UK National He

*Reprint req

Department of

Dewsbury Hosp

E-mail addre

J Shoulder Elbow Surg (2014) 23, 382-387

1058-2746/$ - s

http://dx.doi.org

www.elsevier.com/locate/ymse

Fewer anchors achieves successful arthroscopic shoulderstabilization surgery: 114 patients with 4 years offollow-up

Caroline Witney-Lagen, MRCS*, Namal Perera, MRCS, Sarah Rubin, MBBS,Balachandran Venkateswaran, FRCS

Department of Orthopaedics, Dewsbury and District Hospital, Halifax, W. Yorks, UK

Background: The shoulder is the most frequently dislocated joint, with an incidence of 10 to 20 per100,000 each year. The optimum number of anchors to use in arthroscopic stabilization is a topic ofgrowing interest; most surgeons use 3. Our stabilization technique is to commonly use only a single sutureanchor to purse-string the capsulolabral tissue up and toward the glenoid. This study aimed to ascertainwhether successful stabilization can be achieved with fewer than 3 anchors.Methods: Our study comprised 114 consecutive patients with anterior instability and a Bankart lesion un-dergoing arthroscopic stabilization with 4 years of follow-up. Outcome was measured by Oxford InstabilityScore (OIS) and recurrence of instability or dislocation. Patient demographics were 86.8% male, 13.2%female, mean age of 31 years, 76.3% Hill-Sachs lesions, 13.2% bony Bankart lesions, 13.2% glenoiddefects, and 9.6% SLAP lesions. The majority of patients, 71 patients (62.3%), received only 1 anchor;40 patients (35.1%) received 2 anchors, and 3 patients (2.6%) had 3 anchors.Results: The mean OIS was 44.3 preoperatively and 17.3 postoperatively (P < .0001). There was no dif-ference in OIS improvement between the patients who received a single anchor and those who received2 or 3 anchors (P > .05). Even with minor bony Bankart lesions and glenoid defects, a single suture anchorcan be sufficient. Our failure rate of 6.1% is comparable with that of other published series.Conclusion: Successful shoulder stabilization can be achieved with fewer than 3 anchors, and a single an-chor is usually sufficient.Level of evidence: Level III, Retrospective Cohort, Treatment Study.� 2014 Journal of Shoulder and Elbow Surgery Board of Trustees.

Keywords: Shoulder; dislocation; stabilization; instability; anchor; arthroscopic

The shoulder is the most frequently dislocated joint, with anincidence of 10 to 20 per 100,000 each year.9 Arthroscopicstabilization techniques have become increasingly popular.

alth Service ethics approval is not required for this work.

uests: Caroline Witney-Lagen, Orthopaedic SpR, MRCS,

Orthopaedics, C/o Orthopaedic Secretary Karen Hardy,

ital, Halifax Road, West Yorkshire, WF13 4HS, UK.

ss: carolinejws@gmail.com (C. Witney-Lagen).

ee front matter � 2014 Journal of Shoulder and Elbow Surgery

/10.1016/j.jse.2013.08.010

Data from the American Board of Orthopaedic Surgerydemonstrated that 71% of Bankart repairs were performedarthroscopically from 2003 to 2005, rising to 88% in 2006to 2008.14 Recent evidence also supports the efficacy ofarthroscopic stabilization. A meta-analysis,7 including morethan 3000 operations, reports that arthroscopic shoulder sta-bilization surgery with anchors or bioabsorbable tacks has arate of failure similar to that of open stabilization after 2 years.

Board of Trustees.

Fewer anchors in arthroscopic shoulder stabilization 383

The optimum number of anchors to use in arthroscopicstabilization is a topic of growing interest. The mostcommonly used number of anchors, as reported by theliterature, is 3. A paper by Boileau et al2 recommended that4 suture anchors be used because, in their experience, pa-tients with 3 or fewer anchors were at greater risk forrecurrence of instability. Similarly, van der Linde et al23

reported higher recurrence rates with fewer than 3 an-chors. Although there are no biomechanical studies evalu-ating the effectiveness of a single-anchor technique, thereare clinical reports of successful results being achievedwith fewer than 3 anchors. For example, Levy et al10 re-ported success with use of a single purse-string suture an-chor in a series of 36 patients treated arthroscopically. Inaddition, a study from the Copeland group by Massoudet al12 described a technique for open shoulder stabilizationwith use of a vertical-apical suture rather than any anchorsor tacks. This study aimed to ascertain whether successfulstabilization can be achieved with fewer than 3 anchors.

Figure 1 Diagrammatic representation of anchor positioning forpurse-string technique.

Materials and methods

The study comprised 114 consecutive patients. Inclusion criteriawere all patients with anterior instability and a Bankart lesionundergoing arthroscopic stabilization. Patients with bony glenoiddefects >20% were excluded from the study and treated with aLatarjet procedure. All patients were followed up for a minimumof 4 years postoperatively. The Oxford Instability Score (OIS) wascollected prospectively. All other data were collected retrospec-tively by use of patient case notes, x-ray analysis, and a Bluespier(Bluespier UK, Droitwich, Worcestershire, UK) computer data-base of operation records. Demographic data collected includedthe patient’s age and sex and the presence or absence of a bonyBankart lesion, glenoid defect, and SLAP (superior labral anterior-posterior) lesion. Outcome was measured by prospective evalua-tion of the OIS, which was collected both before and after surgeryin the outpatient clinic, and also by recurrence of instability ordislocation. Failure was defined either as a failure of the OIS toimprove by 10 points after surgery or by postoperative dislocation.

All operations were performed by the same surgeon (B.V.)between 2005 and 2008. Our stabilization technique is tocommonly use only a single suture anchor placed at 4 o’clock.One suture limb of the anchor is placed at 6 o’clock and the othersuture limb at 3 o’clock. Positioning of the anchor is illustrateddiagrammatically in Figure 1. The anchor is then used to snug thecapsulolabral tissue onto a well-prepared glenoid. The most vitalstep is the anteroinferior release of the displaced labrum andpreparation of the glenoid face, followed by the repair of thelabrum to create a bumper anteroinferiorly. We use hand-heldrasps and no power instruments to prepare the anterior face ofthe glenoid, which helps preserve bone stock.

This technique, which purse-strings the capsulolabral tissue upand toward the glenoid, is usually done adequately with only 1anchor. The purse-string allows reduction in the volume ofredundant capsule and eliminates drive-through. If this is achievedwith a single anchor, we do not use any more anchors. The pa-tient’s age and activity level do not influence our choice of therequired number of anchors. However, if the capsulolabral tissue

has not been adequately snugged down onto the glenoid, we do nothesitate to use 2 or 3 anchors as required. In our experience,associated lesions such as SLAP tears, posterior labral tears, bonyBankart lesions >1 cm, and engaging Hill-Sachs lesions with adepth >5 mm occupying more than 25% of the humeral headincrease the likelihood of requiring 2 or 3 anchors. We use thebioabsorbable Lupine anchor with Orthocord (DePuy Orthopae-dics Inc, Warsaw, IN, USA), which is ideal because the suturestrength allows adequate handling and tension without failure.Some of our early patients were treated with single-loaded an-chors, although our more recent preference is for double-loadedanchors as the second suture helps with labrum point fixation atthe 3-o’clock position.

Pre-anchor and post-anchor arthroscopic photographs areshown in Figure 2 to demonstrate the use of our technique. Tocomplete the purse-string, both limbs of the anchor are passed,thereby creating an increased surface area of labrum that isdirectly opposed to the glenoid. The resulting increased volume oflabrum is evident in the photograph and creates a physical bumperor tissue barrier to help prevent dislocation. Figure 3 is anarthroscopic photograph of a simple knot without purse-stringingfor comparison. In Figure 3, a single limb has been passed,providing point fixation only.

Postoperatively, all patients were placed into a polysling ininternal rotation. Pendulum exercises were commenced from day1. At 3 weeks, a physiotherapy appointment was made, and pas-sive range of movement exercises were commenced. Exercisesagainst resistance and overhead activities were introduced in thesixth postoperative week. Noncontact sports were allowed from 2months and contact sports from 6 months.

Statistical analysis was performed by the 2-tailed Student t test.Comparisons between the group of patients treated with 1 anchor

Figure 2 (A) Pre-anchor arthroscopic photograph. (B) Post-anchor and purse-stringing arthroscopic photograph.

Figure 3 Simple knot without purse-stringing.

71

40

3

Number of patients receiving 1, 2 or 3 anchors

1 anchor 2 anchors 3 anchors

Figure 4 Number of patients treated with a single anchor, 2anchors, and 3 anchors.

Table I Comparison of patient demographics in the groupstreated with a single anchor and with 2 or 3 anchors

384 C. Witney-Lagen et al.

and the patients treated with 2 or 3 anchors allowed unequalsample size and assumed equal variance. A value of P < .05 wasconsidered statistically significant.

1 anchor 2 or 3 anchors

No. of patients 71 43Mean age 31 years 10 months 30 years 1 monthMale 59/71 40/43Bankart lesion 71/71 43/43Bony Bankart lesion 5/71 10/43Glenoid defect 10/71 5/43Hill-Sachs lesion 50/71 37/43SLAP lesion 4/71 7/43

Results

Our series comprised 114 patients, of whom 99 (86.8%)were male and 15 (13.2%) female. The mean age was 31years (age range, 15-71 years). In total, 10 patients wereolder than 50 years. There were 87 (76.3%) Hill-Sachslesions, 15 (13.2%) bony Bankart lesions, 15 (13.2%) gle-noid defects, and 11 (9.6%) SLAP lesions. The number ofanchors used is shown in Figure 4. This demonstrates thatthe majority of patients, 71 patients (62.3%), received only1 anchor; 40 patients (35.1%) received 2 anchors, and 3patients (2.6%) had 3 anchors. The patient demographics inthe single-anchor group are compared with the de-mographics in the 2- or 3-anchor group in Table I.

All patients were followed up for a minimum of 4years. The length of follow-up ranged from 48 to 60months. Preoperative and postoperative OIS are shown inFigure 5. The mean OIS was 44.3 preoperatively and

17.3 postoperatively. This is a mean improvement of 27points. This difference between the preoperative andpostoperative scores is statistically significant withP < .0001.

Figure 6 shows the mean preoperative and postoperativeOIS for patients treated with 1 anchor and also for thosetreated with 2 or 3 anchors. For the patients treated with asingle anchor, the mean preoperative OIS was 44.9 points,and the mean postoperative OIS was 16.7 points, giving amean improvement of 28.2 points. The mean OIS for

Figure 5 Oxford Instability Scores (OIS) preoperatively and postoperatively.

Mean Oxford Instability Scores

0

5

10

15

20

25

30

35

40

45

50

3 ro 21

Number of anchors

OIS

Post-op

Pre-op

Figure 6 Mean preoperative and postoperative Oxford Insta-bility Scores (OIS) for patients treated with a single anchor andthose treated with 2 or 3 anchors.

Fewer anchors in arthroscopic shoulder stabilization 385

patients treated with 2 or 3 anchors was 43.3 points pre-operatively and 18.3 points postoperatively, giving a meanimprovement of 25.0 points. This difference between thegroups was not found to be statistically significant withP ¼ .055.

Our overall failure rate was 6.1%. Of 71 patients treatedwith a single anchor, failure occurred in 2; of 40 patientstreated with 2 anchors, failure occurred in 4; and of 3 pa-tients treated with 3 anchors, failure occurred in 1. Recur-rent instability developed in 6 patients (5.3%). All of thesepatients were male, aged between 21 and 29 years (meanage, 25 years) and keen sportsmen. The 2 failures in thesingle-anchor group occurred in a 25-year-old sportsmanwith a small Hill-Sachs lesion who developed recurrentinstability and a 54-year-old man who did not sufferinstability but was unhappy with his surgery, continuing tosuffer with pain in activities of daily living. The 4 failuresin the 2-anchor group included two 29-year-old sportsmen;a 21-year old sportsman who was initially well, but 1 yearafter surgery he suffered a dislocation, falling off amotorbike; and a 26-year-old footballer who sustained adislocation 2 years after surgery in a football tackle. Thefailure in the 3-anchor group occurred in a 25-year-old manwith a large Hill-Sachs lesion and a bony Bankart lesion

who, against our advice, played rugby in his sixth post-operative week and sustained a dislocation followed byrecurrent instability.

Discussion

In our series, we have used only 1 anchor in the majority ofpatients. Although most surgeons commonly use 3 anchors,we have used 3 anchors only in 3 (2.6%) of our patients.Despite our smaller number of anchors, our failure rate of6.1% compares favorably with that of other published se-ries.1-6,8,10,11,13,15-23 Table II shows how our resultscompare with the literature.

Strengths of our study include a relatively large patientsample size, a relatively long minimum follow-up, and astandardized operation technique. Our follow-up timeranged from 48 to 60 months, which is longer than in mostreported studies. However, we do know that rates ofrecurrent instability increase every year. The 3 highestfailure rates shown in Table II correspond to the 3 studieswith the longest follow-up period.3,15,23 Obviously, if wefollowed up our patients for a longer time, such as the 8- to10-year follow-up of van der Linde et al,23 the10-yearfollow-up of Castagna et al,3 or the 13.5-year follow-upof Privitera et al,15 we would expect more failures duringthis extended period. Therefore, some caution needs to beemployed in comparing our results with standard practiceas the long-term outcome is still uncertain. Greater patientnumbers would also provide more accurate data. It ispossible that the greater patient numbers seen in the studiesby Blomquist et al,1 Imhoff et al,8 and Tan et al19 couldaccount for their failure rates being greater than ours.

Another clear limitation of our study is that our patientpopulation treated with a single anchor was not matchedwith the patient population treated with 2 or 3 anchors(Table I). Statistically there was no significant difference inthe OIS improvement between patients treated with a single

Table II Comparison of our results with those in theliterature

Author No. ofpatients

Follow-up(months)

Failures(%)

Our study 114 48-60 6.1Blomquist1 213 12 10.0Privitera15 20 162 35.0van der Linde23 68 96-120 35.0Nepra�s13 129 12-60 4.6Chiang4 45 77 6.7Castagna3 31 120 22.6Imhoff8 190 Average of 37 10.5Cooke5 20 Average of 26 15.0Cooke6 15 17-31 6.7Sedeek17 40 Average of 30 7.5Tischer21 147 Average of 36 6.1Thal20 72 24-84 6.9Marquardt11 54 Average of 44 7.5Boileau2 91 Average of 36 15.4Tan19 124 18-60 5.6Reichl16 65 3-20 6.6Valis22 64 3-36 4.7Speck18 30 12-42 13.3

386 C. Witney-Lagen et al.

anchor and those treated with 2 or 3 anchors. However, wedid find that as the number of anchors increased from 1 to 2to 3, the failure rates also increased from 3% to 10% to33%, respectively. We believe this reflects the fact thatmore anchors were used for more severe disease and thatmore severe disease is more likely to result in failure. Inparticular, our patients with SLAP lesions and bonyBankart lesions were more likely to be treated with 2 or 3anchors. However, we did have 5 patients with minor bonyBankart lesions and another 10 patients with minor glenoiddefects treated with only a single anchor. None of thesepatients failed. In addition, when we did use more anchorsfor our more severe lesions, we were still unlikely torequire the standard surgeon’s preference of 3 anchors.

Conclusion

Although most surgeons choose to use 3 anchors as theirstandard practice, we found that fewer anchors canachieve successful shoulder stabilization. A single an-chor was sufficient for the majority of our patients. Wefound no statistically significant difference in outcomefor patients treated with a single anchor compared withthose treated with 2 or 3 anchors. In fact, our experienceshows that even in patients with minor bony Bankartlesions and minor glenoid defects, a single anchor can besufficient. We believe it is the superomedial shift of thetorn capsulolabral tissue that is important, rather than thenumber of anchors. Therefore, successful stabilizationcan be achieved with fewer than the standard practice of

3 anchors, and in fact a single anchor is sufficient formost patients.

Disclaimer

The authors, their immediate families, and any researchfoundation with which they are affiliated have notreceived any financial payments or other benefits fromany commercial entity related to the subject of thisarticle.

References

1. Blomquist J, Solheim E, Liavaag S, Schroder CP, Espehaug B,

Havelin LI. Shoulder instability surgery in Norway: the first report

from a multicenter register, with 1-year follow-up. Acta Orthop 2012;

83:165-70. http://dx.doi.org/10.3109/17453674.2011.641102

2. Boileau P, Villalba M, H�ery JY, Balg F, Ahrens P, Neyton L. Risk

factors for recurrence of shoulder instability after arthroscopic Bankart

repair. J Bone Joint Surg Am 2006;88:1755-63. http://dx.doi.org/10.

2106/JBJS.E.00817

3. Castagna A, Markopoulos N, Conti M, Delle Rose G, Papadakou E,

Garofalo R. Arthroscopic Bankart suture-anchor repair: radiological

and clinical outcome at minimum 10 years of follow-up. Am J Sports

Med 2010;38:2012-6. http://dx.doi.org/10.1177/0363546510372614

4. Chiang ER, Wang JP, Wang ST, Ma HL, Liu CL, Chen TH. Arthro-

scopic posteroinferior capsular plication and rotator interval closure

after Bankart repair in patients with traumatic anterior glenohumeral

instabilityda minimum follow-up of 5 years. Injury 2010;41:1075-8.

http://dx.doi.org/10.1016/j.injury.2010.05.028

5. Cooke S, Ennis O, Majeed H, Rahmatalla A, Kathuria V, Wade R.

Clinical results and motion analysis following arthroscopic anterior

stabilization of the shoulder using bioknotless anchors. Int J Shoulder

Surg 2010;4:36-40. http://dx.doi.org/10.4103/0973-6042.70821

6. Cooke SJ, Starks I, Kathuria V. The results of arthroscopic anterior

stabilisation of the shoulder using the bioknotless anchor system.

Sports Med Arthrosc Rehabil Ther Technol 2009;1:2. http://dx.doi.

org/10.1186/1758-2555-1-2

7. Hobby J, Griffin D, Dunbar M, Boileau P. Is arthroscopic surgery for

stabilisation of chronic shoulder instability as effective as open sur-

gery? A systematic review and meta-analysis of 62 studies including

3044 arthroscopic operations. J Bone Joint Surg Br 2007;89:1188-96.

http://dx.doi.org/10.1302/0301-620X.89B9.18467

8. Imhoff AB, Ansah P, Tischer T, Reiter C, Bartl C, Hench M, et al.

Arthroscopic repair of anterior-inferior glenohumeral instability using

a portal at the 5:30-o’clock position: analysis of the effects of age,

fixation method, and concomitant shoulder injury on surgical out-

comes. Am J Sports Med 2010;38:1795-803. http://dx.doi.org/10.

1177/0363546510370199

9. Krøner J, Lind T, Jensen J. The epidemiology of shoulder dislocations.

Arch Orthop Trauma Surg 1989;108:288-90.

10. Levy O, Matthews T, Even T. The ‘‘purse-string’’ technique: an

arthroscopic technique for stabilization of anteroinferior instability of

the shoulder with early and medium-term results. Arthroscopy 2007;

23:57-64. http://dx.doi.org/10.1016/j.arthro.2006.10.006

11. Marquardt B, Witt KA, Liem D, Steinbeck J, P€otzl W. Arthroscopic

Bankart repair in traumatic anterior shoulder instability using a suture

anchor technique. Arthroscopy 2006;22:931-6. http://dx.doi.org/10.

1016/j.arthro.2006.04.105

Fewer anchors in arthroscopic shoulder stabilization 387

12. Massoud SN, Levy O, Copeland SA. The vertical-apical suture

Bankart lesion repair for anteroinferior glenohumeral instability. J

Shoulder Elbow Surg 2002;11:481-5. http://dx.doi.org/10.1067/mse.

2002.126207

13. Nepra�s P, Zeman P, Mat�ejka J, Koudela K Jr, Koudela K Sr. Arthro-

scopic stabilisation of post-traumatic ventral instability of the shoulder

by use of bioknotless anchors [in Czech]. Acta Chir Orthop Traumatol

Cech 2011;78:56-60.

14. Owens BD, Harrast JJ, Hurwitz SR, Thompson TL, Wolf JM. Surgical

trends in Bankart repair: an analysis of data from the American Board

of Orthopaedic Surgery Certification Examination. Am J Sports Med

2011;39:1865-9. http://dx.doi.org/10.1177/0363546511406869

15. Privitera DM, Bisson LJ, Marzo JM. Minimum 10-year follow-up of

arthroscopic intra-articular Bankart repair using bioabsorbable tacks.

Am J Sports Med 2012;40:100-7. http://dx.doi.org/10.1177/

0363546511425891

16. Reichl M, Koudela K. Post-traumatic anterior shoulder insta-

bilitydarthroscopic stabilization method using bone anchors [in

Czech]. Acta Chir Orthop Traumatol Cech 2004;71:37-44.

17. Sedeek SM, Tey IK, Tan AH. Arthroscopic Bankart repair for trau-

matic anterior shoulder instability with the use of suture anchors.

Singapore Med J 2008;49:676-81.

18. Speck M, Hertel R. Arthroscopic capsulo-labral repair and refixation

with Mitek anchor in anterior shoulder instability [in German]. Z

Orthop Ihre Grenzgeb 1997;135:348-53.

19. Tan CK, Guisasola I, Machani B, Kemp G, Sinopidis C, Brownson P,

et al. Arthroscopic stabilization of the shoulder: a prospective ran-

domized study of absorbable versus nonabsorbable suture anchors.

Arthroscopy 2006;22:716-20. http://dx.doi.org/10.1016/j.arthro.2006.

03.017

20. Thal R, Nofziger M, Bridges M, Kim JJ. Arthroscopic Bankart repair

using knotless or bioknotless suture anchors: 2- to 7-year results.

Arthroscopy 2007;23:367-75. http://dx.doi.org/10.1016/j.arthro.2006.

11.024

21. Tischer T, Vogt S, Imhoff AB. Arthroscopic stabilization of the

shoulder with suture anchors with special reference to the deep

anterior-inferior portal (5.30 o’clock). Oper Orthop Traumatol 2007;

19:133-54. http://dx.doi.org/10.1007/s00064-007-1199-1

22. Valis P, N�ydrle M. Arthroscopic stabilization of the shoulder using

anchors [in Czech]. Acta Chir Orthop Traumatol Cech 2003;70:233-6.

23. van der Linde JA, van Kampen DA, Terwee CB, Dijksman LM,

Kleinjan G, Willems WJ. Long-term results after arthroscopic shoulder

stabilization using suture anchors: an 8- to 10-year follow-up. Am J Sports

Med 2011;39:2396-403. http://dx.doi.org/10.1177/0363546511415657