From theRadiology (TSurgery andLondon, En(A.Y.), TokySports MedicTennessee, U

The authoand publicat

Received JAddress c

Orthopedic Sku, Tokyo 10

� 2015 b0749-8063http://dx.d

Combined Arthroscopic Bankart Repair and CoracoidProcess Transfer to Anterior Glenoid for ShoulderDislocation in Rugby Players: Evaluation Based

on Ability to Perform Sport-SpecificMovements Effectively

Atsushi Tasaki, M.D., Ph.D., Wataru Morita, M.D., Akira Yamakawa, M.D.,Taiki Nozaki, M.D., Eishi Kuroda, M.D., Yoshimitsu Hoshikawa, M.D.,

and Barry B. Phillips, M.D.

Purpose: To evaluate the outcomes of a combination of an arthroscopic Bankart repair and an open Bristow procedure inrelation to the subjective quality of performance in movements that are typical in rugby. Methods: Forty shoulders in 38players who underwent surgery for traumatic anterior instability of the shoulder were reviewed. In all cases, arthroscopicBankart repair was followed by a Bristow procedure, with preservation of the repaired capsular ligaments, during the sameoperation. The mean age at the time of surgery was 21 years. Patients were asked to describe common rugby maneuvers(tackle, hand-off, jackal, and saving) preoperatively and postoperatively as “no problem,” “insufficient,” or “impossible.”Results: There were no recurrent dislocations at a mean follow-up of 30.5 months. The mean Rowe score improvedsignificantly from 65.0 (range, 55 to 75) to 97.5 (range, 95 to 100) (P < .001) after surgery. Preoperatively, regarding thetackling motion, none of the patients reported having no problem, whereas the ability was described as insufficient for 23shoulders and impossible for 17 shoulders. Postoperatively, no problemwith tackling was reported for 36 shoulders, whereasinsufficiencywas reported for 4. The results for the hand-off, jackal, and savingmaneuvers were similar (P< .001). No patientrated any of the motions as impossible postoperatively.Conclusions: This combined surgical procedure clearly is effective inpreventing recurrent dislocation in rugby players; however, some players complained of insufficiency in the quality of theirplay when they were tackling or performing other rugby-specific movements. Level of Evidence: Level IV, case series.

he treatment strategy for shoulder instability in

Tcollision sport athletes remains controversial.Currently, arthroscopic surgery is the most commonoperative therapy for traumatic anterior dislocation of

Departments of Orthopedic Surgery (A.T., E.K., Y.H.) and.N.), St. Luke’s International Hospital, Tokyo, Japan; Division ofInterventional Science, University College London (W.M.),

gland; Division of Clinical Biotechnology, University of Tokyoo, Japan; and Department of Orthopaedic Surgery, Division ofine, University of TennesseeeCampbell Clinic (B.B.P.), Memphis,.S.A.rs report that they have no conflicts of interest in the authorshipion of this article.une 20, 2014; accepted March 12, 2015.orrespondence to Atsushi Tasaki, M.D., Ph.D., Department ofurgery, St. Luke’s International Hospital, 9-1 Akashicho, Chuo-4-8560, Japan. E-mail: tatsu@luke.ac.jpy the Arthroscopy Association of North America/14511/$36.00oi.org/10.1016/j.arthro.2015.03.013

Arthroscopy: The Journal of Arthroscopic and Related

the shoulder. Arthroscopic Bankart repair enables anaccurate intra-articular assessment, as well as simulta-neous treatment of disorders within the joint, such asretensioning of glenohumeral ligaments and repair ofSLAP lesions and capsular tears. However, a procedurethat repairs only the soft tissues may be insufficient forcollision sport athletes, who frequently present withsevere glenoid bony defects and Hill-Sachs lesions inaddition to the soft-tissue damage.1,2

Coracoid transfer, which is also known as the Bristowprocedure and the Latarjet procedure, has been re-ported to produce positive results in collision sportathletes with traumatic anterior instability of theshoulder1,3; however, this procedure does not repairthe soft tissues anatomically and, when used alone, hasa risk of persistent apprehension if the transferredcoracoid is malpositioned or the bone graft does notunite.4 We have been using a combination of anarthroscopic Bankart repair to anatomically repair the

Surgery, Vol -, No - (Month), 2015: pp 1-9 1

Fig 1. Our procedure in a rightshoulder. (A) Drilling perpendicu-larly from the cancellous part of thecoracoid process (CP) that wasdecorticated convexly for trans-fixion to the anterior aspect of theglenoid. (B) Exposure of the ante-rior surface of the glenoid(asterisk). The capsular ligament(CL) was preserved. (SC,subscapularis.)

2 A. TASAKI ET AL.

intra-articular lesions and glenohumeral ligaments andan open Bristow procedure to provide robust stabilityby extra-articular augmentation for competitive rugbyplayers with traumatic anterior dislocations of theshoulder.To evaluate the effectiveness of this combination of

procedures, we surveyed rugby players before and aftersurgery to determine whether apprehension or limita-tions that prevented the athletes from competing fullyremained, even if there were no recurrent dislocations.We believe that it is important to evaluate the post-operative outcome not only by the rate of recurrentdislocation but also by the quality of play from thepatient’s point of view. This study aimed to evaluate theoutcome of a combination of an arthroscopic Bankartrepair and an open Bristow procedure in relation to thesubjective quality of performance in movements thatare typical in rugby. We hypothesized that the surgicalprocedure would not only prevent recurrent dislocationbut also allow a full return to athletic competitionwithout insufficiency.

MethodsApproval was obtained from our institutional review

board (St. Luke’s International Hospital Research EthicsCommittee, No. 13-R369) before enrollment of patientsin this study. Patients who had undergone an arthro-scopic Bankart repair and open Bristow procedure fortraumatic anterior instability of the shoulder betweenApril 2008 and April 2011 were identified from thesenior author’s (A.T.) database. The inclusion criteriawere competitive rugby players who were participatingat or above the high school level and who continued toactively participate in the sport for more than 2 years.All operations and evaluations were performed at asingle institution by a single surgeon (A.T.).

Operative TechniqueSurgery was performed with the patient in the beach-

chair position using an arm support (Spider LimbPositioner; Tenet Medical Engineering, Calgary,Alberta, Canada). A deltopectoral approach was usedwith an incision of 7 to 10 cm according to the patient’ssize. After the coracoacromial ligament and the tendonof the pectoralis minor were severed from the coracoidprocess, the coracoid process was dissected whereas theconjoined tendon was preserved. The cut end of thecoracoid process that was dissected was decorticatedand shaped so that it was approximately 15 mm longand the cancellous bone was convex (Fig 1A). A4.0-mm cannulated cancellous screw (Meira, Nagoya,Japan) was used to fix the coracoid process to theanterior surface of the glenoid. The coracoid fragmentwas prepared for fixation by measuring the length of ascrew needed to transfix the coracoid to the glenoid andthen tapping and drilling perpendicular to the cancel-lous surface.The subscapularis was separated in the direction of its

fibers, with preservation of the enthesis. The capsularligament of the glenohumeral joint was identified andpreserved. Through the subscapularis interval, a glenoidretractor was placed at the glenoid neck so that theanterior surface of the glenoid was exposed at the4-o’clock position in a right shoulder (slightly inferior tothe center) (Fig 1B).Next, arthroscopic Bankart repair was performed. A

hydraulic pump was used in all procedures, and 3portals were used (anterior, posterior, and ante-rosuperior). The anteroinferior capsular ligament wasmobilized by sufficient detachment and then pulled upand retensioned. Leakage of fluid from the detachedcapsular ligament and separated subscapularis muscledid not interfere with this procedure. The reduced

Fig 2. Bristow method in a rightshoulder. (A) A washer was usedfor fixation of the coracoid bonegraft. (B) True anteroposteriorradiograph.

Fig 3. Arthroscopic view of the repaired Bankart lesion, afteran open Bristow procedure, in a right shoulder. The capsularligaments are conserved after the Bristow procedure.

BANKART REPAIR IN RUGBY PLAYERS 3

ligament was sutured simply using at least 4 sutureanchors that were placed evenly from the 5:30 clock-face position to the 2-o’clock position. Absorbable su-ture anchors (Panalok loops; DePuy Mitek, Raynham,MA) were used for the first 23 procedures and Bio-raptor (Smith & Nephew, Andover, MA) for the latter17 procedures. In all 15 shoulders with a bony Bankartlesion, the middle and inferior glenohumeral ligamentswere attached to the fractured portion of the glenoidand reduced simultaneously with the repair of thecapsular ligament.5 Other noted intra-articular lesionswere treated as needed.Fixation of the prepared coracoid fragment was car-

ried out according to the Bristow procedure.6 With caretaken to preserve the repaired capsular ligaments andto avoid placing the graft lateral to the glenoid articularsurface, a guide pin was inserted from the anteriorsurface of the glenoid to the posterior surface, parallelto the glenoid surface at the 4-o’clock level in a rightshoulder. The anterior edge of the glenoid wasconfirmed over the capsule digitally, and the cortex wasthen decorticated with a cannulated 8-mm bonereamer along the guide pin to obtain sufficient bleedingfrom the bone surface. The fragment was inserted overthe guide pin and fixed onto the anterior surface of theglenoid with a screw and washer that reached thecontralateral cortex of the glenoid (Fig 2). Special carewas taken so that the graft was close to the glenoidarticular surface but did not protrude. The fixed posi-tion was confirmed with plain radiographs (ante-roposterior and axial) during surgery. Finally, the jointspace was inspected arthroscopically, especially therepaired Bankart lesion. The split subscapularis musclewas not sutured. The incisions were closed (Fig 3).A shoulder brace with the shoulder in a neutral po-

sition was used for 3 weeks postoperatively, followedby range-of-motion (ROM) exercises. Plain radiographs

were obtained every 4 weeks until bone union wasachieved. When union was confirmed, patients wereallowed to start running at 10 weeks, to practice passingthe ball and begin weight training at 12 weeks, and toparticipate in practices including physical contact at 4months postoperatively. Computed tomography (CT)was used in 12 cases with unclear bone union or bonyBankart lesions when further evaluation for boneunion and location of the transferred coracoid wasthought to be appropriate. Postoperative dislocations orsubluxations, Rowe scores, ROM, perioperative com-plications, and findings of evaluation of the fixedcoracoid fragment were reviewed.7 Positional accuracyof the grafted bone was evaluated on plain radiographs.The ideal position of the graft is acknowledged to bebetween the 3- and 4-o’clock positions in a rightshoulder on anteroposterior glenoid radiographs.

Fig 4. Typical movements in rugby.The arrows indicate the playersperforming each movement. (A)Tackle: hitting the opponent at theshoulder while extending the arm.(B) Saving: reaching out to obtainthe rolling ball. (C) Hand-off:sticking the hand and arm out toward off the opponent, who isattempting to tackle. (D) Jackal(performed by defensive player):taking the ball from the tackledplayer.

Table 1. Patient Demographic Characteristics

Data

Players (n ¼ 38)Sex

Male 37, 2 bilateral (97)Female 1 (3)

League levelNational top 3 (8)College 34 (89)High school 1 (3)

PositionForward 20 (53)Back 18 (47)

Shoulders (n ¼ 40)Revision surgery 8 (20)Glenoid bony defect

Negative 22 (55)Positive 18 (45)

Median age, yr (range) 21 (17-25)

NOTE. Data are presented as n (percent) unless otherwiseindicated.

4 A. TASAKI ET AL.

Similarly, the ideal placement of the graft relative to theglenoid is within 5 mm medial to the articular surfaceon an axial radiograph.To assess the movements typical in rugby football, we

focused on 4 movements: tackling, saving, hand-off,and jackal (Fig 4). The patients were then categorizedinto 3 groups depending on whether they had “noproblem” performing these maneuvers or the maneu-vers were “insufficient” or “impossible” according to asubjective survey conducted before and after surgery.

Statistical AnalysisThe results were statistically analyzed using the Wil-

coxon signed rank test for the Rowe score and paired ttest for each ROM of the shoulder. An extendedMcNemar test was used for evaluation of rugby-specificmaneuvers, with significance set at P < .05. Statasoftware (version 12.0; StataCorp, College Station, TX)was used for all statistical analyses.

ResultsThe records of 42 shoulders in 40 consecutive top-

level competitive rugby players were reviewed. Nosame-level rugby players were treated surgically byother procedures during this period. All patients hadmultiple dislocation or subluxation episodes. Two

patients were excluded because they retired fromplaying within 2 years after surgery. We included 1female and 37 male players in this study, and 2 of themale players underwent treatment bilaterally, resultingin 40 shoulders in 38 patients (Table 1). Of the players,3 competed in national top leagues, 34 in college varsity

Table 2. Operative Findings, Complications, and Location ofTransferred Coracoid (n ¼ 40 Shoulders)

Parameter n (%)

Concomitant repairSLAP 4 (10)HAGL 3 (7.5)Capsular tear 1 (2.5)Bony Bankart fragment 15 (37.5)

ComplicationInfection 0 (0)Detachment of coracoid 2 (5)Nonunion 3 (7.5)

Location of coracoidAnteroposterior view3-o’clock position 3 (7.5)4-o’clock position 35 (87.5)5-o’clock position 2 (5)

Axial view0-5 mm medial 35 (87.5)5-10 mm medial 5 (12.5)

HAGL, humeral avulsion of glenohumeral ligament.

BANKART REPAIR IN RUGBY PLAYERS 5

leagues, and 1 at the high school level. Twenty playerswere forwards, and 18 were backs.The median age at surgery was 21 years (range, 17 to

25 years), and all patients were top-level competitiverugby players with a Tegner activity level of 9 to 10.8

Eight patients had undergone arthroscopic Bankartrepair but had recurrent instability, 6 of whom under-went surgery at a different hospital. Preoperative eval-uation with 3-dimensional CT showed glenoid rimbone deficits in 18 shoulders, with mean bone loss of12.2% (range, 5% to 22%). The size of the defect wascalculated as the percentage of the glenoid fossa on theen face CT view, with the equation B/A � 100%, whereA is the diameter of the outer fitting circle based on theinferior part of the glenoid contour from the 3- to9-o’clock position and B is the length of the bonydefect.9 In 15 shoulders there was a bony Bankartlesion. In all 7 shoulders with bone deficits exceeding15%, the avulsed fragments were left in place.The mean surgical time including perioperative ra-

diographs was 3 hours 8 minutes (range, 2 hours 30minutes to 4 hours 45 minutes). No patient had sub-stantial hemorrhage. The median number of sutureanchors used was 4.5 (range, 4 to 6). SLAP repair was

Table 3. Comparison of Rowe Scores and Range of Motion Preo

Parameter

Median (Range) or Mean � S

Preoperative Postope

Rowe score 65 (55-75) 97.5 (95Range of motion, �

Forward flexion 175 � 1 170ER at 0� of abduction 62 � 4 56ER at 90� of abduction 91 � 2 90

CI, confidence interval; ER, external rotation.

performed in 4 patients, capsular repair in 3, and repairof a humeral avulsion of the glenohumeral ligament in1 (Table 2). All 15 bony Bankart lesions were repairedwith suture anchors. Final arthroscopic evaluationbefore closure showed no cases in which the screw forthe Bristow procedure had damaged the initial Bankartrepair necessitating further repair.The mean postoperative time until return to play was

5.9 months (range, 4 to 8 months). All patientsreturned to their preoperative athletic levels. Patientswere followed up only during the time they continuedto play. During the mean postoperative follow-upperiod of 30.5 months (range, 24 to 42 months),there were no shoulder dislocations or subluxationsand no patients had an apprehension sign in the end-range positions of abduction and external rotation.The median Rowe score improved significantly from 65(range, 55 to 75) to 97.5 (range, 95 to 100) (P < .001)(Table 3). The postoperative loss of ROM showed asignificant difference between preoperative and post-operative values: 3� in forward flexion and 6� inexternal rotation at 0� of abduction (P < .001)(Table 3).One patient was treated with antibiotics for a super-

ficial surgical-site infection. In 2 patients transientaxillary nerve palsy was noted after surgery, but thisresolved within 3 months.In 2 patients the transplanted bone segment

completely detached from the glenoid and requiredsurgical refixation. These were the third and fifth caseswithin our experience, and displacement was noted onfollow-up radiographs at 1 month and 3 months post-operatively (Fig 5). The coracoid fragment had cut outfrom the screw and had not been resorbed. Refixationwas performed immediately after the displacement wasnoted, and a 4.0-mm cancellous screw with a spikewasher was used instead of the previously used 3.5-mmcortical screw alone; bone union was confirmed byplain radiographs after 2 months.Nonunion of the grafted bone occurred in 3 shoul-

ders. Plain radiographs were used to confirm callusformation at the anteroinferior aspect of the shoulder.The postoperative return to play was delayed in these 3patients because the rehabilitation protocol afterarthroscopic Bankart repair was adopted given the lack

peratively and Postoperatively (n ¼ 40 Shoulders)

D

SE P Value 95% CIrative

-100) 1.2 < .001 �31.9 to �27

� 3 0.4 < .001 2.1 to 3.9� 4 1.0 < .001 4.1 to 8.1� 4 0.8 > .99 �1.2 to 1.6

Fig 5. Radiographs of graftdisplacement (arrows) with frac-ture 1 month after surgery in aright shoulder: (A) anteroposteriorand (B) axial views.

6 A. TASAKI ET AL.

of complete support by the graft. All 3 returned to playat about 6 months postoperatively. In 2 shoulders thescrew had dislodged (Fig 6) and was removed duringthe off-season.In terms of positional accuracy of the coracoid graft,

anteroposterior glenoid radiographs showed that 35 of40 grafts were placed at the 4-o’clock position in theright shoulder. The fragment was placed at the5-o’clock position in 2 shoulders and at the 3-o’clockposition in 3. On axial radiographs, 35 grafts werewithin 5 mm medial to the articular surface, whereas in5 shoulders the graft was placed more than 5 mmmedial to the glenoid surface (Fig 7); bone union wasobtained in all 5 of the latter shoulders. No graftextruded laterally compared with the glenoid surface.Improvement in tackling was found postoperatively

(P < .001) (Table 4). Before surgery, no player wasable to tackle fully, whereas the ability to tackle wasrated as insufficient for 23 shoulders and as impossiblefor 17. Postoperatively, the ability to tackle fully wasreported for 36 shoulders, and tackling was rated asinsufficient for 4 shoulders. Regarding the hand-offmaneuver preoperatively, the ability was rated asinsufficient for 20 shoulders and as impossible for 20

Fig 6. Nonunion of coracoid boneat 6 months after surgery in a rightshoulder: (A) anteroposterior and(B) axial views. The screw dis-lodged, although callus formation ispresent. The patient reported nocomplaints of instability andreturned to play.

shoulders. After surgery, 34 shoulders had no problem,and the ability was rated as insufficient for 6 shoulders.The results for the saving and jackal maneuvers weresimilar (P < .001).

DiscussionAll of the competitive rugby players who had a

combination arthroscopic Bankart repaireopen Bristowprocedure were able to return to competition, and therewere no recurrent anterior dislocations during theminimum 2-year follow-up period. In rugby playerswith traumatic shoulder instability, discussions ofmovements that affect their ability to play have focusedprimarily on tackling; however, our study showed thatplayers complain of insufficiency when they performthe hand-off, jackal, and saving maneuvers that arecommon in rugby.In our patients the hand-off movement, which is also

known as the stiff-arm maneuver, evoked as muchapprehension as tackling or even more. During thehand-off movement, when the affected arm isextended, there is a potential for the arm to be pulled orrotated by an opponent. Similarly, in the jackal andsaving maneuvers, the extended arm is prone to be

Fig 7. Measurement of medialplacement on plain axial radio-graphs in a right shoulder. Thedotted lines indicate the level of theglenoid surface, and the solid lineindicates the medial margin of thecoracoid fragment. (A) “Flashed”position. (B) Measurement of thedistance between the 2 lines(arrow).

BANKART REPAIR IN RUGBY PLAYERS 7

forced to elevate or externally rotate. Hence the insuf-ficiency may be psychological. Limitations in not justtackling but in multiple motions lead to a decrease inperformance, indicating that the presence of a post-operative dislocation and an inability to tackle are notthe only appropriate criteria for evaluating outcomes inrugby players.Shoulders with traumatic instability usually have

bony deficits or capsular damage that may compromisethe results of solitary arthroscopic Bankart repair incollision sport athletes2,10; however, arthroscopicanatomic repair of intra-articular lesions has benefits insupporting instability and regaining shoulder function,including proprioception and ROM.11,12 Moreover,concomitant intra-articular lesions such as bonyBankart lesions, SLAP lesions, and capsular tears thatcould be treated arthroscopically would be more diffi-cult to diagnose and treat by an open approach. Rotatorinterval closure or remplissage could be added during

Table 4. Evaluation of Rugby-Specific Motions: Comparisonof Subjective Ability Between Preoperative and PostoperativeStates

Preoperative Postoperative P Value

Tackle < .001No problem 0 36Insufficient 23 4Impossible 17 0

Hand-off < .001No problem 0 34Insufficient 20 6Impossible 20 0

Saving < .001No problem 0 35Insufficient 27 5Impossible 13 0

Jackal < .001No problem 0 36Insufficient 30 4Impossible 10 0

NOTE. Data are reported as number of shoulders unless otherwiseindicated.

arthroscopic surgery.13 We chose a bony repair or bonyaugmentation.The “sling” effect, by tensioning the subscapularis,

plays a role in supporting the shoulder joint in both theBristow procedure and Latarjet procedure.14,15 Biome-chanical study has identified the sling effect at both theend-range and midrange arm positions.15 However, thisprocedure alone does not repair the capsular ligamentthat was incised when the coracoid graft was fixed tothe glenoid. Therefore malpositioning of the graft maylead to insufficient support at certain positions whenthe shoulder is abducted.16,17 With an open approach,diagnosis and treatment of intra-articular lesions cannotbe performed accurately, and although the possibility ofdislocation recurrence may be low, there is a chance ofresidual symptoms.Postoperative ROM decreased significantly compared

with preoperative measurements in flexion (3�) andexternal rotation at 0� of abduction (6�). Various factorsare known to affect postoperative ROM, such as thepatient’s individual laxity, the rehabilitation protocol,and the surgical methods, including the Bristow pro-cedure. However, these declines in ROM were accept-able regarding returning to playing rugby and did notdiffer greatly from those after standard arthroscopicBankart repair.18

Arthroscopic coracoid process transfer to the anteriorglenoid has been described recently.19,20 This procedureis extremely advanced and requires special instrumen-tation. The transection of the coracoid process must becompleted before the Bankart repair to complete theoperation arthroscopically. In cases with bony Bankartlesions, the site where the bony fragment has to bereduced coincides with the projected place of coracoidtransfer. This may interfere with the surgical procedure,and hence the fragment may have to be resected duringthis all-arthroscopic procedure. Preservation of thefragment is desirable because the bony piece could fillthe deficit when reduced.5 We reduced and repairedthe fractured bone arthroscopically and then fixed the

Fig 8. Computed tomography images in a patient with a bonyBankart fragment in the left shoulder. (A) Sagittal view before

8 A. TASAKI ET AL.

coracoid graft onto the repaired and compressed gle-noid bony fragment (Fig 8). In an all-arthroscopicprocedure, the rotator interval and part of the jointcapsule have to be dissected to obtain sufficient expo-sure for coracoid osteotomy.19,20 Our method has anadvantage of avoiding further damage to the sur-rounding tissues that should be preserved.Preservation of the arthroscopically repaired capsular

ligament during the open Bristow method may limitthe view of the anterior part of the glenoid neck, andthis may account for the 2 grafts that displaced and the3 nonunions with callus formation that occurred in ourstudy early after starting this procedure. A fine nerveretractor was useful to separate the subscapularis in atransverse direction and to obtain a clear view. Inaddition, after the 3 nonunions, more attention waspaid to decortication of both the coracoid graft and theanterior aspect of the glenoid. No nonunions occurredafter this was incorporated into the procedure.Several studies have reported relatively high fre-

quencies of malpositioning of the grafted coracoid withthe Bristow procedure.21-23 We were extremelycautious when placing the graft because we preservedthe capsule and oriented the location by palpating thejoint space and the glenoid digitally. The postoperativeimages did not show any graft projecting into the jointspace. With sufficient repair of the capsular ligaments,no clinical problems, including severe restriction inabduction, were noted even when the graft was moremedial than intended. Therefore the key points in thesurgical procedure to prevent osteoarthritic changes areto avoid placing the graft lateral to the glenoid articularsurface and to thoroughly prepare to obtain boneunion.It is important for surgeons to understand the typical

movements that are required to fully return to athleticcompetition in a particular sport. For rugby players, notonly did the risk of redislocation during normal tacklingmovements affect their play, but also apprehension hada significant influence when the shoulder was at themidrange position and the arm was extended. There-fore surgeons should attempt to obtain shoulder sta-bility at both the end range and midrange of shouldermotion.

LimitationsLimitations of this study include the small number of

cases and the subjective nature of a retrospective studywith postoperative assessment by a questionnaire. The

surgery. The arrows indicate the bony Bankart fragment. (B,C) Sagittal and coronal views after surgery. The coracoidfragment was fixed onto the repaired bony Bankart fragment,and these have achieved union.

BANKART REPAIR IN RUGBY PLAYERS 9

follow-up period was not long enough to evaluatepostoperative arthritis, which has been reported afterthe Bristow procedure. There was no control group orrandomization, and the described surgical procedurewas not compared with other surgical methods such asarthroscopic Bankart repair alone, a combination withthe Latarjet method, or all-arthroscopic surgery.Moreover, although not present in any of our patients,a large bony deficit of the glenoid and absence of thefragment may result in a different outcome.15 Return tosport was at the discretion of the team trainer managingthe athlete’s care in the postoperative period, althoughthis was generally based on our recommendations.

ConclusionsThis combined surgical procedure is effective in pre-

venting recurrent shoulder dislocation in rugby players;however, some players complained of insufficiency inthe quality of their play when they were tackling orperforming other specific movements.

AcknowledgmentThe authors are deeply grateful to Naoyoshi Sueishi

(professional photographer), Kazuhiro Tamura, andKaori Matsumoto (monthly rugby magazine, Tokyo,Japan) for excellent photographs of rugby-specificmovements. In addition, they are particularly gratefulfor the assistance given by Kay Daugherty.

References1. Balg F, Boileau P. The instability severity index score. A

simple pre-operative score to select patients for arthro-scopic or open shoulder stabilisation. J Bone Joint Surg Br2007;89:1470-1477.

2. Boileau P, Villalba M, Hery JY, Balg F, Ahrens P,Neyton L. Risk factors for recurrence of shoulder insta-bility after arthroscopic Bankart repair. J Bone Joint SurgAm 2006;88:1755-1763.

3. Holzer NA, Cunningham G, Ladermann A, Gazielly. Latar-jet-Patte Triple-locking Procedure for Recurrent AnteriorInstability. Tech Shoulder Elbow Surg 2013;14:63-68.

4. Young DC, Rockwood CA Jr. Complications of a failedBristow procedure and their management. J Bone JointSurg Am 1991;73:969-981.

5. Sugaya H, Kon Y, Tsuchiya A. Arthroscopic repair ofglenoid fractures using suture anchors. Arthroscopy2005;21:635.

6. Schroder DT, Provencher MT, Mologne TS, Muldoon MP,Cox JS. The modified Bristow procedure for anteriorshoulder instability: 26-year outcomes in Naval Academymidshipmen. Am J Sports Med 2006;34:778-786.

7. Rowe CR, Patel D, Southmayd WW. The Bankart proce-dure: A long-term end-result study. J Bone Joint Surg Am1978;60:1-16.

8. Tegner Y, Lysholm J. Rating systems in the evaluation ofknee ligament injuries. Clin Orthop Relat Res 1985;198:43-49.

9. Sugaya H, Moriishi J, Dohi M, Kon Y, Tsuchiya A. Glenoidrim morphology in recurrent anterior glenohumeralinstability. J Bone Joint Surg Am 2003;85:878-884.

10. Mizuno N, Yoneda M, Hayashida K, Nakagawa S, Mae T,Izawa K. Recurrent anterior shoulder dislocation causedby a midsubstance complete capsular tear. J Bone JointSurg Am 2003;85:878-884.

11. Fabbriciani C, Milano G, Demontis A, Fadda S, Ziranu F,Mulas PD. Arthroscopic versus open treatment of Bankartlesion of the shoulder: A prospective randomized study.Arthroscopy 2004;20:456-462.

12. Potzl W, Thorwesten L, Gotze C, Garmann S, Steinbeck J.Proprioception of the shoulder joint after surgical repairfor Instability: A long-term follow-up study. Am J SportsMed 2004;32:425-430.

13. Leroux T, Bhatti A, Khoshbin A, et al. Combinedarthroscopic Bankart repair and remplissage for recurrentshoulder instability. Arthroscopy 2013;29:1693-1701.

14. Boileau P,Mercier N, Old J. Arthroscopic Bankart-Bristow-Latarjet (2B3) procedure: How to do it and tricks tomake iteasier and safe. Orthop Clin N Am 2010;41:381-392.

15. Yamamoto N, Muraki T, An KN, et al. The stabilizingmechanism of the Latarjet procedure: A cadaveric study.J Bone Joint Surg Am 2013;95:1390-1397.

16. Boileau P, Bicknell RT, El Fegoun AB, Chuinard C.Arthroscopic Bristow procedure for anterior instability inshoulders with a stretched or deficient capsule: The “belt-and-suspenders” operative technique and preliminaryresults. Arthroscopy 2007;23:593-601.

17. Collin P, Rochcongar P, Thomazeau H. Treatment ofchronic anterior shoulder instability using a coracoid boneblock (Latarjet procedure): 74 cases. Rev Chir Orthop Rep-aratrice Appar Mot 2007;93:126-132 [in French].

18. Kim SH, Ha KI, Cho YB, Ryu BD, Oh I. Arthroscopicanterior stabilization of the shoulder: Two to six-yearfollow-up. J Bone Joint Surg Am 2003;85:1511-1518.

19. Boileau P, Mercier N, Roussanne Y, Thelu CE, Old J.Arthroscopic Bankart-Bristow-Latarjet procedure: Thedevelopment and early results of a safe and reproducibletechnique. Arthroscopy 2010;26:1434-1450.

20. Lafosse L, Lejeune E, Bouchard A, Kakuda C, Gobezie R,Kochhar T. The arthroscopic Latarjet procedure for thetreatment of anterior shoulder instability. Arthroscopy2007;23:1242.e1-1242.e5.

21. Allain J, Goutallier D, Glorion C. Long-term results of theLatarjet procedure for the treatment of anterior instabilityof the shoulder. J Bone Joint Surg Am 1998;80:841-852.

22. Hovelius L, Akermark C, Albrektsson B, et al. Bris-tow-Latarjet procedure for recurrent anterior disloca-tion of the shoulder. A 2-5 year follow-up study onthe results of 112 cases. Acta Orthop Scand 1983;54:284-290.

23. Hovelius L, Korner L, Lundberg B, et al. The coracoidtransfer for recurrent dislocation of the shoulder. Tech-nical aspects of the Bristow-Latarjet procedure. J BoneJoint Surg Am 1983;65:926-934.