Comparison between all-suture and biocomposite anchors in the arthroscopic treatment of traumatic anterior shoulder instability: A retrospective cohort study

Abstract

Purpose

Suture anchors have revolutionized arthroscopic surgery, enabling direct soft tissue-to-bone repair. There are many types of anchors still used in arthroscopic shoulder operations. We sought to compare the clinical outcome of all-suture and biocomposite anchors when used in arthroscopic Bankart repair for patients suffering from anterior shoulder instability.

Methods

A single-center retrospective cohort study of 30 patients (mean age: 26.6 years, SD: 8.8 years, male/female ratio: 5/1, mean follow up: 28 months, SD: 23.8, range: 12–92) with anterior shoulder instability was conducted. Patients were divided into 2 groups based upon the type of suture anchors used for the Bankart repair: group A (14 patients) used only all-suture anchors and group B (16 patients) used only biocomposite anchors. Outcomes reported were postoperative dislocations, positive shoulder apprehension test, self-reported sense of shoulder instability, return to activities of daily living, return to sports, patient satisfaction and complications. Patient reported outcome measures (PROMs) used were the Rowe Score for Instability, Constant Shoulder Score, Walch Duplay Score, The American Shoulder and Elbow Surgeons (ASES) Shoulder Score, Oxford Shoulder Instability Score and external rotation at 90° of arm abduction, external rotation at 0° of arm abduction, forward flexion, abduction, adduction and internal rotation.

Results

Rates of postoperative shoulder dislocation demonstrated no significant difference between the 2 groups (p > .05). Four postoperative dislocations happened, two in each group (14.3% and 12.5% for all suture only and biocomposite only groups, respectively), with three of them being traumatic. In addition, no significant differences were observed amongst groups regarding shoulder apprehension test (group A: 85.7% vs. group B: 93.8%), sense of shoulder instability (7.1% vs. 6.3%), return to activities of daily living (group A: 85.7% vs. group B: 93.8%), return to sports (group A: 85.7% vs. group B: 87.5%), patient satisfaction (moderate level: group A 21.4% vs. group B 12.5%), and PROMs.

Conclusion

The short-term failure rate and clinical/functional outcomes of arthroscopic Bankart repair using all-suture anchors is similar to the use of biocompatible anchors.

Level of evidence

Level III, retrospective cohort study.

1. Introduction

Several types of anchors have been used for the arthroscopic repair of anterior shoulder instability. Based on the type of material, these can be stratified in four different types: metallic, biodegradable, biocomposite and all-suture anchors, none of which is considered a golden standard and all are still used in shoulder arthroscopy.1, 2, 3, 4, 5 Historically, the first anchors used were metallic, which had great resistance to pullout strength.^1,6 Metallic anchors correlated with high risk of suture's failure, bone loss in the glenoid due to their bigger size, cartilage damage and signal alteration in postoperative MRI, but newer generation metal anchors are still in use in arthroscopic shoulder surgeries.^2,4,7, 8, 9

Biodegradable anchors combined similar fixation strength to metallic anchors with less signal interference in postoperative MRI.^2,7,10,11 However, biodegradable anchors were associated with high rate of breakage during insertion, rapid anchor degradation, anchors’ migration, cystic formation in the glenoid bone, soft tissue aseptic inflammatory reaction, and glenoid bone osteolysis, but new materials have been introduced and they are still in wide use in shoulder arthroscopy.^2,7,10,12, 13, 14, 15, 16

To bridge the time gap from initial anchor fixation of a biodegradable anchor to bone formation as it dissolved, biocomposite anchors were introduced.^2,17,18 Containing a bioabsorbable polymer-based part (e.g. hydroxyapatite) and an osteoconductive bio-ceramic part such as b-tricalcium phosphate or polyether ether ketone (PEEK) they managed to eliminate early degradation and preserve the advantages of bioabsorbable anchors thanks to the bio-ceramic part, which provides a basis of bone formation.^2,3,18,19 They are introduced in shoulder arthroscopy around 2010, they are still in use and low rates of bone cyst formation, tunnel widening and inflammation reaction were reported.^2,20, 21, 22, 23, 24, 25

At around the same time, another type of anchors was introduced consisting of suture-only material.^1,2,26 The absence of rigid components was considered to minimize invasiveness and eliminate some of the complications of metallic anchors, while their smaller size and smaller pit hole needed made it easier to provide more anchorage points and benefited shoulder revision surgery if required, respectively, compared to all other anchor types.^1,20,27 Biomechanical studies on types of all-suture anchors showed similar ultimate load to failure and displacement at ultimate failure and less osteolysis between some of the all-suture,biodegradable and biocomposite anchors, though there was variance in strength and performance between all-suture anchor types.^20,27, 28, 29, 30, 31, 32 Some reports of tunnel widening and displacement created questions about their in-vivo properties.^1,28,29

Despite promising data from preclinical studies suggesting the clinical use of all-suture anchors, there is still a lack of clinical studies comparing all-suture anchors to older types of anchors. Therefore, we sought to compare dislocation, reoperation and complication rates between all-suture anchors and biocomposite anchors when used in arthroscopic Bankart repair for patients suffering from anterior shoulder instability.

We hypothesized that the clinical and functional outcomes of all-suture anchors when used in arthroscopic Bankart repair are similar to biocomposite anchors.

2. Materials and methods

This is a single-center retrospective cohort study based on prospectively collected data from January 2011 until January 2019.

Inclusion criteria were: (1) patients of any age with at least one radiologically confirmed episode of anterior shoulder dislocation; (2) patients treated with all-arthroscopic Bankart repair either with or without Remplissage; (3) patients whose Bankart repair was performed either with the use of biocomposite anchors or all-suture anchors; (4) patients with at least 12 months of clinical and functional follow-up, (5) patients who underwent soft tissue procedures.

Exclusion criteria were: (1) Patients with glenoid bone loss >25% evaluated preoperatively by MRI; (2) Patients with <12 months follow-up (3) Patients who received arthroscopic Bankart repair (with or without Remplissage) with metallic, biodegradable anchors or both biocomposite and all-suture anchors; (4) Patients with other-than-anterior shoulder instability (posterior, multidirectional instability); (5) Patients treated with open surgical procedure for anterior shoulder instability.

Based on the predefined criteria, by searching in our database with the key words: “Instability”, “Dislocation”, “Bankart” we identified 30 eligible to be included in the final cohort of this study (Fig. 1). The patient cohort consisted from 2 groups based upon the type of anchors used: (1) group A (14 patients) which used only all-suture anchors, and (2) group B (16 patients) which used only biocomposite anchors.

Fig. 1.

Modified CONSORT flow chart.

Shoulder function was examined preoperatively by an experienced shoulder surgeon in our institution. Preoperative radiological imaging with MRI scan was ordered. After standard clinical examination and imaging, preoperative data were collected, including: date, age of patient, gender, side of the dislocated shoulder and hand side dominance, clinical evidence of loose joints, overhead profession (or not), level of sport activity before the first dislocation or between the episodes (none, recreational, competitive, professional), number of prior dislocations, age at first dislocation, period of instability (time period between the first confirmed dislocation and time of surgery).

Data from surgery included: type of surgery, concomitant lesions, type, brand name and number of anchors used and type, brand name and number of anchors used for remplissage.

2.1. Surgical technique

The repair of anterior shoulder instability was performed arthroscopically according to standard procedure for labral repair.³³ Patients were positioned in lateral decubitus position using traction on the involved arm and surgery was performed under general anesthesia. A posterior portal was first created for camera insertion and diagnostic arthroscopy was performed to visualize Bankart lesion and identify any associated pathology. Afterwards, an anterior (antero-superior) and an antero-inferior portal were created. Capsulolabral lesions were repaired using anchors, either only biocomposite or only all-suture- depending on the judgement of the surgeon. Depending on the degree and the location of labral tear, the points of anchor insertion were chosen. For reducing Bankart lesions, anchors were placed from inferior to superior direction, with the inferiormost one being double loaded, while for SLAP lesions were positioned postero-superiorly.

Arthroscopic remplissage was performed additionally to the Bankart repair when other co-lesions existed, such as engaging Hill-Sachs.^34,35 This procedure combined posterior capsular plication and infraspinatus tenodesis within the Hill-Sachs lesion, thus making the bony defect extra-articular resulting in decreasing the possibility of recurrent instability.³⁴ Specifically, in the case of engaging Hill-Sachs, the lesion was firstly abraded, next, anchors were inserted in the lesion, sutures were passed through the posterior capsule and the infraspinatus tendon and firmly tied the myo-capsular complex to the humeral head. All anchors used in remplissage procedure were double loaded, either all-suture (in group A) or biocomposite (in group B).

2.2. Outcomes

Postoperative evaluation was documented at the last follow-up including: recurrence of shoulder dislocation after the procedure and postoperative month of recurrence of dislocation (if happened), self-reported sense of instability, return to average daily activity, level of return to sports compared to pre-dislocation level, patients satisfaction level, shoulder apprehension test, external rotation at 90° of abduction, external rotation at 0° of arm abduction, forward flexion, abduction, adduction, internal rotation The Rowe Score for Instability,³⁶ Constant Shoulder Score,³⁷ Walch Duplay Score,³⁸ The American Shoulder and Elbow Surgeons (ASES) Shoulder Score,³⁹ Oxford Shoulder Instability Score.⁴⁰

2.3. Postoperative rehabilitation

Following surgery, all shoulders were protected and immobilized in a sling for 4–6 weeks. During this time, passive pendulum exercises were initiated and active mobility maintenance exercises for the accessory joints – hand, wrist, elbow-were allowed and encouraged. At 15 days postoperative, active assisted exercises were started by gradually increasing the range of motion (ROM), avoiding external rotation until 4 weeks postoperatively. At the beginning of 5th week, use of sling was recommended only during sleep and in crowds and at 6–8 weeks postoperatively strengthening exercises were started, with isometrics first, avoiding terminal external rotation and abduction positions. Overhead activities were allowed after 3 months, return to non-contact sports after 4–5 months and close contact activities 6 months after surgery. During rehabilitation time both supervised physical therapy and home exercise programs were implemented.⁴¹

2.4. Statistical analysis

Data analysis was performed with IBM SPSS Statistics software (Version 25.0, Armonk, NY: IBM Corp). All p values were evaluated and compared with a significance level of 5%. Standard Deviations, means, and t-tests (or Mann Whitney U test for non-parametric variables) were used for continues and chi-square tests for the analysis of cross tables, absolute and relative frequencies for categorical variables.

3. Results

The power analysis of this study was 80%. No significant difference was found between the two groups regarding the baseline demographic, pre- and intraoperative characteristics, either using independent sample t-test (age, period of instability, number of previous dislocations, total anchors used, total single loaded, total double loaded, double loaded anchors used in remplissage) or Chi square (gender, side of the dislocated arm, hand dominance, sports activity before the first dislocation, clinical evidence of loose joints, complementary arthroscopic remplissage) (Table 1).

Table 1.

Baseline Characteristics of patient groups.

Baseline Characteristics	Group A (All-Suture Only)	Group B (Biocomposite Only)	P value
Demographic Characteristics
Gender			.743
Male	12 (85.71%)	13 (81.25%)
Female	2 (14.29%)	3 (18.75%)
Age At Surgery [Mean ± SD, In Years]	26.21 ± 10.66	26.94 ± 7.21	0.313
Side			0.151
Left	6 (42.9%)	3 (18.75%)
Right	8 (57.1%)	13 (81.25%)
Dominant Limb Affected			0.156
Yes	7 (50%)	12 (75%)
No	7 (50%)	4 (25%)
Sports Activity			0.132
No	0 (0%)	2 (12.5%)
Recreational	9 (64.3%)	13 (81.25%)
Competitive	3 (21.4%)	1 (6.25%)
Proffesional	2 (14.3%)	0 (0%)
Loose Joints			0.743
Yes	2 (14.3%)	3 (18.85%)
No	12 (85.7%)	13 (81.25%)
Period Of Instability [Mean ± SD, In Years]	3.93 ± 6.50	6.00 ± 4.50	0.077
Number Of Previous Dislocations [Mean ± SD]	2.85 ± 2.97	6 ± 5.99	0.062
Total Anchors Used [Mean ± SD, In Number]	3.71 ± 1.14	4.31 ± 1.35	0.294
Single Loaded	0.86 ± 1.1	0.69 ± 0.6	0.951
Double Loaded	2.93 ± 1.64	3.56 ± 0.84	0.473
Additional Surgical Interventions
Remplissage			0.143
No	9 (64.3%)	6 (37.5%)
Yes	5 (35.7%)	10 (62.5%)
Double Loaded Anchors [Mean ± SD, In Number]	1.6 ± 0.55	1.1 ± 0.32	0.129
Single Loaded Anchors [Mean ± SD, In Number]	0	0	1

SD: standard deviation.

The mean follow-up of the patient cohort was 30.8 months (SD: 24.3, 95% Confidence interval [CI]: 21.8–39.9). The mean age of the patients at the date of the surgery was 26.6 years (SD: 8.8, 95% CI: 23.3–23.9). The female-to-male ratio was 1/5. The mean number of anchors used per patient was 4.03 (SD: 1,27, 95% CI: 3.56–4.51).

The recurrence rate at last follow-up was 14.3% (2/14) for group A and 12.5% (2/16) for group B, (n.s.). One patient from group A suffered traumatic dislocation at 18 months postoperatively, while the other one suffered atraumatic dislocation at 20 months postoperatively. Regarding to group B, both patients suffered traumatic dislocations, the first one at 22 months postoperatively, while the other one suffered recurrent dislocations at 6,12, 14, 16 months postoperatively, after a falling during rehabilitation period at 6 months post-surgery. Three of those patients were lost in the follow up after their postoperative dislocations and only one patient, from group A, was treated arthroscopically in our department for bony Bankart and partial subscapularis and partial supraspinatus tear, with last follow up at eight months postoperatively.

No significant difference was observed amongst groups in the rate of negative apprehension test (group A: 85.7%, 12/14; group B: 93.8%, 15/16) showed no significant difference (n.s.).

Furthermore, no significant difference was observed between the two groups in the self-reported sense of instability, since only one patient per group reported feeling of instability in everyday life; both patients suffered postoperative dislocations.

The mean postoperative Rowe Score,³⁶ Constant Score,³⁷ Walch Duplay Score,³⁸ ASES,³⁹ and Oxford Shoulder Instability Score⁴⁰ demonstrated no significant differences amongst groups (Table 2).

Table 2.

Shoulder scores (PROMs) and comparison.

Shoulder Scores	Group A (All-Suture Only)	Group B (Biocomposite Only)	P value
Oxford Instability [Mean ± SD 95% CI]	44.5 ± 4.1 41.6–47.5	46.7 ± 3.3 44.1–49.2	0.182
ROWE [Mean ± SD 95% CI]	87.8 ± 14.2 76.9–98.7	98.3 ± 5 95–102.2	0.094
Walch Duplay Score [Mean ± SD 95% CI]	95.6 ± 5.3 91.5–99.6	97.8 ± 4.4 94.4–101.2	0.436
Constant Score [Mean ± SD 95% CI]	91.5 ± 7.1 85.6–97.4	96.9 ± 3.7 94.2–99.6	0.173
ASES [Mean ± SD 95% CI]	94.7 ± 6.1 90.3–99.1	98.5 ± 2.7 96.3–100.8	0.122

SD: Standard deviation; CI: Confidence interval; ROWE: Rowe Instability score; ASES: American Shoulder and Elbow Surgeons score.

At univariate analysis of the number of anchors used per patient in the primary outcome - recurrence rate – no significant association was found (n.s.).

The mean external rotation at 90° of arm abduction for group A was 84.6° (SD: 13.1°, 95% CI: 76.3°–92.9°) and for group B was 86.5° (SD: 8°, 95% CI: 81.7°–91.38°). (n.s.)

The mean external rotation at 0° of arm abduction for group A was 72.5° (SD: 11.8°, 95% CI: 64°–81°) and for group B was 62° (SD: 12.3°, 95% CI: 53.7°–70.3°). (n.s.)

The mean forward flexion for group A was 178.2° (SD: 4°, 95% CI: 175.5°–180.9°) and for group B was 177.1° (SD: 9.2°, 95% CI: 171.3°–182.9°). (n.s.)

The mean abduction for group A was 175.5° (SD: 9.3° 95% CI: 169.2°–181.7°) and for group B was 175° (SD: 7.4°, 95% CI: 170°–180°). (n.s.)

The mean adduction for group A was 30° (SD: 2.9°, 95% CI: 27.3°–32.7°) and for group B was 26.3° (SD: 4.8°, 95% CI: 18.6°–33.9°). (n.s.) Fig. 2 represents the measured range of motion.

Fig. 2.

Measured range of motion.

The mean of internal rotation for group A was T9 (IQR: T7 – T10) and for group B was t8 (IQR: T7 – T12) (n.s.)

No significant differences were observed amongst groups regarding return to activities of daily living (n.s.). Specifically, all but two patients from group A (85.7%, 112/14) and all but one patient from group B returned to the same level of activities of daily living (93.8%, 15/16). From the two patients of group A, the first one suffered a postoperative dislocation, while the second one did not. The only patient of group B who did not return to the same level of activities of daily living suffered a postoperative dislocation.

No significant differences were observed amongst groups regarding return to sports (n.s.). All patients returned to the same level of sport activities, except two patients per group, who reported return to a lower level of sport activities. One of these patients per group had suffered a postoperative shoulder dislocation.

No significant differences were observed amongst groups regarding patient satisfaction (n.s.). Three patients from group A reported moderate level of satisfaction, whereas all the rest reported full or almost full satisfaction by the surgery outcome (Fig. 3). All patients reported full or almost full satisfaction, Two patients from group B reported moderate satisfaction, whereas all the rest reported full or almost full satisfaction by the surgery outcome.

Fig. 3.

Patient reported satisfaction level.

3.1. Complications

From group A, two patients reported mild complications: pain at night and at moderate daily use and mild pain at extreme positions of forward flexion and abduction. From group B, four patients reported mild complications: intermittent pain at the coracoid process, mild pain in the morning, pain at abduction at 180°, mild pain at night.

4. Discussion

The key finding of the present study was that arthroscopic Bankart repair using all-suture anchors resulted in similar short-term outcomes compared to biocomposite anchors for the treatment of anterior shoulder instability. Specifically, no difference was observed amongst groups in postoperative recurrence rates (14.3% and 12.5% for group A and B respectively) and rates of positive shoulder apprehension test (group A: 85.7%, 12/14; group B: 93.8%, 15/16)). The study primarily focused on the clinical outcome and efficacy of all-suture anchors in shoulder Bankart repair. Based on these findings, we feel that all-suture anchors could be safely used in arthroscopic Bankart repair.

To our best knowledge this is the second clinical study comparing results of arthroscopic Bankart repair between patients with only all-suture anchors versus patients with only biocomposite anchors. In a retrospective comparative study, Lee et al.¹ compared the clinical outcomes and CT analysis of tunnel diameter of 67 patients- 33 with all suture and 34 with biocomposite anchors- and reported a postoperative redislocation rate of 6% (6.1% vs 5.9%) – lower than our study-with no significant difference in recurrence rate, subjective instability, ASES and Rowe Score. They used 2 types of all-suture anchors – 1.3 mm (single-loaded) and 1.8 mm (double-loaded) and a single type of biocomposite anchor. CT analysis of the tunnel diameter increment was significantly greater with the 1.8 mm all-suture anchor than 1.3 mm all-suture anchor or biocomposite anchor, with the authors concluding that it did not influence the clinical outcome. Goschka et al.⁴² compared in a cadaveric study the biomechanical properties of all-suture anchors versus 4 other types of anchors-included a biocomposite anchor-in rotator cuff repair, dissecting the supraspinatus and repairing it with 4 different double rows. Biomechanical properties of anterior or posterior gap formation, construct stiffness, ultimate load and load to 5 mm of gap formation were evaluated, none of which showed significant difference between groups.

Similarly to our results regarding postoperative recurrence rate in biocomposite anchors, Tan et al.²⁶ found no significant difference in recurrence rate between biocomposite (13.1%) and titanium/metallic (6.3%) groups. In a case series on 72 patients, Thal et al.⁴³ found no significant difference between biocomposite (3.7%) versus biodegradable (8.7%) suture anchors, but lower recurrence rate with a follow up of 2–7 years. In a systematic review, Brown et al.⁴⁴ reported recurrence rates of 10.1% and 7.8% for biodegradable and nonabsorbable anchors (biocomposite and metallic included) respectively. Milano et al.⁴⁵ compared the clinical outcome of arthroscopic treatment of shoulder instability between metallic and biodegradable anchors in 78 patients at 2-year follow-up, showing no significant difference but stating a total recurrence rate of 4.3% (2.8% and 5.9%, respectively).

Furthermore, no significant differences were observed amongst groups in terms of postoperative - at last follow-up - mean PROMs: ASES, Constant score, Rowe score, etc, mean external rotation at 90° of abduction, return to activities of daily living, return to sports, patient satisfaction, and self-reported sense of instability. All Oxford Shoulder Instability and Constant Scores surpassed the threshold level - 35 and 68, respectively - for successful arthroscopic Bankart treatment, as identified by Xu et al.⁴⁶ Willemot et al.³¹ in a case series of 20 patients at 1 year postoperatively treated with all-suture anchors demonstrated similar shoulder assessment scores – Constant: 89.3 – but no postoperative shoulder dislocation at all. Peters et al.¹⁴ in a retrospective cohort study of 155 cases compared the 4 different types of anchors in arthroscopic shoulder stabilization- 2 biodegradable, 1 biocomposite and 1 metal. They reported a 5-year recurrence rate of 5% in biocomposite anchor, 6% in metallic anchor and 19% and 37% in biodegradable anchors, showing a significant difference of the biodegradable to the nonabsorbable - biocomposite and metallic-anchors.

In the present study, the number of anchors used was mainly 3 or more: 3.71 (SD: 1.14) for group A and 4.31 (SD: 1.35) for group B, as using less than 3 anchors has been linked with increased rates of shoulder redislocation.^47,48

This study had several limitations. This was a retrospective chart review cohort study subjected to potential selection, detection and publication bias. The number of patients was small but not until the point that we could reach to statistically accurate results (the study was adequately powered). In addition, the two groups were found comparable in all baseline demographic, clinical pre- and intra-operative characteristics measured, while a single surgeon (the senior author) performed all the operations by using the same surgical technique.

While the follow-up was short, it has been proved that the vast majority of re-dislocations occur early after anterior instability surgery. Finally, other limitations were that we examined a mixed population of patients with history of one, two or recurrent dislocations, patients with either soft tissue capsule damage or bone lesions of various degrees and use of remplissage or not.

5. Conclusion

The short-term failure rate and clinical/functional outcomes of arthroscopic Bankart repair using all-suture anchors is similar to the use of biocompatible anchors. Based on these findings, it is suggested that all-suture anchors can be safely used in arthroscopic Bankart repair for the treatment of anterior shoulder instability.

Disclaimer

Declaration of interests: MM: Unpaid Consultant for Lima Inc. Udine, IT, outside the submitted work. No other conditions present potential conflicts of interest.

Funding statement

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

IRB approval

Not applicable-not required in our country for retrospective chart reviews, such as the current study.

Data statement

Not available or applicable.

CRediT authorship contribution statement

Ioannis Pantekidis: Data curation, Investigation, Formal analysis, Visualization, Writing – original draft, Writing – review & editing. Michael-Alexander Malahias: Conceptualization, Methodology, Methodology, Data curation, Writing – review & editing. Stefania Kokkineli: Writing – review & editing. Emmanouil Brilakis: Validation, Writing – review & editing. Emmanouil Antonogiannakis: Supervision, Resources, Project administration.

Declaration of competing interest

MM: Unpaid Consultant for Lima Inc. Udine, IT, outside the submitted work.

No other conditions present potential conflicts of interest.