Abstract
Background:
The use of isolated soft tissue repair versus bone block stabilization for the treatment of recurrent anterior shoulder instability in adolescents has no scientific evidence.
Purpose:
To compare the clinical outcomes of adolescent patients who underwent isolated arthroscopic Bankart (iB) repair with those who underwent the arthroscopic Bristow-Latarjet procedure in addition to Bankart (BLB) repair.
Study Design:
Cohort study; Level of evidence, 3.
Methods:
A total of 60 shoulders in adolescents (aged 13-18 years) were reviewed with a minimum 2 years’ follow-up: iB repair (n = 36) and arthroscopic Bankart repair with an additional Bristow-Latarjet procedure (BLB; n = 24). The characteristics of the patients in each group in terms of age at the first instability episode, age at surgery, hyperlaxity, participation in at-risk sports, and Instability Severity Index Score were comparable. The mean follow-up was longer in the iB group (7.7 vs 4.1 years, respectively), whereas the rates of patients engaged in competition and those with glenoid lesions were higher in the BLB group. The primary outcome measures were failure, defined as the recurrence of instability (clinical dislocation or subluxation), and return to sports. The mean follow-up was 6.2 years (range, 2-16 years).
Results:
At the last follow-up, the rate of recurrence was significantly higher in the iB group, with 22% (8/36) failures, than in the BLB group, with 8% (2/24) instability recurrences (P < .05). The rate of return to sports at the same level was significantly higher after the BLB repair than after iB repair (79% vs 47%, respectively; P < .001). No statistical difference was found in patient-reported outcome scores between treatment groups (P > .05). Although failures occurred early after the BLB repair, 88% of failures after iB repair occurred after 2 years. On multivariate analysis, adolescents in the iB group with >3 episodes of preoperative dislocation and shoulder hyperlaxity (external rotation >90°) had a 60% recurrence rate (P < .005).
Conclusion:
Adolescent patients undergoing the BLB repair had a lower rate of recurrent instability and higher rates of return to sports and competition than those undergoing iB repair. Patients with shoulder hyperlaxity (external rotation >90°) and >3 dislocations had an unacceptable failure rate of 60% after iB repair.
Keywords
Shoulder instability is an increasingly common problem in the adolescent population (aged 13-18 years), as participation in at-risk sports (collision or contact-overhead) continues to rise.46,48 The high failure rates reported with shoulder stabilization of traumatic recurrent anterior shoulder instability in adolescents practicing sports are a concern.48,49 After a first episode of anterior dislocation or subluxation, >80% of adolescents have a recurrent dislocation, particularly during competitive and contact or collision sports.16,23,33,39,42-44 The lower incidence of dislocations in adolescents <14 years can be explained by the specificity of lesions such as humeral physeal fractures, anterior labral periosteal sleeve avulsions, or a combination of these lesions.1,20 For patients over 15 years of age, it is necessary to adapt the therapeutic strategy, unlike children under 15, for whom specific lesions should not be overlooked.20,28
Chronic anterior instability is characterized by humeral and glenoid bone lesions of varying severity, which have a direct effect on the type of surgical treatment required. Basically, there are 3 main surgical options to stabilize the shoulder in adolescents with recurrent instability who participate in high-risk (contact or collision) sports: isolated suturing of the labrum (Bankart repair), 49 Bankart repair with Hill-Sachs remplissage, 13 or a bone block procedure (Bristow-Latarjet procedure with coracoid transfer, anterior bone block autograft/allograft) with or without Bankart repair.11,49 The latter option, under open surgery, has a recurrence rate in adults of between 0% and 13%, depending on the type of fixation (screw or suture button) or the position of the bone block (standing [Bristow] or lying [Latarjet]).5,6,18,19,26,27,30,34 The rate of instability recurrence in adults reported in the literature is higher with Bankart repair only, ranging from 11% to 25%.16,17,21,32,36,38,40 Results are satisfactory in adults when glenoid and humeral bony lesions are small (“on track”) 25 and the Instability Severity Index Score is <3 points. 47 In a recent study that considered the recurrence of instability and anterior apprehension as failure, 31% of adolescents aged 12 to 18 years had failed results after isolated Bankart (iB) repair, with a medial Hill-Sachs lesion (large Hill-Sachs interval) as a risk factor for failure. 28
Shoulder stabilization under arthroscopic surgery started being performed in the 1990s through labral repair with capsular shift using suture anchors (iB repair).9,15 In the early experience of Boileau et al, 13 arthroscopic iB repair, being minimally invasive, was the main procedure performed for the treatment of recurrent anterior shoulder instability in adolescents practicing sports. In 2010, Boileau et al11,12 developed a guided technique to perform the arthroscopic Bristow-Latarjet procedure with Bankart (BLB) repair. This combined procedure became the treatment of choice for stabilizing an adolescent's shoulder in which bony lesions and hyperlaxity were present. A cannulated screw was first used for fixation of the coracoid bone block onto the scapula, and later, a suture button was developed and preferred. There is inadequate evidence to suggest which procedure should be offered to an adolescent athlete: iB repair or BLB repair.
The purpose of this study was to compare the outcomes in adolescents with recurrent shoulder dislocations who underwent iB repair versus BLB repair. Our hypothesis was that adolescent patients undergoing BLB repair would have a lower rate of recurrent instability and higher rates of return to sports and competition than those undergoing iB repair.
Methods
Study Design
All adolescent patients (aged 13-18 years at surgery) with recurrent anterior shoulder instability treated with either iB repair or BLB repair by a single shoulder surgeon (P.B.), with a follow-up of >2 years, were included (study approval No. ICR-2020-SA-12-3). We excluded adolescents with voluntary instability, those with multidirectional instability, those with failed stabilization surgery, those with underlying connective tissue abnormalities (Ehlers-Danlos syndrome), those whose medical history revealed a neurological problem (epileptic seizures, neuromuscular disorders), and those who had a psychiatric disorder (ie, had been diagnosed with a psychiatric disorder by a psychiatrist and/or was being treated with an antidepressant or anxiolytic drugs). All included patients had undergone failed nonoperative treatment and rehabilitation for ≥6 months before surgery. Nonoperative treatment was conducted by a physical therapist, and the patients were followed up twice within 6 months by the referring clinician.
Patients
A total of 70 consecutive adolescent patients with recurrent anterior instability met the inclusion criteria and underwent either iB repair or BLB repair between 1998 and 2017. Overall, 10 patients were either lost to follow-up or had missing data, leaving 60 shoulders for the present study. iB was performed in the first 36 shoulders, whereas the next 24 shoulders underwent BLB.
Patient data and computed tomography (CT) findings of bipolar bone loss are presented in Table 1. The 2 groups were comparable for age at the first instability episode, age at surgery, shoulder hyperlaxity, participation in at-risk sports, and Instability Severity Index Score. 3 However, the rates of male patients, those engaged in competition, and those with glenoid erosion (>15%) were higher in the BLB group. The mean follow-up was 6.2 years (range, 2-16 years). The mean follow-up was significantly longer in the iB group than in the BLB group (7.7 vs 4.1 years, respectively; P < .001).
Patient Characteristics a
Data are presented as mean ± SD, mean (range), or n (%) unless otherwise specified. NS, not significant; SSV, Subjective Shoulder Value.
Surgical Technique
All patients underwent surgery under general anesthesia and locoregional anesthesia (interscalene block) in the lazy beach-chair position, with the consent of their legal guardians. For the iB group, a minimum of 3 absorbable anchors were placed between the 3- and 6-o’clock positions at the anterior part of the glenoid, starting with the most inferior one.9,15 The temporary outside traction suture technique was used for patients operated on after 2002 (n = 26) with the same anchor system. 9 For the BLB group, the procedure was performed in 5 steps as described for the original technique.10-15 Fixation of the bone block was performed in the lying position using a cannulated screw (n = 14) for the Latarjet procedure or in the standing position using 1 suture button (n = 10) for the Bristow procedure. The fifth stage corresponded to Bankart repair and capsular shift with 2 anchors at the 3- and 5-o’clock positions.
Postoperative Management
Postoperative management was identical for both groups of patients. Patients were immobilized in an internal rotation sling for 4 weeks from 1998 to 2010 (n = 45) and in a neutral rotation sling (n = 15) for 4 weeks month after 2010. At 30 days after surgery, physical therapy was started in combination with balneotherapy and without elastic or strength training, with return to sports allowed after 3 to 6 months postoperatively.
Outcome Assessment
All patients were assessed by 1 observer (O.R.), independent of the operating surgeon (P.B.), with a questionnaire that included satisfaction (rated as very satisfied, satisfied, dissatisfied, or very dissatisfied), the Subjective Shoulder Value (SSV) score 24 and the SSV for Sport (SSV–Sport) score,6,24 return to sports, competitive level, pain evaluated on a visual analog scale, range of motion, the Western Ontario Shoulder Instability Index score, 37 and the Rowe score. 45 The recurrence of instability was defined as ≥1 episodes of anterior dislocation or subluxation. Persistent anterior apprehension was defined as a “fear that the humeral head would come out of [the] joint with the arm placed in the throwing position (abduction–external rotation).” Patients completed the questionnaire themselves during a clinical assessment or a telephone interview for subjective data.
Radiological Assessment
Preoperatively, we evaluated glenoid bone loss on CT with the best-fit circle technique, in the en face view,10,31 and with the Barchilon method. 2 The Instability Severity Index Score was established using clinical and radiological criteria. 3 Bone block positioning and healing were analyzed at final follow-up on radiographs and CT images. The glenoid was examined in axial and sagittal sections: 3 successive slices in each plane were retained to analyze the position of the bone block.10,31 The ideal position of the bone block was therefore defined as a flush position in the horizontal plane and between 3 and 6 o'clock in the vertical plane. The coracoid bone block was considered too lateral if it was 5 mm behind the glenoid and too medial if it exceeded the edge of 5 mm. The graft was considered subequatorial if the bone block was positioned between 3 and 6 o'clock, equatorial if the bone block was ≤25% over the equator, and above equatorial if >50%.10,31 In addition to bone healing, we looked for possible coracoid fractures, migration, or osteolysis.4,31
Statistical Analysis
The Shapiro-Wilk test was used to analyze the distribution of data. The paired t test compared normally distributed data, and the Mann-Whitney U test compared nonnormally distributed data. The chi-square test was used to compare categorical data. If any cell contained ≤3 numbers, we used the Fisher exact test. Multivariate analysis was performed to assess the association between recurrence and the factors studied. The significance level was set at a P value of <.05. We performed statistical analyses using XLSTAT software (Addinsoft).
Results
Complications and Reoperations
The overall failure rate, corresponding to the recurrence of instability, was 17% (10/60). In the iB group, 8 patients (22%) had a recurrence of anterior instability (4 dislocations and 4 subluxations), at a mean follow-up of 7.7 years, and <10% glenoid bone loss on preoperative CT. The 4 patients with a dislocation were revised by BLB repair and had no failures at final follow-up, while those with a subluxation were treated with muscle strengthening and physical therapy.
In the BLB group, 2 patients (8%) had a recurrence of anterior instability, at a mean follow-up of 4.1 years, due to 2 migrations of the bone block after a probable perioperative fracture after BLB repair (1 screw and 1 suture button). The 2 patients were revised with the arthroscopic Eden-Hybinette procedure (iliac crest bone graft) with Bankart repair and concomitant Hill-Sachs remplissage. They both had a stable shoulder at the last follow-up. Another patient presenting with an infection required a reoperation under arthroscopic surgery for washout and screw removal, followed by an antibiotic treatment; at the last follow-up, the shoulder was stable, and the patient was satisfied.
Clinical and Subjective Outcomes
At the last follow-up, the rates of instability recurrence and apprehension were significantly higher after iB repair than after BLB repair, and more adolescent athletes returned to the same level of sports in the BLB group than in the iB group (Table 2). Functional scores and range of motion (external rotation limitation) measurements showed no significant difference between the 2 groups. On multivariate analysis, patients who underwent iB repair with >3 episodes of preoperative dislocations and shoulder hyperlaxity (external rotation >90°) had a 60% recurrence rate (P < .005). No difference was found in terms of instability recurrence or return to sports between patients younger or older than 15 years in the present series. No statistical difference was found in patient-reported outcome scores between treatment groups (P > .05).
Clinical and Subjective Outcomes a
Data are presented as mean ± SD, mean (range), or n (%) unless otherwise specified. NS, not significant; SSV, Subjective Shoulder Value.
Failure
The 2 cases of failure in the BLB group occurred before 1-year follow-up, with no clear risk factors identified. However, 88% of recurrences in the iB group occurred after the first year. As mentioned above, 2 risk factors were found: hyperlaxity (P = .01) and >3 preoperative subluxations (P = .01). On multivariate analysis, the rate of failure was 60% in the iB group when these 2 factors were present (Figure 1).

Survival diagram: failure over time, showing that, unlike isolated arthroscopic Bankart (iB) repair, all recurrences after BLB occurred within the first year of follow-up. BLB, arthroscopic Bristow-Latarjet procedure with Bankart repair; FU, follow-up; Postop, postoperative.
Radiological Findings
No difference in bone block positioning was found between the screw and suture button groups in either the horizontal or the vertical plane on postoperative CT (Table 3). The rate of bone healing was higher with suture button fixation than with screw fixation (80% vs 57%, respectively; P = .04). The rate of osteolysis was greater in the screw group than in the suture button group (50% vs 10%, respectively; P = .01).
Radiological Findings in the Patients Undergoing BLB Repair a
Data are presented as n (%). NS, not significant. b Only the group of bone block.
Discussion
The results of the present study demonstrate that BLB repair was a more effective surgical option for adolescents with recurrent anterior shoulder dislocations than iB repair. Adolescent patients undergoing BLB repair had a lower rate of recurrent instability and higher rates of return to sports and competition than those undergoing iB repair. The rate of instability recurrence was almost 3 times higher after iB repair (22%) than after BLB repair (8%). The rate of return to sports at the same level was significantly higher after the BLB repair than after iB repair (79% vs 47%, respectively; P < .001). Although the 2 groups were not matched, the patients were comparable for age at the first instability episode, age at surgery, hyperlaxity, participation in at-risk sports, and Instability Severity Index Score. With the follow-up being shorter in the BLB group than in the iB group, we acknowledge that the rate of recurrence may increase after BLB repair in the future. However, this risk appears to be low: similar to what we observed in adults, failures after BLB repair occurred early (during the first year), while they continued to occur late (after 2 years) after iB repair. 6 Furthermore, the risk of recurrence was theoretically higher in the BLB group with more patients engaged in competition and more patients with severe glenoid bone loss. Finally, patients with bipolar bone lesions seemed to require BLB repair.
The Bristow-Latarjet procedure is considered to be a nonanatomic procedure because of tenodesis of the conjoint tendon, which acts as a sling to stabilize the humeral head in the glenoid socket (sling effect). 51 This procedure is thought to limit postoperative range of motion (particularly external rotation) and strength, which may be a concern for these patients to return to athletics.8,29 In our study, no significant difference was found in postoperative range of motion between the iB and BLB groups. Our data did not demonstrate any significant difference in the amount of final external rotation motion asymmetry between patients who had undergone the BLB procedure and those who had undergone iB repair. The decreased range of motion in both groups in the current study was low and did not appear to negatively affect the patients’ perceived motion and return to sports, including throwing and contact-overhead sports. The absence of any reduction in clinically relevant external rotation or in the rest of the range of motion is consistent with recent literature.7,48
Several risk factors for the recurrence of shoulder instability after iB repair have been reported in the literature. 3 The number of episodes of preoperative instability was found to be a risk factor for failure of arthroscopic stabilization. 22 Shoulder hyperlaxity, defined as external rotation above 90° with the elbow at the side, has also been recognized as a risk factor for failure after arthroscopic Bankart repair. 15 In our study, we found that patients who underwent iB repair and who had >3 episodes of preoperative dislocations together with shoulder hyperlaxity had an unacceptable risk of recurrence of 60%. Age, sex, type of sport, level of sport, and glenoid and/or humeral bone loss were not found to be risk factors for recurrence. On the basis of this finding, we suggest that adolescent patients with these risk factors (external rotation with the elbow at the side >90° and >3 dislocations) should be contraindicated for iB repair.
Young age and sports practice, specifically in competition, are other well-known risk factors for recurrent instability after iB repair. 3 In a series of anterior shoulder instability in adolescents, with a population comparable with that in our study, Egger et al 28 found a failure rate of over 30% after iB repair. Subjective instability (feeling of apprehension) was considered a failure, while 17% of patients were revised for a recurrence of clinical instability. In our series, the overall failure rate of iB repair was 22%, with 11% of patients undergoing revision for recurrent instability. The high recurrence rate after iB repair (22% at a mean follow-up of 7.7 years) is consistent with that reported in the literature.16,21,32,36,40,50
Promising outcomes have been reported after arthroscopic Bankart repair with remplissage in adolescent patients with large Hill-Sachs lesions without significant glenoid loss. 35 Hughes et al 35 assessed the outcomes of arthroscopic Bankart repair with and without Hill-Sachs remplissage for the treatment of recurrent anterior shoulder instability in adolescents. Patients were matched for age, sex, and size of the Hill-Sachs lesion. A total of 21 adolescents underwent a remplissage procedure, and 20 matched controls underwent only Bankart repair. The authors found a significantly higher rate of recurrence in the Bankart repair group (8/17) compared with the remplissage group (2/15) (P = .04). As in our study, no statistical difference was found in patient-reported outcome scores between treatment groups or in range of motion measurements. The remplissage technique was not used in our series; thus, it cannot be evaluated. It appears to be a surgical option in the absence of bone loss in the glenoid associated with Bankart repair.
Our data suggest that the Bristow-Latarjet procedure remains one of the best surgical techniques to address recurrent anterior instability with significant glenoid bone loss. 14 The rate of recurrence after BLB repair in our adolescent population was low (8%), despite 75% of shoulders having glenoid bone lesions as well as 79% and 88% of the adolescents participating in high-risk sports and/or competition respectively.
In our study, coracoid transfer was performed arthroscopically, which has never been published on in an exclusively pediatric/adolescent population. To our knowledge, this is the first study to compare iB repair and BLB repair in a population of adolescent athletes with a midterm to long-term follow-up. This study has several limitations, including the retrospective study design and lack of randomization. One should consider this study a historical series including surgical techniques that have progressively evolved over time. This study is a witness to the evolution of the arthroscopic techniques developed by Boileau et al.10-14 Initially, only iB repair was possible. With the development of guiding instruments and the improved skills of the surgeon, it has become possible to perform BLB. Similarly, the fixation technique for the Bristow-Latarjet procedure has evolved from cannulated screws to suture buttons, which are much easier and safer in adolescents, who often have small coracoids (specifically young female patients).41,50 Overall, 10 patients were lost to follow-up (14% of the cohort); this can be explained by the characteristics of this young and mobile population, which limit the possibilities of long-term clinical follow-up. The study results are limited by the small sample size, and a larger cohort will be needed to determine significant correlations regarding outcomes. However, the significant differences in the rates of recurrence and return to sports indicate that the type II error (power) was acceptable for these variables. As mentioned above, the duration of follow-up differed between the groups, with a longer follow-up in the iB group than in the BLB group. Although this could be seen as disadvantageous for Bankart repair, one should consider that failures after iB repair were spread out over time, unlike failures after BLB repair, which occurred in the first year.
Conclusion
In a population of adolescent athletes with high functional demands, the addition of the Bristow-Latarjet procedure to Bankart repair provided better shoulder stability and higher rates of return to sports and competition than iB repair. Patients with >3 preoperative episodes of dislocations and shoulder hyperlaxity (external rotation >90°) had a 60% rate of failure and should be contraindicated for iB repair. In adolescent patients treated with BLB repair, suture buttons appeared to be a reliable means of fixation with a high rate of bone block healing and lower rates of fractures and osteolysis than with screw fixation. The use of an arthroscopically guided technique for the Bristow-Latarjet procedure resulted in correct bone block positioning.
Footnotes
Submitted October 16, 2023; accepted January 24, 2024.
One or more of the authors has declared the following potential conflict of interest or source of funding: P.B. has received consulting fees from Smith & Nephew and support to develop the specific instruments and implants used to perform the discussed procedure. AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD and disclaims any liability or responsibility relating thereto.
