collect preoperative values, including age, sex, body mass index, and details of injury (traumatic vs. atraumatic), prior type of care received on the affected shoulder, and frequency of instability (single vs. recurrent). Post-operative clinical data were collected, including post-operative recurrent dislocations or subluxations, complications, return to the operating room for revision surgery, and to determine the duration of follow-up.

Radiological data

All patients underwent a computed tomography (CT) scan with a 3-dimensional reconstruction of the glenoid to characterize the bone loss.³⁵ The length, width, and thickness of the coracoid and distal clavicle for each patient were measured on the CT scans (Fig. 1 A–C). Similar measurements of the coracoid were performed for comparison, ensuring the DCA was not significantly undersized in comparison. Measurement of the distal clavicle was performed 10 mm medial to the acromioclavicular (AC) joint. Distal clavicle length was defined as the anteroposterior (AP) dimension of the distal clavicle at this point, as that edge is turned into the length of the bone along the glenoid. Distal clavicle width was defined as the superior–inferior dimension of the clavicle at this point. Glenoid bone loss, measurement of the Hill–Sachs interval, and characterization of the Hill–Sachs lesion as on-track or off-track were performed for all patients based on 3-dimensional reconstruction and best-fit circle methods as previously described (Fig. 2).⁵ Coracoclavicular (CC) distances were measured as the distance between superior border of the coracoid and the inferior border of the clavicle on anteroposterior (AP) or Grashey radiographs both pre-operatively and post-operatively to assess for AC joint instability.⁹

Figure 1.

(A) A sagittal slice of a shoulder CT scan (left) demonstrates coracoid and clavicle thickness measurements at 9.87 mm and 11.2 mm, respectively. (B) An axial slice of a shoulder CT demonstrates clavicle length measurement at 18.8 mm. (C) An axial slice of a shoulder CT demonstrates measurements of the coracoid length and width measurements at 11.6 mm and 19.7 mm, respectively. CT, computed tomography; 3D, three-dimensional.

Figure 2.

Three-dimensional CT reconstruction of the glenoid demonstrating measurement of glenoid bone loss. CT, computed tomography.

Surgical technique

All patients underwent surgery between November 2020 and August 2023 using the procedure described by Levin et al.²⁴ Patients were placed in the lateral decubitus position. A transverse incision was made over the AC joint, exposing the lateral margin of the distal clavicle. A transverse osteotomy was performed 10 mm from the lateral margin to harvest the DCA. All grafts were decorticated, and 2 2.8-mm holes, superior to inferior, were made in the graft a centimeter apart for passage of suture buttons (Endobutton, Smith & Nephew, Watford, England). The final graft was fashioned to measure approximately 22.8 mm (length) x 10 mm (thickness) x 10 mm (width) on average (Fig. 3). Thickness and width were determined according to the technique guide using their graft preparation device to appropriately traverse the bone-block cannula. Using a transglenoid drill guide placed in the center of the defect, 2 2.8-mm transglenoid tunnels are made 1 cm apart and 5 mm posterior to the joint line (Fig. 4 A and B). After preparing the glenoid surface, the bone graft transfer was introduced into the joint, paying special attention to the orientation and positioning to ensure good anterior coverage with no overhang. The graft suture buttons are shuttled through the tunnels, exiting out the posterior incision (Fig. 5); 2 respective buttons are placed over the sutures posteriorly and are secured extra-articularly and outside the capsule against the posterior glenoid using Nice knots, followed by dual tensioners and multiple half-hitches to lock the construct (Fig. 6). After confirmation of satisfactory fixation and position (Fig. 7A), the capsule was imbricated over the graft, rendering it extra-articular (Fig. 7B).

Figure 3.

Harvested distal clavicle autograft.

Figure 4.

(A) Arthroscopic image viewing from the anterosuperior portal showing introduction of transglenoid drill guide placed through the posterior portal. (B) Arthroscopic image viewing from the anterosuperior portal showing the 2 respective tunnels made to shuttle EndoButton sutures.

Figure 5.

Arthroscopic image viewing from the posterior portal showing shuttling of the distal clavicle graft into the joint.

Figure 6.

Intraoperative photo demonstrating tensioning of the EndoButton sutures posteriorly.

Figure 7.

(A) Arthroscopic image viewing from posterior portal showing placement of the intra-articular graft in anterior–inferior position without lateral overhang. (B) Arthroscopic image viewing from posterior portal showing capsular repair over the secured graft, making it extra-articular.

Post-operative protocol

Postoperatively, patients were immobilized in a sling with an abduction pillow, and the shoulder in a neutral position. Sling wear is continued for 6 weeks, with passive range of motion (ROM) starting week 1. Pendulum exercises are initiated immediately. A postoperative follow-up visit occurs 2 weeks after surgery, at which time sutures are removed and passive ROM is supervised by an outpatient physical therapist. Passive motion is progressed to full by the 8-week follow-up visit. Active assisted ROM begins at 6 weeks and strengthening at 12 weeks, respectively, with return to full activity by 6 months. Post-operative radiographs are obtained at 6 weeks, 12 weeks, 6 months, and 1 year post-operatively.

Data analysis

Statistical analyses were performed using Microsoft Excel (version 16.0; Microsoft Corporation, Redmond, Washington). Data were presented as mean ± standard deviation. Pre-operative and post-operative CC distances were compared using paired t-tests. P values < .05 were considered statistically significant.

Results

A total of 8 arthroscopic DCA fixed with suture buttons were performed between 2020 and 2023 and met all inclusion criteria. All 8 had a minimum 1-year follow-up. The mean follow-up time was 2.6 years (range: 1.5-4.0 years) (Table I). Seven patients were male (87.5%) and 1 patient was female (12.5%), with a mean age of 32.6 years (range: 20-53 years). All patients had a history of traumatic dislocation and had sustained multiple dislocations, and were treated conservatively prior to surgical intervention.

Table I.

Patient demographics.

Patient	Age (yr)	Sex	BMI (kg/m2)	Duration of follow up (yr)
1	30	Male	24.5	2.3
2	36	Female	32.17	2.4
3	53	Male	29.27	3.4
4	40	Male	27.6	3.3
5	20	Male	20.98	2.1
6	22	Male	25.83	1.5
7	34	Male	36.0	4.0
8	26	Male	24.1	1.5

BMI, body mass index.

On preoperative CT evaluation, the average glenoid bone loss was 23.3 ± 4.1%, and all patients were found to have a Hill–Sachs deformity; 4 of 8 (50.0%) had off-track lesions, and the average Hill–Sachs interval was 17.4 ± 4.4 mm (Table II). The average length (AP dimension) of the distal clavicle was 22.8 ± 3.0 mm. The distal clavicle thickness, inferior–superior dimension, at 10 mm from the distal end averaged 11.6 ± 1.4 mm. For comparison, the coracoid was found to have an average length of 23.2 ± 5.1 mm, width of 11.3 ± 1.4 mm, and thickness of 11.4 ± 1.9 mm in this patient sample.

Table II.

Preoperative CT measurements and evaluations of the coracoid, distal clavicle, and glenoid.

Patient	Coracoid			Distal clavicle		Glenoid bone loss (%)	Hill–Sachs interval (mm)	Hill–Sachs (y/N)	Bony Bankart (y/N)	On-track/Off-track
	Length (mm)	Width (mm)	Thickness (mm)	Length (mm)	Thickness (mm)
1	19.7	11.6	9.87	23.6	11.2	24.7	17.5	Y	Y	Off-track
2	20.1	8.57	11.3	19.5	12.2	23.7	11.8	Y	N	On-track
3	31.7	12.5	12.8	21.1	13.7	22.4	26.5	Y	Y	Off-track
4	19.4	10.8	11.2	22.5	13.4	30.8	18.3	Y	Y	Off-track
5	20	11	9.53	19.8	11.7	25.3	13.5	Y	Y	On-track
6	23.5	10.6	10.8	27	10.6	23.3	18.7	Y	Y	Off-track
7	30.8	13.2	15.2	27.3	10.3	17.8	16.3	Y	N	On-track
8	20.5	12.2	11.3	21.3	9.9	18.4	16.5	Y	Y	On-track
Average	23.2 ± 5.1	11.3 ± 1.4	11.4 ± 1.9	22.8 ± 3.0	11.6 ± 1.4	23.3 ± 4.1	17.4 ± 4.4

CT, computed tomography; Y, yes; N, no; mm, millimeters.

CC distances were measured on AP or Grashey radiographs both pre-operatively and post-operatively (Table III). Average pre-operative and post-operative CC distances were 11.5 ± 2.8 and 11.3 ± 3.3 mm, respectively. Mean change in CC distance pre-operatively to post-operatively was −0.7 mm (95% confidence interval: −4.8 to +3.4 mm, P = .69).

Table III.

Radiographic measurements of the coracoclavicular distances and preoperative and postoperatively.

Patient	Pre-operative CC distance (mm)	Post-operative CC distance (mm)	Change in CC distance (mm)
1	11.0	9.7	−1.3
2	9.8	18.3	8.5
3	14.3	12.2	−2.1
4	10.2	9.3	−0.9
5	7.4	No radiograph	n/a
6	11.7	11.1	−0.6
7	16.4	10.2	−6.2
8	10.8	8.5	−2.3
Average	11.5 ± 2.8	11.3 ± 3.3	−1.3 ± 4.47

CC, coracoclavicular; mm, millimeters.

At the final clinical follow-up, no patients had evidence of recurrent instability, revision surgery, or transient neurological deficits. All patients had returned to full activities. One patient was noted to have one loose posterior extracapsular button noted on radiographs without clinical complication. No other complications, including changes in the graft position, joint space narrowing, or graft failures, were noted from reviewing final radiographs postsurgery.

Discussion

This case series of patients who underwent arthroscopic DCA augmentation for glenoid bone loss leading to shoulder instability demonstrated low complication rates and no evidence of recurrent instability at minimum 1 year follow-up.

Glenoid bone loss has been recognized as a critical factor contributing to recurrent anterior glenohumeral instability.³⁹ The Latarjet procedure has long been recognized as the standard of care for addressing recurrent anterior instability with anterior glenoid bone loss; however, the procedure carries a high complication rate, with high morbidity if musculocutaneous nerve complications do occur.^8,12,15,18 Consequently, alternative grafting options have been described; however, the optimal procedure remains controversial.

Although the Latarjet procedure has historically proven to be an excellent option for addressing glenoid bone loss, it is not without significant complications, including musculocutaneous nerve palsy, graft nonunion, and graft–glenoid incongruity, which can accelerate arthritic change.³ Several other bone augmentation options have been explored given the deficiencies associated with the Latarjet, including DTA. Provencher et al³³ first described this technique, which has gained popularity, particularly in patients with significant bone loss or in the revision setting. Bhatia et al³ demonstrated improved glenoid contact area with DTA when compared to the Latarjet in terms of increased contact area and decreasing peak force. As a result, studies show good functional outcomes and similar effectiveness in restoring glenoid bone stock when compared to the Latarjet.¹³ However, DTA is associated with significantly higher cost and resorptive rates compared to the Latarjet.⁴⁰

The DCA bone-block procedure has garnered increasing interest as an easier, more cost effective, and safer alternative to current options. The distal clavicle presents a safe and easily accessible osteochondral autograft that is capable of comparably reconstituting glenoid bone loss with greater versatility in restoring larger glenoid bone defects compared to other grafts,^21,29,34 as further evidenced in this series where the average glenoid bone loss was 27%. Furthermore, as opposed to allografts, another advantage of autografts includes the lack of antigenicity, decreased infection risk, and increased rate of graft healing and incorporation.¹ Compared to fresh osteochondral allografts, which can cost thousands of dollars with often prolonged wait times, the DCA is readily accessible both in availability and cost.

Furthermore, DCAs have been shown to be noninferior to traditional Latarjet with regard to glenoid width and surface area restoration.⁶ Several studies have shown that distal clavicle bone blocks can effectively restore glenoid bone loss while restoring the articular cartilage. In addition to having similar contact areas and pressures within the shoulder when compared to coracoid grafts, distal clavicle bone blocks have been shown to have more versatility when compared to other grafts.²¹ In this study, maximization of the superior-to-inferior coverage was attained by using the AP surface (23 mm in this series) against the glenoid defect, as opposed to using the medial–lateral surface, which is shorter (10-12 mm based on resection). The optimal graft orientation remains unclear; however, this would be an important variable to further evaluate clinical outcomes compared to the Latarjet and iliac crest bone graft.

Drawbacks of the DCA includes risk of damage to the CC ligaments and injury to the AC joint when harvesting the autograft. There remains a paucity of data in the current literature that points to conclusive evidence regarding the amount of clavicle resection that produces AC joint instability. In a biomechanical cadaveric study, Beitzel et al² noted an increase in translation with a 10-mm distal clavicle resection despite having an intact AC joint capsule. We believe that this risk can be mitigated by harvesting the clavicle no more than 12 mm medial to the joint, as a more medial resection may risk injury to the CC ligaments, which attach approximately 16 mm medial to the AC joint and span 26 mm further medially, according to the cadaveric study by Chahla et al.⁹ In addition, the graft should be peeled off the AC capsule to prevent any capsular damage that may prevent an adequate repair of the joint capsule, therefore preventing instability.¹¹ In addition, the technique described in the present study may confer a secondary capsulorrhaphy-like effect, which could contribute to capsular tightening and a further reduced risk of future instability. Using this approach, we did not note any patients in our series to suffer any AC joint injuries. However, as with the Latarjet procedure where injury to the CC ligament can occur with overly aggressive coracoid resections,⁹ AC joint instability is a risk, and the surgeon must take great surgical care to avoid this.

Previous techniques for distal clavicle bone graft augmentation have been described by Kwapisz and Boileau.^4,22 Both techniques describe a technique similar to the one used in this study of placing the undersurface of the distal clavicle along the glenoid to restore the congruent arc of the glenoid. Rather than screw fixation, which has been associated with hardware complications, we utilized the suture button technique, which can provide high compression force over a relatively large surface area and affords the advantages of smaller drill holes and consequently a lower risk of fracture.^10,14 Suture button fixation also allows for more flexibility, particularly when trying to avoid the subscapularis, compared to screw fixation. Furthermore, as noted in one of our patients who had a suture button complication without clinical significance, screw complications may be more likely to present more significant patient morbidity compared to buttons.^16,25 Efficacy of button fixation has been demonstrated; for example, Xu et al⁴¹ noted a 98% union rate among 102 patients that underwent arthroscopic Latarjet fixed with suture buttons. We recommend having 2 points of fixation by using 4 suture buttons as opposed to single-point fixation with 2 buttons, to increase compressive force, reduce rotational displacement, and reduce failure risk should one suture button fail. While its use during allograft fixation has been documented,^17,28 it is unclear if suture buttons fixation has similar efficacy in autografts or allografts. In addition, further research is needed to understand if performance changes as the grafts increase in size.

This study is not without limitations, which include first its retrospective nature and limited number of patients. This small case series represents a select subset of patients with shoulder instability and critical bone loss who underwent glenoid reconstruction with DCA augmentation with suture button fixation. This sample may not be applicable to all patients, such as young contact athletes. While no patients were noted to have clinical or radiographic complications, we did not employ postoperative CT scans due to cost concerns and therefore cannot conclusively comment on union rates. Furthermore, our follow-up period may be insufficient to fully capture final rates of recurrent instability and longer-term studies are needed to fully ascertain clinical outcomes. Finally, this series consists of patients operated on by a single surgeon and may not be generalizable to all surgeons. Further research is warranted to further understand this treatment paradigm and its efficacy, including patient-reported outcomes and further investigation with post-operative CT imaging to evaluate for any degree of remodeling and union rates.

Conclusion

In this case series of patients who underwent arthroscopic DCA augmentation for glenoid bone loss leading to shoulder instability, we noted low complication rates and no evidence of recurrent instability.

Disclaimers:

Funding: No outside funding or grants assisted in the study.

Conflicts of interest: Jonathan Bravman is a board or committee member for American Orthopaedic Society for Sports Medicine, receives research support from Biomet and Stryker, is a paid consultant for DJ Orthopaedics and Smith & Nephew, receives other financial or material support from Mitek, Smith & Nephew, and Stryker, and receives IP royalties and is an unpaid consultant for Shukla Medical. Robert T. Tisherman receives an educational grant through Arthrex. Christopher Klifto is a consultant for Smith and Nephew, Styker, Restore3D and Acumed. Oke Anakwenze is a consultant and receives royalties and educational support from Exactech, is a shareholder of SutureTech, consultant for Johnson & Johnson, JSES board member, and associate editor of JSES RRT. The other authors, their immediate families, and any research foundation with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.