- Email: [email protected]
Arthroscopic extra-articular Bankart repair for the treatment of recurrent anterior shoulder dislocation
Arthroscopic Extra-articular Bankart Repair for the Treatment of Recurrent Anterior Shoulder Dislocation Herbert Resch, M.D., Paul Povacz, M.D., Markus Wambacher, M.D., Gernot Sperner, M.D., and Karl Golser, M.D.
S u m m a r y : This study describes a new arthroscopic procedure for the stabilization of recurrent anterior shoulder dislocations. The technique involves two important features. The first is the anterior inferior transmuscular approach through the subscapularis muscle, which permits self-locking implants to be inserted into the anterior inferior third of the glenoid rim so that they oppose the direction of pull of the capsule. This approach was studied on 79 cadaveric shoulders before clinical application. The second feature is the extracapsular (extra-articular) location of the self-locking implants, which permits a superomedial capsular shift as required. The technique offers a high degree of capsular stability. Of a total of 318 patients undergoing this procedure, the first 100 shoulders (98 patients) were evaluated postoperatively at an average of 35 months (range, 18 to 62 months). The diagnosis in all cases was traumatic recurrent anterior shoulder dislocation. Repair of the capsule was performed initially with screws and later with absorbable tacks. The overall recurrence rate was 9% (9 shoulders). Excluding the first 30 shoulders to take account of the learning curve, the recurrence rate for the subsequent 70 shoulders was only 5.7%. Limitation of external rotation at 0 ° abduction averaged 6.7 ° and 6.1 ° at 90 ° abduction for all shoulders; 61% of participants in overhead sports and 70% of participants in contact sports resumed their preinjury activities. The recurrence rate for patients involved in overhead sports was 10% and for collision sports it was 14%. There were no recurrences in the case of patients whose sports involve minimum risk to the shoulder (cycling, jogging). Most recurrences were observed in patients with lax shoulders and small Bankart lesions. Key Words: Anterior inferior portal--Transmuscular approach--Extra-articular fixation.
ince the first arthroscopic operation for shoulder instability p e r f o r m e d b y L. Johnson in 1982, num e r o u s articles have b e e n published on a variety o f arthroscopic stabilisation techniques. 1-9 In spite o f some g o o d results p u b l i s h e d b y Caspari, 2 Morgan, m and W o l f et al. H the recurrence rate quoted in m o s t reports is in the range o f 10% to 2 0 % 3'4'12-14 c o m p a r e d
S
From the General Hospital of Salzburg (H. R., P.P.), and the University Hospital of lnnsbruck (M. W., G. S., K. G.), Austria. Address correspondence and reprint requests to Herbert Resch, M.D., Landeskrankenanstalten Salzburg, Unfallchirurgie, M~illnerHauptstra~e 48, 5020 Salzburg, Austria. © 1997 by the Arthroscopy Association of North America 0749-8063/97/1302-145053.00/0
188
with a figure o f 5% or lower for open surgery in most series. 15-19One o f the m a i n contributing factors responsible for the poorer results achieved with arthroscopic procedures is doubtless the location o f the subscapularis muscle, w h i c h is an obstacle to access to the anterior inferior region o f the glenoid. In the open p r o c e d u r e described b y Bankart, 2° the subscapularis tendon is d e t a c h e d to create a direct approach to the anterior inferior region o f the glenoid. 17-2° If a portal is selected superior to the subscapularis tendon, the instrument will be located at an acute angle or tangentially to the inferior region o f the glenoid rim, which poses p r o b l e m s in terms o f the stability o f the repair in that region. 21 To c o m p l e t e l y imitate the open Bankart
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 13, No 2 (April), 1997: pp 188-200
EXTRA-ARTICULAR B A N K A R T REPAIR
procedure using arthroscopic techniques, the surgeon would have to travel through the subscapularis with a blunt instrument. As such, this portal would lie inferior to the coracoid process and, therefore, would be in contradiction to the accepted rule that approaches on a line inferior to the coracoid process be avoided to prevent injury to the vessels and nerves of the brachial plexus. 22'23 Like Davidson et al., 21 we performed an anatomic study to find an anterior portal that travels through the subscapularis muscle to reach the glenoid rim without putting the adjoining neurovascular structures at risk. 24 This portal can be used to permit selflocking implants to be inserted at any point on the anterior glenoid rim for capsular repair. The implant is secured against the outside surface of the capsule (extra-articular fixation), which permits not only fixation on the glenoid rim regardless of the condition of the labrum but also a superomedial shift of an overstretched capsule, which seems to be extremely important for the success of the repairY -27
MATERIAL AND METHODS Between 1989 and 1995, a total of 318 patients presenting with recurrent anterior dislocation of the shoulder underwent surgery at Innsbruck University Hospital and Salzburg General Hospital using the arthroscopic extra-articular Bankart repair technique (AEBR). (The senior author moved from Innsbruck to Salzburg in 1993.) Only those patients who had a substantial bony Bankart lesion (more than cortical thickness) or poor capsule condition were operated on by open techniques. Of these 318 patients, the first 100 shoulders (98 patients), which had been operated on between 1989 and 1993, were given a clinical and radiological follow-up after an average of 35 months (range, 18 to 62 months). The study is based on a consecutive sequence of patients treated exclusively for traumatic recurrent anterior shoulder dislocations. No cases of atraumatic instability were included. Patients presenting with a single dislocation and patients with recurrent subluxation were excluded from the study. All patients had agreed to the arthroscopic procedure after explanation of the risks, and the technique used was identical in every case. The only difference involved relates to the implant used for repair of the capsulolabral complex (see below). The mean age at first dislocation was 22.6 years (range, 13 to 52 years) and at operation 26.8 years (range, 14 to 55 years). The group comprised 76 men and 22 women. In 56 cases, it was the right shoulder
189
and in 40 cases, the left shoulder that was affected; in 2 patients both shoulders were affected. The dominant shoulder was involved in 56 cases. The average number of dislocations before reconstruction was 4.7 (range 2 to 50). In 72 shoulders, dislocations occurred fewer than 10 times, and in 28 shoulders more than 10 times. An average of 4.6 years (range, 5 to 120 months) had elapsed between the first dislocation and repair. Ninety-seven of the 98 patients indicated that they participated in some kind of sport preoperatively. In 79% of the cases, the first dislocation occurred during such sports activity. The type of athletic activity was classified into 3 groups based on the degree of shoulder involvement: (1) overhead sports, (2) Shoulder-demanding sports (e.g., football, skiing, icehockey, golf) and (3) Shoulder-sparing sports (e.g., cycling, jogging). Forty-nine patients participated in an overhead sport preoperatively, 44 patients in a shoulder-demanding sport, and 4 patients in a shoulder-sparing sport. In addition to this classification, a separate group of contact sports was also identified comprising 20 patients. At preoperative clinical examination, all patients had a positive apprehension test result. Eight patients had a positive sulcus sign with a gap of more than 2 cm between the acromion and the humeral head. These patients were classified as clinically lax. For all patients, the preoperative radiographic protocol consisted of anteroposterior radiographs taken in external rotation plus Bernageau's glenoid view. 2s Computed tomographic scans were also taken on 62 patients. Postoperative review included further anteroposterior radiographs taken on discharge from the hospital, at 6 months, and also at 1 year postoperatively to monitor the tightness of the screws. Fisher's Exact Test was used to calculate statistical significance. The special feature of the operative technique used is the transmuscular (trans-subscapular) approach to the anterior inferior glenoid rim. The approach was studied on cadaveric shoulders 24 before clinical application.
Anatomic Study The anatomic study was performed on 79 cadaveric shoulders. 24 The reference point selected on the anterior glenoid rim was 4:30 o'clock (right shoulder) and 7:30 o'clock (left shoulder). The distance was measured from the reference point on the glenoid rim to the coracoid process and also to the adjacent nerves. Musculocutaneous Nerve: The distance between the palpable coracoid process and the intersection of
190
H. RESCH E T AL.
the musculocutaneous nerve and the medial margin of the conjoined tendon averaged 5.3 cm (2 to 11.2 cm). In only three shoulders (3.8%) was the distance less than 3 cm, and in no case was it less than 2 cm. Axillary Nerve: The shortest distance between the axillary nerve and the 4:30 (7:30) o'clock position on the glenoid rim measured transversely was 2.5 cm on average (1.8 to 3.2 cm). Coraeoid Process: The average distance between the coracoid process and the 4:30 (7:30) o'clock position on the glenoid rim was 1.8 cm (1.6 to 2.7 cm). To reach the 4:30 (7:30) o'clock position by a direct approach, the portal must be located 1.5 to 2 cm inferior to the coracoid process, and the instrumentation must travel through the subscapularis muscle. It is important that the initial incision not be made medial to an imaginary line originating at the coracoid process and extending inferiorly and parallel to the shaft of the upper arm. To further reduce the risk of injury to the musculocutaneous nerve on clinical application, the conjoined tendon of coracobrachialis and short head of biceps is passed laterally by changing the direction of advance with the cannula (slalom approach, see below).
Technique Used for AEBR Instruments: Repair of the capsulolabral complex to the glenoid rim was performed with two different types of implants: Screws (Oswald Leibinger, Freiburg, Germany):, the screws used were 16 × 2.7 mm cannulated self-tapping titanium screws with washers. They were screwed in with the help of a 1-mm thick guidewire using a special cannulated screwdriver. 7 Absorbable tacks (Suretac; Smith & Nephew Endoscopy Inc, Andover, MA): At the beginning of the study, tacks with a 6-mm head (Suretac I) were used. In the last 3 years, a particularly suitable type of tack has been employed (Suretac II). This model has an 8mm head fitted with spikes on the underside. The tacks are inserted by means of a metal cannula system (IXC instrumentation; Smith & Nephew Endoscopy Inc) Drilling for the guidewire and holes for the implants is performed in one step using a cannulated drill system. When drilling is complete, the locking mechanism is released and the drill withdrawn leaving the guidewire in place. Patient Positioning: The patient is placed in a beach-chair position (Fig l) with the upper part of the body at an angle of about 50°) 9 The patient's arm is placed in an arthroscopic elbow brace (Smith &
FIG 1. Supine position with patient's arm in an elbow brace for 90° immobilizationwith 2-kg traction weight.
Nephew Endoscopy Inc) and the elbow immobilized at an angle of 90 °. A traction weight of 2 kg is applied to the brace via a pulley. The traction maintains the arm in 20 ° to 30 ° of abduction. Portals: Besides the classic posterior portal for the arthroscope, two anterior portals are required (Fig 2) (1) The anterior superior portal is located slightly superior to the coracoid process. This more cranial orientation increases the distance from the anterior inferior portal, which facilitates simultaneous working via the two portals. (2) The anterior inferior portal is located 1.5 cm inferior to the coracoid process on a line originating from the palpable coracoid process and running inferiorly and parallel to the humeral shaft. This portal is for the transsubscapular approach. Preparation of the Glenoid Rim: Working via the anterior superior portal, a shoulder elevator is used to detach the traumatised glenoid labrum and the capsule from the bony glenoid rim and adjacent scapular neck. The glenoid rim is then abraded with a 4.5 mm arthroplasty burr and two or three shallow notches (about 1 mm deep) cut in the glenoid rim at 5 o'clock, 3.30 and possibly at 2 o'clock (right shoulder) (Fig 3). These notches must be just deep enough to ensure that the full width of the bony glenoid rim and the cortex are visible in the 30 ° wide-angle arthroscope. Repair Technique: With the surgeon's assistant handling the arthroscope, the surgeon positions himself in front of the patient's shoulder. The patient's arm, which is immobilised at a right angle in the elbow brace, is placed in 20 ° to 30 ° of external rotation. The patient's forearm rests against the surgeon's chest. At
EXTRA-ARTICULAR BANKART REPAIR
/
/
/
]
/ F
F I G 2. Anterior portals for AEBR. The anterior inferior portal is located between 1.5 and 2 cm inferior to the palpable coracoid process on a line running parallel to the axis of the upper arm. The anterior superior portal is slightly superior to the apex of the coracoid process.
191
through this portal and advanced straight ahead until the subcutis is perforated. The trocar sheath is then swivelled posterolaterally approximately 45 ° and advanced until firm resistance is encountered (humeral head). Sliding along the humeral head, the trocar sheath is redirected medially and advanced in a posteromedial direction (Fig 5). In this phase, the blunt end of the trocar is kept constantly pressed against the humeral head. After passing over the tendon-muscle junction, the blunt trocar penetrates the muscle tissue of the subscapularis and, on arrival at the joint line, depresses the capsule into the joint. The trocar does not pass through the tendon but penetrates the muscle tissue of subscapularis as the small amount of external rotation in the patient's arm holds the tendon lateral to the anterior glenoid rim. Depression of the capsule into the joint is clearly visible in the arthroscope. A 1-mm guidewire is then introduced through the cannulated blunt trocar and the depressed capsule, and respective of the inferior glenohumeral ligament perforated at the desired point (Fig 6). The speared capsule is then repositioned on the inferiormost notch made on the glenoid rim and the capsule shift assessed (Fig. 7). The correct perforation point on the capsule has been found if the capsule is properly tightened superomedially with the patient's arm in the position described above. If this is not the case, the capsule is perforated again at a different point. Once the correct perforation point has been located, the trocar sheath with the capsule is returned to the joint gap, and the sheath is advanced as far as possible. The serrated end of the trocar prevents
/
this point, an approximately 10-cm thick sterilized Arthroscopic Shoulder Roll (Smith & Nephew Endoscopy Inc) is placed in the patient's axilla to serve as a fulcrum to distract the joint. The cord for the traction weight passes round the surgeon's hip on the patient's side and stabilizes the patient's arm in this position so that the surgeon has both hands free for the operation (Fig 4). The anterior inferior portal is located about 1.5 cm inferior to the coracoid process on a line running distally from the coracoid process and parallel to the humeral shaft. A spinal needle is then inserted into the joint to determine under arthroscopic control the exact location of this portal on the patient's skin. A metal trocar sheath with a blunt trocar is introduced
F I G 3. Preparation of the anterior glenoid rim; burring two or three small notches in the 5, 3.30, and 2 o'clock positions (right shoulder).
192
H. R E S C H E T AL.
/I FIG 8. Slalom approach: the trocar sheath with blunt trocar is inserted via the anterior inferior portal and at first advanced posterolaterally. On contact with the humeral head, the line of advance is changed to posteromedial and the trocar slides along the tendon of subscapularis. The muscle tissue is penetrated and the capsule depressed into the joint.
J
!!
F I G 4. Surgeon's position relative to the patient. Patient's arm flexed at a right angle in approximately 30 ° of external rotation and resting on the surgeon's chest, with the cord for the traction weight passing round the surgeon's hip. Arthroscopic roll in the patient's axilla used to distract the joint.
the capsule from slipping. The blunt trocar with the K-wire is then removed. Now the cannulated drill holding the guidewire is introduced into the trocar, and the capsule is perforated at the same point and moved back to the inferior notch (Fig 8) (The capsule is not speared with the drill-guidewire combination from the very beginning, because repeated perforation with the thicker guidewire or even drill would destroy the capsule). The drill with the guidewire is placed in the bony notch close to the cartilage under arthroscopic control. It is advanced to 18-mm maximum depth (Fig 9), the locking mechanism released to free the guidewire, and the guidewire impacted with a driver to ensure optimum anchorage in the bone. Then the drill is extracted. An absorbable tack is introduced via the guidewire and seated with the cannulated Suretac
driver (Figs 10 and 11). This procedure is repeated for the middle notch, with attention paid once again to good capsule tension superomedially. If the glenoid labrum is also detached superior to the midline of the glenoid cavity, repair is performed by working intra-
/ J7
/z/ , i/
FIG 6. Depressing the capsule into the joint with the blunt end of the trocar and perforation with the K-wire.
EXTRA-ARTICULAR B A N K A R T REPAIR
193
A
.
/"
FIG 7. Repositioning of the speared capsule on the inferiormost notch (5 o'clock) and assessment of craniomedial capsule tension. Where tension is excessive or inadequate, a new perforation point is selected.
articularly at this level. G i v e n the undulating shape o f the glenoid rim, p l a c e m e n t o f the implants within the notches tightens the capsule b e t w e e n t h e m so that a watertight seal is obtained (Figs 12 and 13). Postoperative Management: Postoperatively, the p a t i e n t ' s arm is i m m o b i l i z e d in a light shoulder bandage for 3 weeks. F r o m w e e k s 4 to 6, up to 90 ° of flexion and external rotation to neutral are permitted. A f t e r the sixth week, exercise is p e r m i t t e d in all planes, including external rotation. Sports activities m a y be r e s u m e d f r o m the fourth m o n t h p o s t o p e r a t i v e l y except for those involving throwing actions, which are not a l l o w e d until the fifth month.
RESULTS O v e r a l l R e s u l t s : On the basis o f the R o w e score, 18 74% o f the results were classified as very good, 9%
FIG 8. Placement of the tip of the guidewire and drill. (A) Placement of the guidewire in the middle of the inferiormost notch under arthroscopic control. (B) Arthroscopic view of placement of the guidewire in the middle of the notch. C, cartilage; T, trough, CO, cortex.
J
194
H. RESCH E T AL.
/ /;
FIG 11. Extra-articular (extracapsular) location of the tack. The capsule is tightened (arm in 30 ° external rotation!).
i
F I G 9. Drilling in the guidewire and drill as far as the stop. The locking mechanism is released and the guidewire hammered home.
as good, 8% fair, and 9% poor. Discounting the first 30 shoulders to account for the development of the technique, and evaluating only the subsequent 70 shoulders, the overall results graded on the Rowe system were 76% very good, 12% good, 6% fair, and 6% poor. Stability: At the time of follow-up, 9 patients (9 shoulders) suffered a recurrence (9% recurrence rate), with 7 dislocations and 2 subluxations. Again, discounting the first 30 shoulders which had a recurrence rate of almost 17% to take account of the developmental phase of the technique, the recurrence rate for the subsequent 70 shoulders was only 5.7%. In 7 of
/
,/ // i
the 9 patients who suffered a recurrence, trauma of varying severity was the causal factor. Glenoid Rim Pathology: Intraoperative findings included varying degrees of displacement or detachment of the labrum from the glenoid rim in 72 shoulders. In 25 cases, the capsulolabral complex had formed a deep pocket following detachment from the glenoid rim and scapular neck. In three shoulders, no detachment of the labrum was observed (Fig 14). Changes to the labrum were restricted to the anterior inferior half of the glenoid in 42 shoulders (type A), while they affected the entire anterior glenoid rim from
\
i )
FIG 10. After removal of the drill with the help of a grip, an absorbable tack is inserted and hammered home.
FIG 12. Capsule refixed on the glenoid rim. The two lower implants are always extra-articular, the upper implant often intra-articular. Placement of the implants in the notches tensions the capsule between the notches, creating a watertight seal.
EXTRA-ARTICULAR BANKART REPAIR
195
shoulders 80 ,0
~"' ~ !!:~I~i
2 0
t detatchment from rim
FIG 14.
FIG 13. Radiographic view with implant position (Suretacs) shown as tunnels.
the 12 o'clock to the 6 o'clock position in 40 shoulders (type B) and extended beyond the 12 to 6 o'clock zone in 15 shoulders (type C). In 6 shoulders (6%), an additional type II SLAP lesion 3° was noted. Patients with no or only minor Bankart lesions (type A) and 3 patients without labral detachment had a recurrence rate of 13%, whereas patients with extensive Bankart lesions (types B and C) had a recurrence rate of 5% (Fig 15). The difference is statistically significant. In a total of 11 shoulders, a small glenoid rim avulsion fracture was noted preoperatively either on the computed tomographic arthrogram or the plain radiographs. Patients with larger defects underwent reconstruction using open techniques. The recurrence rate for these patients was 9%, which is identical to the overall recurrence rate. N u m b e r a n d T y p e of I m p l a n t s : A single implant was used in 11 shoulders, two in 52 shoulders, three in 34 shoulders, and four in 3 shoulders (Fig 16). The recurrence rate for shoulders with one implant was
detatchment
no detatchment
from rim and neck
Aspect of labrum lesion in 100 shoulders.
about 18% compared with 2.9% for three implants. The difference was statistically significant. In 25 shoulders, screws alone were used and in 53 shoulders, Suretac anchors were used. A combination of the two was used in 22 shoulders. The recurrence rate for screws alone was 8% compared with 9.4% for the absorbable tacks. The difference was not significant. Laxity: Of the 8 shoulders displaying laxity at preoperative examination (positive sulcus sign), recurrent instability occurred in 3 shoulders (37%), whereas the recurrence rate for all other shoulders was 6.5%. The difference was statistically significant. Hill-Sachs Lesion: A Hill-Sachs lesion of variable size was noted in 92 shoulders. In 15 shoulders, the lesion was only chondral or subchondral. In 59 shoulders, it was less than 5-ram deep whereas in 8 shoulders it was more than 5 m m deep (Fig 17). The recurrence rate for patients with chondral or subchondral Hill-
s h : : l ~
iA = =
[]
i
N,~
35
30 J % ~
A FIG 18.
B
C
inf. half ant. rim ant. rim ant. rim + posterior no detatchment
I
D
Extent of detachment of labrum in 100 shoulders.
196
H. RESCH E T AL.
shoulders
patients
6o
~-~
60 50
40
~i~i::i,i;;i:
40 30
20 10
I
:I:,N
~,,,~,o
10
N 1 F I G 16.
2
4
3
Implants
N u m b e r of implants in 100 shoulders.
Sachs lesions was 20%, and for lesions with a depth of more than 5 m m it was 11%. The difference was not significant. N u m b e r of Dislocations: The number of dislocations suffered preoperatively was below ten in 78 shoulders and above 10 in 22 shoulders. In the former group the recurrence rate was 7.6% compared with 13.6% for the second group. The difference was not significant. Mobility: Average limitation of external rotation measured at 0 ° abduction compared with the contralateral shoulder was 6.7 ° and at 90 ° abduction it was 6.1 °. In 6 shoulders, limitation of flexion and in 8 shoulders limitation of abduction by more than 10 ° was noted compared with the contralateral shoulder. Internal rotation was unrestricted in all cases. R e t u r n to Sports: All patients were asked before and after their operation to indicate their sports inter-
shoulders 60
40
~:#i,
I
subchondr. F I G 17.
< 5mm
> 5mm
I
none
Hill-Sachs lesion in 100 shoulders.
0
Overhead
demand, sparing collision no sport
FIG 18. Return to sports activity groups for 98 patients.
ests. Only one patient claimed not to have any athletic pursuits preoperatively. With the exception of this one case, all patients indicated that they returned to some degree of athletic participation after the operation, albeit not always the sport played preoperatively. Of a total of 49 patients who played a form of overhead sport preoperatively, 30 returned to that type of activity postoperatively. This represents a return rate of 61%. Of the 44 patients who played a shoulder-demanding form of sport preoperatively, 37 (84%) returned to their sport. Overall, there was an increase of 16 patients within this group if one includes all those who either changed from one group of sports to the other or changed their sport within the group (35% increase). This increase derives primarily from a switch from the overhead group to the shoulder-demanding group. All four patients who said they had participated in a shoulder-sparing type of sport preoperatively returned to that activity (100% return rate). This group was joined postoperatively by a further three patients from the shoulder-demanding group, making a total of seven patients. Before the operation, 20 patients were involved in some kind of contact sport, 14 of whom (70%) returned to their sport postoperatively (Fig 18). The figures for returning to the individual sports are shown in Fig 19. Figure 20 shows the case of a change of sport taking the preoperative tennis players as an example. Regaining premorbid levels of performance in sport was found to occur at two peaks on the time axis. The first peak was seen at about 6 months postoperatively (14%) with the second ocurring at 12 months (21°7o) (Fig 21). Strength: A spring balance was used to measure internal rotation force from the fully externally rotated
EXTRA-ARTICULAR BANKART REPAIR
197
patie~ 45 40 35 30 25 20 15 10 5
0
ski
tennis
soccer climbing snow- volleyboard ball
2
4
6
8
10
12
20
mos.
FIG 21. Time of return to former performancelevel in sports.
FIG19. Return to the individual sports activities.
position in comparison with the contralateral shoulder. The patient's arm was adducted and flexed to form a right angle at the elbow. The difference between the two sides was 0.53 kg on average and was not statistically significant. Analysis of Postoperative Recurrences: Of a total of nine shoulders that suffered a recurrence postoperatively, repair of the capsule had been performed with just one implant in two shoulders (representing a recurrence rate of 18% for all shoulders reconstructed with just one implant). In both cases, a minor Bankart lesion (type A) had been noted, which was the reason for using one implant None. Three shoulders that suffered a recurrence had been described as lax preoperatively (37% of all such shoulders). These three shoulders include the above two shoulders with minor Bankart lesions in which just one implant had been used. The third shoulder presented with a medium-size Bankart lesion (type B) and was reconstructed with two implants. Seven of the nine patients who suffered a recurrence did so while playing some form of sport. At the time of the recurrence, three patients were active in an overhead sport (1 handball, 1 tennis, 1 rock-climbing)
FIG 20. Changefrom tennis to other sports activities after surgery in 11 patients.
and four patients in a shoulder-demanding sport (2 snowboarding, 1 skiing, 1 football). An additional two patients sustained a recurrence while participating in a contact sport. Relative to the total number of patients in the various sports groups, the recurrence rate was 10% for the overhead group, 6.6% for the shoulderdemanding group, 0% for the shoulder-sparing group, and 14.2% for the group involved in contact sports. With the exception of the shoulder-sparing group, the differences were not significant. C h a n g e in Technique: In six of the cases in this study, the original goal of AEBR had to be abandoned intraoperatively and an open procedure substituted. In three cases with a clinically lax shoulder, no Bankart lesions and a large, thin joint capsule was found necessitating an open capsular shift procedure. In three other cases, the joint capsule was found intraoperatively to be very thin and weak (plastic deformation) in the area of the inferior glenohumeral ligament, and open Bankart repair was performed. Complications: A distinction must be made between complications involving screws and those involving the absorbable tacks. In four patients, the screw was placed too close to the cartilage with the result that the washer penetrated the capsule, causing metal impingement on the humeral head requiring removal. No screws fractured but one did eventually loosen. In the case of the absorbable tacks, the head penetrated the capsule causing button holing in six cases, and an additional tack had to be inserted. In four cases of reconstruction using the absorbable tacks, a slight increase in body temperature was noted between weeks 3 and 5 postoperatively, along with an increase in C-reactive protein accompanied by a feeling of general fatigue. These symptoms disappeared within 3 days in three of the patients, whereas an arthroscopic
198
H. RESCH ET AL.
lavage was required in the fourth patient, after which the symptoms immediately abated. We believe this may have represented a foreign body reaction to the synthetic material. DISCUSSION The open Bankart technique offers stable repair of the capsulolabral complex to the glenoid rim over the whole area of detachment using transosseous sutures or suture anchors. 16-2°'31 At the same time, a capsular shift can be performed to shorten the overstreched joint capsule. At arthroscopy, the two key elements, namely insertion of the implants in the region of the inferior third of the glenoid rim and execution of a capsular shift superomedially, can be satisfied by selecting the transsubscapular approach with extracapsular placement of the implants. As noted in the literature, subcoracoid portals are located in a potential hazard zone for the neurovascular structures) 2'23 With the help of anatomic studies, however, it has been possible to minimize the risks posed by this approach providing certain criteria are observed. 24 Unlike the procedure adopted by Davidson et al., who describe a similar transmuscular portal, 23 the slalom technique as described above offers an additional degree of safety, because the conjoined tendon is passed laterally. With a skin incision made at a maximum of 2 cm inferior to the coracoid process, the path of the blunt trocar lies not only superior to the musculocutaneous nerve but also lateral to it by the width of the conjoined tendon. As the anatomic studies show, there can be no risk to the neurovascular structures as long as the blunt trocar slides horizontally and maintains contact with the humeral head. In the anatomic studies performed, it was also shown that, on external rotation of the arm starting from the neutral position, the tendon of the subscapularis muscle lies lateral to the anterior glenoid rim. Therefore, as the blunt trocar glides along the humeral head, it does not pass through the tendon but easily penetrates the muscle tissue of the subscapularis. Movement of the humeral head beyond the glenoid rim overstretches the joint capsule and may cause plastic deformation of the capsuleY 27'32 Consequently, shifting the anterior capsule is just as important a part of arthroscopic stabilization as it is in open surgery. 16-18 The question of how much capsular shift is needed is best determined by shifting and tightening the capsule superomedially with the patient's arm maintained in an appropriate position. Thirty degrees of external rotation was selected for the position of the patient's arm
to duplicate the position the authors have used successfully during the open Bankart technique. 17 The right angle formed by the patient's arm in the elbow brace permits accurate adjustment to be made to arm rotation via the patient's forearm. There are a number of advantages to the notches burred in the glenoid rim. They permit the guidewire to be placed not only lateral to the cortex under arthroscopic control, thus avoiding the risk of deflection medially, but also a precise placement close to the cartilage to obtain a stepless transition between glenoid and capsule. In addition, placement of the implants in the notches also tensions the capsule between the notches, thus creating a watertight seal between the capsule and the glenoid rim, thereby preventing seepage of synovial fluid between the capsule and the rim over it' s entire length. Both the watertight seal and the smooth transition between glenoid and capsule help to regain the suction cup effect of the glenoid. 33 The screws used exclusively for fixation in the early phase led to complications in a number of cases in the form of metal impingement on the humeral head. For that reason, screws are not used anymore for that purpose. In the case of the absorbable tacks with the 6mm diameter head, the head repeatedly penetrated the capsule. The almost exclusive use over the last 2 years of the absorbable tacks with the bigger diameter head (Suretac II) has completely solved this problem. The systemic foreign body reaction observed in four patients was self-limited in three cases, disappearing after a few days. The arthroscopic irrigation performed on the fourth patient led to an immediate abatement of the symptoms. In general, the clinical course was similar to that reported by Edwards et al.34 The introduction of a new operating technique always involves a developmental phase and a learning curve. That explains the 17% recurrence rate in the first 30 shoulders. For the subsequent 70 shoulders, the rate was only 5.7%, which is comparable with the results achieved with open repair procedures. The minimum follow-up period of 18 months can be considered adequate for an assessment of the results as there is general agreement in the literature that recurrences peak between the 13th and 15th months postoperatively, and decrease strongly after the 18th month, 26'27although recurrences still can occur. 35 An analysis of the failures reveals a significantly large proportion of patients who were classified as " l a x " on preoperative clinical examination. As a result of the minor shearing forces that develop on release of the humeral head from the glenoid, most of these patients also had minor Bankart lesions.
EXTRA-ARTICULAR BANKART REPAIR
It was initially felt that a small Bankart lesion represented a good indication for arthroscopic repair. The smaller the Bankart lesion, the smaller were the n u m ber of implants. However, patients who u n d e r w e n t reconstruction using only one i m p l a n t suffered a significant increase in the n u m b e r of recurrences. Similar observations have been made by other authors. 4'26'27'36 Consequently, it must be concluded that the smaller the Bankart lesions, the more attention must be paid during surgery to the size and condition of the j o i n t capsule. In the case of very p r o n o u n c e d laxity, which is n o r m a l l y accompanied by a large and thin capsule, an open capsular shift is i n d i c a t e d ) 7 The 61% return rate for overhead sports is largely in agreement with the return rate quoted by W a r n e r and W a r r e n s and is well above the return rates after open surgery as reported by Rowe et el. 18 Torg et al., 38 and the authors/7 O n l y Jobe et al. 16 report a higher return rate on the basis of their open repair technique. A limitation on external rotation of only 6 ° is the key factor in the high return rate noted for overhead sports. The arm position selected for repair of the capsule as described above permits the correct degree of capsular shift to be performed. A l m o s t all patients who gave up an overhead sport postoperatively, took up a sport from the s h o u l d e r - d e m a n d i n g group instead, so that this group grew by essentially the same amount. According to our results, the risk of suffering a recurrence while participating in a contact sport was only slightly higher than for other activities. The only significant difference was in comparison with shoulder-sparing sports, which proves that a system for classifying sports on the basis of the loads imposed on the glenohumeral j o i n t is urgently required,
REFERENCES 1. Benedetto KP, G1/StzerW. Arthroscopic Bankart procedure by suture fixation technique: Indications, technique, and results. Arthroscopy 1992;8:111-115. 2. Caspari RB. Arthroscopic reconstruction for anterior shoulder instability. Tech Orthop 1988;3:59-66. 3. Johnson LL. Diagnostic and surgical arthroscopy of the shoulder. St Louis: Mosby, 1993;276-352. 4. Landsiedl F. Arthroscopic therapy of recurrent anterior luxation of the shoulder by capsular repair. Arthroscopy 1992;8:296-304. 5. Maki NJ. Arthroscopic stabilizationfor recurrent shoulder instability. Orthop Trans 1989;13:508. 6. Morgan CD, Bodenstab AB. Arthroscopic Bankart suture repair: Techniques and early results. Arthroscopy 1987;3:111-122. 7. Resch H, Golser K, Kathrein A. Arthroscopic screw fixation techniques. In: Resch H, Beck E, eds. Arthroscopy of the shoulder. New York: Springer-Vertag, 1992;99-112. 8. Warner JP, Warren RF. Arthroscopic Bankart repair using a cannulated, absorbable, fixation device. Oper Tech Orthop Surg 1991; 1:192-198.
199
9. Wiley AM. Arthroscopyfor shoulder instability and a technique for arthroscopic repair. Arthroscopy 1988;4:25-30. 10. Morgan CD. Arthroscopic transglenoid Bankart suture repair. Oper Tech Orthop 1991;1:171. 11. Wolf EM, Wilke RM, Richmond JC. Arthroscopic Bankart Repair using suture anchors. Oper Tech Orthop 1991;1:184. 12. Habermeyer P, Wiedmann E. Arthroscopic three-point Bankart suture repair. In: Resch H, Beck E, eds. Arthroscopy of the shoulder. New York: Springer-Verlag, 1992;90-98. 13. Hawkins RB. Arthroscopic stapling repair for shoulder instability: A retrospective study of 50 cases. Arthroscopy 1989;5:122128. 14. Weber S. A prospective evaluation comparing open and arthroscopic treatment of recurrent anterior glenohumeral dislocation. Presented at the Annual Meeting of the American Orthopaedic Society for Sports Medicine, New Orleans, LA, February 11, t990. 15. Hovelius L, Akermark C, Albrektsson B, Berg El K6rner L, Lundberg B, WredmarkT. Brislow-LetarjetProcedurefor recurrent anterior dislocation of the shoulder. Acta Orthop Scand 1983;54:284-290. 16. Jobe FW, Giangarra CE, Kvitne RS, Glousman RE. Anterior capsulo labral reconstructionof the shoulder in athletes in overhand sports. Am J Sports Med 1991;19:428-434. 17. Resch H. The Bankart Procedure: A Modificationusing Transosseous Sutures. Orthop Traumatol 1993;2:19-28. 18. Rowe CR, Patel D, ShouthmaydWW. The Bankart procedure-A long-term End-result study. J Bone Joint Surg Am 1978;60: 1-18. 19. Thomas SC, Matsen FA. An approach to the repair of avulsion of the glenohumeral ligaments in the management of traumatic anterior glenohumeral instability. J Bone Joint Surg Am 1989; 71:506-513. 20. Bankart ASB. The pathology and treatment of recurrent dislocation of the shoulder joint. Br J Surg 1938;26:23-29. 21. Davidson PA, Tibone JE. Anterior inferior (5 o'clock) portal for shoulder arthrscopy. Arthroscopy 1995;5:519-525. 22. Matthews LS, Vetter WL, Oweida SJ, Spearman J, Helfet DL. Arthroscopic staple capsulorrhaphyfor reccurent anterior shoulder instability. Arthroscopy 1988;4:106-111. 23. Ogilvie-Harris DJ, Wiley AM. Arthroscopic surgery of the shoulder. J Bone Joint Surg Br 1986;68:201-207. 24. Resch H, Wykypiel HF, Maurer H, Wambacher M. The anteroinferior (transmuscular)approach for arthroscopic repair of the Bankart lesion--An anatomical and clinical study. Arthroscopy 1996; 12:309-322. 25. Bigliani LU, Pollock RG, Soslowski LJ, Flatow EL, Pawluk RJ, Mow VC. Tensile properties of the inferior glenohumeral ligament. J Orthop Res 1992;10:187-197. 26. Grana WA, Buckley PD, Yates CK. Arthroscopic Bankart suture repair. Am J Sports Med 1993;21:348-353. 27. Walch G, Boileau P, Levigne C, Mandrino A, Neyret P, Donell S. Arthroscopic stabilization for recurrent anterior shoulder dislocation: Results of 59 cases. Arthroscopy 1995;11:173-179. 28. Bernageau J, Platte D, Debeyre J, Ferrane J. IntEret du profil gl6noidiendans les luxations r6ccidivantesde l'6paule. Rev Chit Orthop 1976;62:142-147 (suppl II). 29. Skyhar MJ, AltcheckDW, Warren RF, WiekecwiczTL, O'Brien SJ. Shoulder arthroscopy with the patient in the beach-chair position. Arthroscopy 1988;4:265-269. 30. Snyder SJ, Karzel RP, DelPizzo W, Ferkel RD, Friedman MJ. SLAP lesions of the shoulder. Arthroscopy 1990;6:274-279. 31. Richmond JC, Donaldson WR, Fu F, Harner C. Modificationof the Bankart reconstruction with a suture anchor. Am J Sports Med 1991;19:343-346. 32. Adolfsson L, Lysholm J. Arthroscopy and stability testing for anterior shoulder instability. Arthroscopy 1989;5:315-320. 33. Habermeyer P, Schuller U, Wiedemann E. The Intra-articular
200
H. R E S C H E T AL.
pressure of the shoulder: An experimental study on the role of the glenoid labrum in stabilizing the joint. Arthroscopy 1992; 8: 166-172. 34. Edwards DJ, Hoy G, Sales AD, Hayes MG. Adverse reactions to an absorbable shoulder fixation device. J Shoulder Elbow Surg 1994;3:230-233. 35. Morrey BF, Jones JM. Recurrent anterior dislocation of the shoulder. Long-term follow-up of the Putti-Platt and Bankart procedures. J Bone Joint Surg Am 1976;56:252. 36. Green MR, Christiansen KP. Arthroscopic versus open Bankart
procedure. A comparison of early morbidity and complications. Arthroscopy 1993;9:371-374. 37. Neer CS II, Foster CR. Inferior capsular shift for involontary inferior and multidirectional instability of the shoulder. J Bone Joint Surg Am 1980;62:897-908. 38. Torg JS, Balduini FC, Bonci C, Lehman RC, Gregg JR, Esterhai JL, Hensal FJ. A modified Bristow-Helfet-May Procedure for recurrent dislocation and subluxation of the shoulder: A report of 212 cases. J Bone Joint Surg Am 1987; 69:904-913.
S u m m a r y : This study describes a new arthroscopic procedure for the stabilization of recurrent anterior shoulder dislocations. The technique involves two important features. The first is the anterior inferior transmuscular approach through the subscapularis muscle, which permits self-locking implants to be inserted into the anterior inferior third of the glenoid rim so that they oppose the direction of pull of the capsule. This approach was studied on 79 cadaveric shoulders before clinical application. The second feature is the extracapsular (extra-articular) location of the self-locking implants, which permits a superomedial capsular shift as required. The technique offers a high degree of capsular stability. Of a total of 318 patients undergoing this procedure, the first 100 shoulders (98 patients) were evaluated postoperatively at an average of 35 months (range, 18 to 62 months). The diagnosis in all cases was traumatic recurrent anterior shoulder dislocation. Repair of the capsule was performed initially with screws and later with absorbable tacks. The overall recurrence rate was 9% (9 shoulders). Excluding the first 30 shoulders to take account of the learning curve, the recurrence rate for the subsequent 70 shoulders was only 5.7%. Limitation of external rotation at 0 ° abduction averaged 6.7 ° and 6.1 ° at 90 ° abduction for all shoulders; 61% of participants in overhead sports and 70% of participants in contact sports resumed their preinjury activities. The recurrence rate for patients involved in overhead sports was 10% and for collision sports it was 14%. There were no recurrences in the case of patients whose sports involve minimum risk to the shoulder (cycling, jogging). Most recurrences were observed in patients with lax shoulders and small Bankart lesions. Key Words: Anterior inferior portal--Transmuscular approach--Extra-articular fixation.
ince the first arthroscopic operation for shoulder instability p e r f o r m e d b y L. Johnson in 1982, num e r o u s articles have b e e n published on a variety o f arthroscopic stabilisation techniques. 1-9 In spite o f some g o o d results p u b l i s h e d b y Caspari, 2 Morgan, m and W o l f et al. H the recurrence rate quoted in m o s t reports is in the range o f 10% to 2 0 % 3'4'12-14 c o m p a r e d
S
From the General Hospital of Salzburg (H. R., P.P.), and the University Hospital of lnnsbruck (M. W., G. S., K. G.), Austria. Address correspondence and reprint requests to Herbert Resch, M.D., Landeskrankenanstalten Salzburg, Unfallchirurgie, M~illnerHauptstra~e 48, 5020 Salzburg, Austria. © 1997 by the Arthroscopy Association of North America 0749-8063/97/1302-145053.00/0
188
with a figure o f 5% or lower for open surgery in most series. 15-19One o f the m a i n contributing factors responsible for the poorer results achieved with arthroscopic procedures is doubtless the location o f the subscapularis muscle, w h i c h is an obstacle to access to the anterior inferior region o f the glenoid. In the open p r o c e d u r e described b y Bankart, 2° the subscapularis tendon is d e t a c h e d to create a direct approach to the anterior inferior region o f the glenoid. 17-2° If a portal is selected superior to the subscapularis tendon, the instrument will be located at an acute angle or tangentially to the inferior region o f the glenoid rim, which poses p r o b l e m s in terms o f the stability o f the repair in that region. 21 To c o m p l e t e l y imitate the open Bankart
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 13, No 2 (April), 1997: pp 188-200
EXTRA-ARTICULAR B A N K A R T REPAIR
procedure using arthroscopic techniques, the surgeon would have to travel through the subscapularis with a blunt instrument. As such, this portal would lie inferior to the coracoid process and, therefore, would be in contradiction to the accepted rule that approaches on a line inferior to the coracoid process be avoided to prevent injury to the vessels and nerves of the brachial plexus. 22'23 Like Davidson et al., 21 we performed an anatomic study to find an anterior portal that travels through the subscapularis muscle to reach the glenoid rim without putting the adjoining neurovascular structures at risk. 24 This portal can be used to permit selflocking implants to be inserted at any point on the anterior glenoid rim for capsular repair. The implant is secured against the outside surface of the capsule (extra-articular fixation), which permits not only fixation on the glenoid rim regardless of the condition of the labrum but also a superomedial shift of an overstretched capsule, which seems to be extremely important for the success of the repairY -27
MATERIAL AND METHODS Between 1989 and 1995, a total of 318 patients presenting with recurrent anterior dislocation of the shoulder underwent surgery at Innsbruck University Hospital and Salzburg General Hospital using the arthroscopic extra-articular Bankart repair technique (AEBR). (The senior author moved from Innsbruck to Salzburg in 1993.) Only those patients who had a substantial bony Bankart lesion (more than cortical thickness) or poor capsule condition were operated on by open techniques. Of these 318 patients, the first 100 shoulders (98 patients), which had been operated on between 1989 and 1993, were given a clinical and radiological follow-up after an average of 35 months (range, 18 to 62 months). The study is based on a consecutive sequence of patients treated exclusively for traumatic recurrent anterior shoulder dislocations. No cases of atraumatic instability were included. Patients presenting with a single dislocation and patients with recurrent subluxation were excluded from the study. All patients had agreed to the arthroscopic procedure after explanation of the risks, and the technique used was identical in every case. The only difference involved relates to the implant used for repair of the capsulolabral complex (see below). The mean age at first dislocation was 22.6 years (range, 13 to 52 years) and at operation 26.8 years (range, 14 to 55 years). The group comprised 76 men and 22 women. In 56 cases, it was the right shoulder
189
and in 40 cases, the left shoulder that was affected; in 2 patients both shoulders were affected. The dominant shoulder was involved in 56 cases. The average number of dislocations before reconstruction was 4.7 (range 2 to 50). In 72 shoulders, dislocations occurred fewer than 10 times, and in 28 shoulders more than 10 times. An average of 4.6 years (range, 5 to 120 months) had elapsed between the first dislocation and repair. Ninety-seven of the 98 patients indicated that they participated in some kind of sport preoperatively. In 79% of the cases, the first dislocation occurred during such sports activity. The type of athletic activity was classified into 3 groups based on the degree of shoulder involvement: (1) overhead sports, (2) Shoulder-demanding sports (e.g., football, skiing, icehockey, golf) and (3) Shoulder-sparing sports (e.g., cycling, jogging). Forty-nine patients participated in an overhead sport preoperatively, 44 patients in a shoulder-demanding sport, and 4 patients in a shoulder-sparing sport. In addition to this classification, a separate group of contact sports was also identified comprising 20 patients. At preoperative clinical examination, all patients had a positive apprehension test result. Eight patients had a positive sulcus sign with a gap of more than 2 cm between the acromion and the humeral head. These patients were classified as clinically lax. For all patients, the preoperative radiographic protocol consisted of anteroposterior radiographs taken in external rotation plus Bernageau's glenoid view. 2s Computed tomographic scans were also taken on 62 patients. Postoperative review included further anteroposterior radiographs taken on discharge from the hospital, at 6 months, and also at 1 year postoperatively to monitor the tightness of the screws. Fisher's Exact Test was used to calculate statistical significance. The special feature of the operative technique used is the transmuscular (trans-subscapular) approach to the anterior inferior glenoid rim. The approach was studied on cadaveric shoulders 24 before clinical application.
Anatomic Study The anatomic study was performed on 79 cadaveric shoulders. 24 The reference point selected on the anterior glenoid rim was 4:30 o'clock (right shoulder) and 7:30 o'clock (left shoulder). The distance was measured from the reference point on the glenoid rim to the coracoid process and also to the adjacent nerves. Musculocutaneous Nerve: The distance between the palpable coracoid process and the intersection of
190
H. RESCH E T AL.
the musculocutaneous nerve and the medial margin of the conjoined tendon averaged 5.3 cm (2 to 11.2 cm). In only three shoulders (3.8%) was the distance less than 3 cm, and in no case was it less than 2 cm. Axillary Nerve: The shortest distance between the axillary nerve and the 4:30 (7:30) o'clock position on the glenoid rim measured transversely was 2.5 cm on average (1.8 to 3.2 cm). Coraeoid Process: The average distance between the coracoid process and the 4:30 (7:30) o'clock position on the glenoid rim was 1.8 cm (1.6 to 2.7 cm). To reach the 4:30 (7:30) o'clock position by a direct approach, the portal must be located 1.5 to 2 cm inferior to the coracoid process, and the instrumentation must travel through the subscapularis muscle. It is important that the initial incision not be made medial to an imaginary line originating at the coracoid process and extending inferiorly and parallel to the shaft of the upper arm. To further reduce the risk of injury to the musculocutaneous nerve on clinical application, the conjoined tendon of coracobrachialis and short head of biceps is passed laterally by changing the direction of advance with the cannula (slalom approach, see below).
Technique Used for AEBR Instruments: Repair of the capsulolabral complex to the glenoid rim was performed with two different types of implants: Screws (Oswald Leibinger, Freiburg, Germany):, the screws used were 16 × 2.7 mm cannulated self-tapping titanium screws with washers. They were screwed in with the help of a 1-mm thick guidewire using a special cannulated screwdriver. 7 Absorbable tacks (Suretac; Smith & Nephew Endoscopy Inc, Andover, MA): At the beginning of the study, tacks with a 6-mm head (Suretac I) were used. In the last 3 years, a particularly suitable type of tack has been employed (Suretac II). This model has an 8mm head fitted with spikes on the underside. The tacks are inserted by means of a metal cannula system (IXC instrumentation; Smith & Nephew Endoscopy Inc) Drilling for the guidewire and holes for the implants is performed in one step using a cannulated drill system. When drilling is complete, the locking mechanism is released and the drill withdrawn leaving the guidewire in place. Patient Positioning: The patient is placed in a beach-chair position (Fig l) with the upper part of the body at an angle of about 50°) 9 The patient's arm is placed in an arthroscopic elbow brace (Smith &
FIG 1. Supine position with patient's arm in an elbow brace for 90° immobilizationwith 2-kg traction weight.
Nephew Endoscopy Inc) and the elbow immobilized at an angle of 90 °. A traction weight of 2 kg is applied to the brace via a pulley. The traction maintains the arm in 20 ° to 30 ° of abduction. Portals: Besides the classic posterior portal for the arthroscope, two anterior portals are required (Fig 2) (1) The anterior superior portal is located slightly superior to the coracoid process. This more cranial orientation increases the distance from the anterior inferior portal, which facilitates simultaneous working via the two portals. (2) The anterior inferior portal is located 1.5 cm inferior to the coracoid process on a line originating from the palpable coracoid process and running inferiorly and parallel to the humeral shaft. This portal is for the transsubscapular approach. Preparation of the Glenoid Rim: Working via the anterior superior portal, a shoulder elevator is used to detach the traumatised glenoid labrum and the capsule from the bony glenoid rim and adjacent scapular neck. The glenoid rim is then abraded with a 4.5 mm arthroplasty burr and two or three shallow notches (about 1 mm deep) cut in the glenoid rim at 5 o'clock, 3.30 and possibly at 2 o'clock (right shoulder) (Fig 3). These notches must be just deep enough to ensure that the full width of the bony glenoid rim and the cortex are visible in the 30 ° wide-angle arthroscope. Repair Technique: With the surgeon's assistant handling the arthroscope, the surgeon positions himself in front of the patient's shoulder. The patient's arm, which is immobilised at a right angle in the elbow brace, is placed in 20 ° to 30 ° of external rotation. The patient's forearm rests against the surgeon's chest. At
EXTRA-ARTICULAR BANKART REPAIR
/
/
/
]
/ F
F I G 2. Anterior portals for AEBR. The anterior inferior portal is located between 1.5 and 2 cm inferior to the palpable coracoid process on a line running parallel to the axis of the upper arm. The anterior superior portal is slightly superior to the apex of the coracoid process.
191
through this portal and advanced straight ahead until the subcutis is perforated. The trocar sheath is then swivelled posterolaterally approximately 45 ° and advanced until firm resistance is encountered (humeral head). Sliding along the humeral head, the trocar sheath is redirected medially and advanced in a posteromedial direction (Fig 5). In this phase, the blunt end of the trocar is kept constantly pressed against the humeral head. After passing over the tendon-muscle junction, the blunt trocar penetrates the muscle tissue of the subscapularis and, on arrival at the joint line, depresses the capsule into the joint. The trocar does not pass through the tendon but penetrates the muscle tissue of subscapularis as the small amount of external rotation in the patient's arm holds the tendon lateral to the anterior glenoid rim. Depression of the capsule into the joint is clearly visible in the arthroscope. A 1-mm guidewire is then introduced through the cannulated blunt trocar and the depressed capsule, and respective of the inferior glenohumeral ligament perforated at the desired point (Fig 6). The speared capsule is then repositioned on the inferiormost notch made on the glenoid rim and the capsule shift assessed (Fig. 7). The correct perforation point on the capsule has been found if the capsule is properly tightened superomedially with the patient's arm in the position described above. If this is not the case, the capsule is perforated again at a different point. Once the correct perforation point has been located, the trocar sheath with the capsule is returned to the joint gap, and the sheath is advanced as far as possible. The serrated end of the trocar prevents
/
this point, an approximately 10-cm thick sterilized Arthroscopic Shoulder Roll (Smith & Nephew Endoscopy Inc) is placed in the patient's axilla to serve as a fulcrum to distract the joint. The cord for the traction weight passes round the surgeon's hip on the patient's side and stabilizes the patient's arm in this position so that the surgeon has both hands free for the operation (Fig 4). The anterior inferior portal is located about 1.5 cm inferior to the coracoid process on a line running distally from the coracoid process and parallel to the humeral shaft. A spinal needle is then inserted into the joint to determine under arthroscopic control the exact location of this portal on the patient's skin. A metal trocar sheath with a blunt trocar is introduced
F I G 3. Preparation of the anterior glenoid rim; burring two or three small notches in the 5, 3.30, and 2 o'clock positions (right shoulder).
192
H. R E S C H E T AL.
/I FIG 8. Slalom approach: the trocar sheath with blunt trocar is inserted via the anterior inferior portal and at first advanced posterolaterally. On contact with the humeral head, the line of advance is changed to posteromedial and the trocar slides along the tendon of subscapularis. The muscle tissue is penetrated and the capsule depressed into the joint.
J
!!
F I G 4. Surgeon's position relative to the patient. Patient's arm flexed at a right angle in approximately 30 ° of external rotation and resting on the surgeon's chest, with the cord for the traction weight passing round the surgeon's hip. Arthroscopic roll in the patient's axilla used to distract the joint.
the capsule from slipping. The blunt trocar with the K-wire is then removed. Now the cannulated drill holding the guidewire is introduced into the trocar, and the capsule is perforated at the same point and moved back to the inferior notch (Fig 8) (The capsule is not speared with the drill-guidewire combination from the very beginning, because repeated perforation with the thicker guidewire or even drill would destroy the capsule). The drill with the guidewire is placed in the bony notch close to the cartilage under arthroscopic control. It is advanced to 18-mm maximum depth (Fig 9), the locking mechanism released to free the guidewire, and the guidewire impacted with a driver to ensure optimum anchorage in the bone. Then the drill is extracted. An absorbable tack is introduced via the guidewire and seated with the cannulated Suretac
driver (Figs 10 and 11). This procedure is repeated for the middle notch, with attention paid once again to good capsule tension superomedially. If the glenoid labrum is also detached superior to the midline of the glenoid cavity, repair is performed by working intra-
/ J7
/z/ , i/
FIG 6. Depressing the capsule into the joint with the blunt end of the trocar and perforation with the K-wire.
EXTRA-ARTICULAR B A N K A R T REPAIR
193
A
.
/"
FIG 7. Repositioning of the speared capsule on the inferiormost notch (5 o'clock) and assessment of craniomedial capsule tension. Where tension is excessive or inadequate, a new perforation point is selected.
articularly at this level. G i v e n the undulating shape o f the glenoid rim, p l a c e m e n t o f the implants within the notches tightens the capsule b e t w e e n t h e m so that a watertight seal is obtained (Figs 12 and 13). Postoperative Management: Postoperatively, the p a t i e n t ' s arm is i m m o b i l i z e d in a light shoulder bandage for 3 weeks. F r o m w e e k s 4 to 6, up to 90 ° of flexion and external rotation to neutral are permitted. A f t e r the sixth week, exercise is p e r m i t t e d in all planes, including external rotation. Sports activities m a y be r e s u m e d f r o m the fourth m o n t h p o s t o p e r a t i v e l y except for those involving throwing actions, which are not a l l o w e d until the fifth month.
RESULTS O v e r a l l R e s u l t s : On the basis o f the R o w e score, 18 74% o f the results were classified as very good, 9%
FIG 8. Placement of the tip of the guidewire and drill. (A) Placement of the guidewire in the middle of the inferiormost notch under arthroscopic control. (B) Arthroscopic view of placement of the guidewire in the middle of the notch. C, cartilage; T, trough, CO, cortex.
J
194
H. RESCH E T AL.
/ /;
FIG 11. Extra-articular (extracapsular) location of the tack. The capsule is tightened (arm in 30 ° external rotation!).
i
F I G 9. Drilling in the guidewire and drill as far as the stop. The locking mechanism is released and the guidewire hammered home.
as good, 8% fair, and 9% poor. Discounting the first 30 shoulders to account for the development of the technique, and evaluating only the subsequent 70 shoulders, the overall results graded on the Rowe system were 76% very good, 12% good, 6% fair, and 6% poor. Stability: At the time of follow-up, 9 patients (9 shoulders) suffered a recurrence (9% recurrence rate), with 7 dislocations and 2 subluxations. Again, discounting the first 30 shoulders which had a recurrence rate of almost 17% to take account of the developmental phase of the technique, the recurrence rate for the subsequent 70 shoulders was only 5.7%. In 7 of
/
,/ // i
the 9 patients who suffered a recurrence, trauma of varying severity was the causal factor. Glenoid Rim Pathology: Intraoperative findings included varying degrees of displacement or detachment of the labrum from the glenoid rim in 72 shoulders. In 25 cases, the capsulolabral complex had formed a deep pocket following detachment from the glenoid rim and scapular neck. In three shoulders, no detachment of the labrum was observed (Fig 14). Changes to the labrum were restricted to the anterior inferior half of the glenoid in 42 shoulders (type A), while they affected the entire anterior glenoid rim from
\
i )
FIG 10. After removal of the drill with the help of a grip, an absorbable tack is inserted and hammered home.
FIG 12. Capsule refixed on the glenoid rim. The two lower implants are always extra-articular, the upper implant often intra-articular. Placement of the implants in the notches tensions the capsule between the notches, creating a watertight seal.
EXTRA-ARTICULAR BANKART REPAIR
195
shoulders 80 ,0
~"' ~ !!:~I~i
2 0
t detatchment from rim
FIG 14.
FIG 13. Radiographic view with implant position (Suretacs) shown as tunnels.
the 12 o'clock to the 6 o'clock position in 40 shoulders (type B) and extended beyond the 12 to 6 o'clock zone in 15 shoulders (type C). In 6 shoulders (6%), an additional type II SLAP lesion 3° was noted. Patients with no or only minor Bankart lesions (type A) and 3 patients without labral detachment had a recurrence rate of 13%, whereas patients with extensive Bankart lesions (types B and C) had a recurrence rate of 5% (Fig 15). The difference is statistically significant. In a total of 11 shoulders, a small glenoid rim avulsion fracture was noted preoperatively either on the computed tomographic arthrogram or the plain radiographs. Patients with larger defects underwent reconstruction using open techniques. The recurrence rate for these patients was 9%, which is identical to the overall recurrence rate. N u m b e r a n d T y p e of I m p l a n t s : A single implant was used in 11 shoulders, two in 52 shoulders, three in 34 shoulders, and four in 3 shoulders (Fig 16). The recurrence rate for shoulders with one implant was
detatchment
no detatchment
from rim and neck
Aspect of labrum lesion in 100 shoulders.
about 18% compared with 2.9% for three implants. The difference was statistically significant. In 25 shoulders, screws alone were used and in 53 shoulders, Suretac anchors were used. A combination of the two was used in 22 shoulders. The recurrence rate for screws alone was 8% compared with 9.4% for the absorbable tacks. The difference was not significant. Laxity: Of the 8 shoulders displaying laxity at preoperative examination (positive sulcus sign), recurrent instability occurred in 3 shoulders (37%), whereas the recurrence rate for all other shoulders was 6.5%. The difference was statistically significant. Hill-Sachs Lesion: A Hill-Sachs lesion of variable size was noted in 92 shoulders. In 15 shoulders, the lesion was only chondral or subchondral. In 59 shoulders, it was less than 5-ram deep whereas in 8 shoulders it was more than 5 m m deep (Fig 17). The recurrence rate for patients with chondral or subchondral Hill-
s h : : l ~
iA = =
[]
i
N,~
35
30 J % ~
A FIG 18.
B
C
inf. half ant. rim ant. rim ant. rim + posterior no detatchment
I
D
Extent of detachment of labrum in 100 shoulders.
196
H. RESCH E T AL.
shoulders
patients
6o
~-~
60 50
40
~i~i::i,i;;i:
40 30
20 10
I
:I:,N
~,,,~,o
10
N 1 F I G 16.
2
4
3
Implants
N u m b e r of implants in 100 shoulders.
Sachs lesions was 20%, and for lesions with a depth of more than 5 m m it was 11%. The difference was not significant. N u m b e r of Dislocations: The number of dislocations suffered preoperatively was below ten in 78 shoulders and above 10 in 22 shoulders. In the former group the recurrence rate was 7.6% compared with 13.6% for the second group. The difference was not significant. Mobility: Average limitation of external rotation measured at 0 ° abduction compared with the contralateral shoulder was 6.7 ° and at 90 ° abduction it was 6.1 °. In 6 shoulders, limitation of flexion and in 8 shoulders limitation of abduction by more than 10 ° was noted compared with the contralateral shoulder. Internal rotation was unrestricted in all cases. R e t u r n to Sports: All patients were asked before and after their operation to indicate their sports inter-
shoulders 60
40
~:#i,
I
subchondr. F I G 17.
< 5mm
> 5mm
I
none
Hill-Sachs lesion in 100 shoulders.
0
Overhead
demand, sparing collision no sport
FIG 18. Return to sports activity groups for 98 patients.
ests. Only one patient claimed not to have any athletic pursuits preoperatively. With the exception of this one case, all patients indicated that they returned to some degree of athletic participation after the operation, albeit not always the sport played preoperatively. Of a total of 49 patients who played a form of overhead sport preoperatively, 30 returned to that type of activity postoperatively. This represents a return rate of 61%. Of the 44 patients who played a shoulder-demanding form of sport preoperatively, 37 (84%) returned to their sport. Overall, there was an increase of 16 patients within this group if one includes all those who either changed from one group of sports to the other or changed their sport within the group (35% increase). This increase derives primarily from a switch from the overhead group to the shoulder-demanding group. All four patients who said they had participated in a shoulder-sparing type of sport preoperatively returned to that activity (100% return rate). This group was joined postoperatively by a further three patients from the shoulder-demanding group, making a total of seven patients. Before the operation, 20 patients were involved in some kind of contact sport, 14 of whom (70%) returned to their sport postoperatively (Fig 18). The figures for returning to the individual sports are shown in Fig 19. Figure 20 shows the case of a change of sport taking the preoperative tennis players as an example. Regaining premorbid levels of performance in sport was found to occur at two peaks on the time axis. The first peak was seen at about 6 months postoperatively (14%) with the second ocurring at 12 months (21°7o) (Fig 21). Strength: A spring balance was used to measure internal rotation force from the fully externally rotated
EXTRA-ARTICULAR BANKART REPAIR
197
patie~ 45 40 35 30 25 20 15 10 5
0
ski
tennis
soccer climbing snow- volleyboard ball
2
4
6
8
10
12
20
mos.
FIG 21. Time of return to former performancelevel in sports.
FIG19. Return to the individual sports activities.
position in comparison with the contralateral shoulder. The patient's arm was adducted and flexed to form a right angle at the elbow. The difference between the two sides was 0.53 kg on average and was not statistically significant. Analysis of Postoperative Recurrences: Of a total of nine shoulders that suffered a recurrence postoperatively, repair of the capsule had been performed with just one implant in two shoulders (representing a recurrence rate of 18% for all shoulders reconstructed with just one implant). In both cases, a minor Bankart lesion (type A) had been noted, which was the reason for using one implant None. Three shoulders that suffered a recurrence had been described as lax preoperatively (37% of all such shoulders). These three shoulders include the above two shoulders with minor Bankart lesions in which just one implant had been used. The third shoulder presented with a medium-size Bankart lesion (type B) and was reconstructed with two implants. Seven of the nine patients who suffered a recurrence did so while playing some form of sport. At the time of the recurrence, three patients were active in an overhead sport (1 handball, 1 tennis, 1 rock-climbing)
FIG 20. Changefrom tennis to other sports activities after surgery in 11 patients.
and four patients in a shoulder-demanding sport (2 snowboarding, 1 skiing, 1 football). An additional two patients sustained a recurrence while participating in a contact sport. Relative to the total number of patients in the various sports groups, the recurrence rate was 10% for the overhead group, 6.6% for the shoulderdemanding group, 0% for the shoulder-sparing group, and 14.2% for the group involved in contact sports. With the exception of the shoulder-sparing group, the differences were not significant. C h a n g e in Technique: In six of the cases in this study, the original goal of AEBR had to be abandoned intraoperatively and an open procedure substituted. In three cases with a clinically lax shoulder, no Bankart lesions and a large, thin joint capsule was found necessitating an open capsular shift procedure. In three other cases, the joint capsule was found intraoperatively to be very thin and weak (plastic deformation) in the area of the inferior glenohumeral ligament, and open Bankart repair was performed. Complications: A distinction must be made between complications involving screws and those involving the absorbable tacks. In four patients, the screw was placed too close to the cartilage with the result that the washer penetrated the capsule, causing metal impingement on the humeral head requiring removal. No screws fractured but one did eventually loosen. In the case of the absorbable tacks, the head penetrated the capsule causing button holing in six cases, and an additional tack had to be inserted. In four cases of reconstruction using the absorbable tacks, a slight increase in body temperature was noted between weeks 3 and 5 postoperatively, along with an increase in C-reactive protein accompanied by a feeling of general fatigue. These symptoms disappeared within 3 days in three of the patients, whereas an arthroscopic
198
H. RESCH ET AL.
lavage was required in the fourth patient, after which the symptoms immediately abated. We believe this may have represented a foreign body reaction to the synthetic material. DISCUSSION The open Bankart technique offers stable repair of the capsulolabral complex to the glenoid rim over the whole area of detachment using transosseous sutures or suture anchors. 16-2°'31 At the same time, a capsular shift can be performed to shorten the overstreched joint capsule. At arthroscopy, the two key elements, namely insertion of the implants in the region of the inferior third of the glenoid rim and execution of a capsular shift superomedially, can be satisfied by selecting the transsubscapular approach with extracapsular placement of the implants. As noted in the literature, subcoracoid portals are located in a potential hazard zone for the neurovascular structures) 2'23 With the help of anatomic studies, however, it has been possible to minimize the risks posed by this approach providing certain criteria are observed. 24 Unlike the procedure adopted by Davidson et al., who describe a similar transmuscular portal, 23 the slalom technique as described above offers an additional degree of safety, because the conjoined tendon is passed laterally. With a skin incision made at a maximum of 2 cm inferior to the coracoid process, the path of the blunt trocar lies not only superior to the musculocutaneous nerve but also lateral to it by the width of the conjoined tendon. As the anatomic studies show, there can be no risk to the neurovascular structures as long as the blunt trocar slides horizontally and maintains contact with the humeral head. In the anatomic studies performed, it was also shown that, on external rotation of the arm starting from the neutral position, the tendon of the subscapularis muscle lies lateral to the anterior glenoid rim. Therefore, as the blunt trocar glides along the humeral head, it does not pass through the tendon but easily penetrates the muscle tissue of the subscapularis. Movement of the humeral head beyond the glenoid rim overstretches the joint capsule and may cause plastic deformation of the capsuleY 27'32 Consequently, shifting the anterior capsule is just as important a part of arthroscopic stabilization as it is in open surgery. 16-18 The question of how much capsular shift is needed is best determined by shifting and tightening the capsule superomedially with the patient's arm maintained in an appropriate position. Thirty degrees of external rotation was selected for the position of the patient's arm
to duplicate the position the authors have used successfully during the open Bankart technique. 17 The right angle formed by the patient's arm in the elbow brace permits accurate adjustment to be made to arm rotation via the patient's forearm. There are a number of advantages to the notches burred in the glenoid rim. They permit the guidewire to be placed not only lateral to the cortex under arthroscopic control, thus avoiding the risk of deflection medially, but also a precise placement close to the cartilage to obtain a stepless transition between glenoid and capsule. In addition, placement of the implants in the notches also tensions the capsule between the notches, thus creating a watertight seal between the capsule and the glenoid rim, thereby preventing seepage of synovial fluid between the capsule and the rim over it' s entire length. Both the watertight seal and the smooth transition between glenoid and capsule help to regain the suction cup effect of the glenoid. 33 The screws used exclusively for fixation in the early phase led to complications in a number of cases in the form of metal impingement on the humeral head. For that reason, screws are not used anymore for that purpose. In the case of the absorbable tacks with the 6mm diameter head, the head repeatedly penetrated the capsule. The almost exclusive use over the last 2 years of the absorbable tacks with the bigger diameter head (Suretac II) has completely solved this problem. The systemic foreign body reaction observed in four patients was self-limited in three cases, disappearing after a few days. The arthroscopic irrigation performed on the fourth patient led to an immediate abatement of the symptoms. In general, the clinical course was similar to that reported by Edwards et al.34 The introduction of a new operating technique always involves a developmental phase and a learning curve. That explains the 17% recurrence rate in the first 30 shoulders. For the subsequent 70 shoulders, the rate was only 5.7%, which is comparable with the results achieved with open repair procedures. The minimum follow-up period of 18 months can be considered adequate for an assessment of the results as there is general agreement in the literature that recurrences peak between the 13th and 15th months postoperatively, and decrease strongly after the 18th month, 26'27although recurrences still can occur. 35 An analysis of the failures reveals a significantly large proportion of patients who were classified as " l a x " on preoperative clinical examination. As a result of the minor shearing forces that develop on release of the humeral head from the glenoid, most of these patients also had minor Bankart lesions.
EXTRA-ARTICULAR BANKART REPAIR
It was initially felt that a small Bankart lesion represented a good indication for arthroscopic repair. The smaller the Bankart lesion, the smaller were the n u m ber of implants. However, patients who u n d e r w e n t reconstruction using only one i m p l a n t suffered a significant increase in the n u m b e r of recurrences. Similar observations have been made by other authors. 4'26'27'36 Consequently, it must be concluded that the smaller the Bankart lesions, the more attention must be paid during surgery to the size and condition of the j o i n t capsule. In the case of very p r o n o u n c e d laxity, which is n o r m a l l y accompanied by a large and thin capsule, an open capsular shift is i n d i c a t e d ) 7 The 61% return rate for overhead sports is largely in agreement with the return rate quoted by W a r n e r and W a r r e n s and is well above the return rates after open surgery as reported by Rowe et el. 18 Torg et al., 38 and the authors/7 O n l y Jobe et al. 16 report a higher return rate on the basis of their open repair technique. A limitation on external rotation of only 6 ° is the key factor in the high return rate noted for overhead sports. The arm position selected for repair of the capsule as described above permits the correct degree of capsular shift to be performed. A l m o s t all patients who gave up an overhead sport postoperatively, took up a sport from the s h o u l d e r - d e m a n d i n g group instead, so that this group grew by essentially the same amount. According to our results, the risk of suffering a recurrence while participating in a contact sport was only slightly higher than for other activities. The only significant difference was in comparison with shoulder-sparing sports, which proves that a system for classifying sports on the basis of the loads imposed on the glenohumeral j o i n t is urgently required,
REFERENCES 1. Benedetto KP, G1/StzerW. Arthroscopic Bankart procedure by suture fixation technique: Indications, technique, and results. Arthroscopy 1992;8:111-115. 2. Caspari RB. Arthroscopic reconstruction for anterior shoulder instability. Tech Orthop 1988;3:59-66. 3. Johnson LL. Diagnostic and surgical arthroscopy of the shoulder. St Louis: Mosby, 1993;276-352. 4. Landsiedl F. Arthroscopic therapy of recurrent anterior luxation of the shoulder by capsular repair. Arthroscopy 1992;8:296-304. 5. Maki NJ. Arthroscopic stabilizationfor recurrent shoulder instability. Orthop Trans 1989;13:508. 6. Morgan CD, Bodenstab AB. Arthroscopic Bankart suture repair: Techniques and early results. Arthroscopy 1987;3:111-122. 7. Resch H, Golser K, Kathrein A. Arthroscopic screw fixation techniques. In: Resch H, Beck E, eds. Arthroscopy of the shoulder. New York: Springer-Vertag, 1992;99-112. 8. Warner JP, Warren RF. Arthroscopic Bankart repair using a cannulated, absorbable, fixation device. Oper Tech Orthop Surg 1991; 1:192-198.
199
9. Wiley AM. Arthroscopyfor shoulder instability and a technique for arthroscopic repair. Arthroscopy 1988;4:25-30. 10. Morgan CD. Arthroscopic transglenoid Bankart suture repair. Oper Tech Orthop 1991;1:171. 11. Wolf EM, Wilke RM, Richmond JC. Arthroscopic Bankart Repair using suture anchors. Oper Tech Orthop 1991;1:184. 12. Habermeyer P, Wiedmann E. Arthroscopic three-point Bankart suture repair. In: Resch H, Beck E, eds. Arthroscopy of the shoulder. New York: Springer-Verlag, 1992;90-98. 13. Hawkins RB. Arthroscopic stapling repair for shoulder instability: A retrospective study of 50 cases. Arthroscopy 1989;5:122128. 14. Weber S. A prospective evaluation comparing open and arthroscopic treatment of recurrent anterior glenohumeral dislocation. Presented at the Annual Meeting of the American Orthopaedic Society for Sports Medicine, New Orleans, LA, February 11, t990. 15. Hovelius L, Akermark C, Albrektsson B, Berg El K6rner L, Lundberg B, WredmarkT. Brislow-LetarjetProcedurefor recurrent anterior dislocation of the shoulder. Acta Orthop Scand 1983;54:284-290. 16. Jobe FW, Giangarra CE, Kvitne RS, Glousman RE. Anterior capsulo labral reconstructionof the shoulder in athletes in overhand sports. Am J Sports Med 1991;19:428-434. 17. Resch H. The Bankart Procedure: A Modificationusing Transosseous Sutures. Orthop Traumatol 1993;2:19-28. 18. Rowe CR, Patel D, ShouthmaydWW. The Bankart procedure-A long-term End-result study. J Bone Joint Surg Am 1978;60: 1-18. 19. Thomas SC, Matsen FA. An approach to the repair of avulsion of the glenohumeral ligaments in the management of traumatic anterior glenohumeral instability. J Bone Joint Surg Am 1989; 71:506-513. 20. Bankart ASB. The pathology and treatment of recurrent dislocation of the shoulder joint. Br J Surg 1938;26:23-29. 21. Davidson PA, Tibone JE. Anterior inferior (5 o'clock) portal for shoulder arthrscopy. Arthroscopy 1995;5:519-525. 22. Matthews LS, Vetter WL, Oweida SJ, Spearman J, Helfet DL. Arthroscopic staple capsulorrhaphyfor reccurent anterior shoulder instability. Arthroscopy 1988;4:106-111. 23. Ogilvie-Harris DJ, Wiley AM. Arthroscopic surgery of the shoulder. J Bone Joint Surg Br 1986;68:201-207. 24. Resch H, Wykypiel HF, Maurer H, Wambacher M. The anteroinferior (transmuscular)approach for arthroscopic repair of the Bankart lesion--An anatomical and clinical study. Arthroscopy 1996; 12:309-322. 25. Bigliani LU, Pollock RG, Soslowski LJ, Flatow EL, Pawluk RJ, Mow VC. Tensile properties of the inferior glenohumeral ligament. J Orthop Res 1992;10:187-197. 26. Grana WA, Buckley PD, Yates CK. Arthroscopic Bankart suture repair. Am J Sports Med 1993;21:348-353. 27. Walch G, Boileau P, Levigne C, Mandrino A, Neyret P, Donell S. Arthroscopic stabilization for recurrent anterior shoulder dislocation: Results of 59 cases. Arthroscopy 1995;11:173-179. 28. Bernageau J, Platte D, Debeyre J, Ferrane J. IntEret du profil gl6noidiendans les luxations r6ccidivantesde l'6paule. Rev Chit Orthop 1976;62:142-147 (suppl II). 29. Skyhar MJ, AltcheckDW, Warren RF, WiekecwiczTL, O'Brien SJ. Shoulder arthroscopy with the patient in the beach-chair position. Arthroscopy 1988;4:265-269. 30. Snyder SJ, Karzel RP, DelPizzo W, Ferkel RD, Friedman MJ. SLAP lesions of the shoulder. Arthroscopy 1990;6:274-279. 31. Richmond JC, Donaldson WR, Fu F, Harner C. Modificationof the Bankart reconstruction with a suture anchor. Am J Sports Med 1991;19:343-346. 32. Adolfsson L, Lysholm J. Arthroscopy and stability testing for anterior shoulder instability. Arthroscopy 1989;5:315-320. 33. Habermeyer P, Schuller U, Wiedemann E. The Intra-articular
200
H. R E S C H E T AL.
pressure of the shoulder: An experimental study on the role of the glenoid labrum in stabilizing the joint. Arthroscopy 1992; 8: 166-172. 34. Edwards DJ, Hoy G, Sales AD, Hayes MG. Adverse reactions to an absorbable shoulder fixation device. J Shoulder Elbow Surg 1994;3:230-233. 35. Morrey BF, Jones JM. Recurrent anterior dislocation of the shoulder. Long-term follow-up of the Putti-Platt and Bankart procedures. J Bone Joint Surg Am 1976;56:252. 36. Green MR, Christiansen KP. Arthroscopic versus open Bankart
procedure. A comparison of early morbidity and complications. Arthroscopy 1993;9:371-374. 37. Neer CS II, Foster CR. Inferior capsular shift for involontary inferior and multidirectional instability of the shoulder. J Bone Joint Surg Am 1980;62:897-908. 38. Torg JS, Balduini FC, Bonci C, Lehman RC, Gregg JR, Esterhai JL, Hensal FJ. A modified Bristow-Helfet-May Procedure for recurrent dislocation and subluxation of the shoulder: A report of 212 cases. J Bone Joint Surg Am 1987; 69:904-913.