- Email: [email protected]
Arthroscopic Bankart suture repair for traumatic anterior shoulder instability: analysis of the causes of a recurrence
Arthroscopic Bankart Suture Repair for Traumatic Anterior Shoulder Instability: Analysis of the Causes of a Recurrence Kenji Hayashida, M.D., Ph.D., Minoru Yoneda, M.D., Ph.D., Shigeto Nakagawa, M.D., Ph.D., Kenji Okamura, M.D., and Sunao Fukushima, M.D.
Summary: Eighty-two patients with traumatic anterior shoulder instability were treated with an arthroscopic transglenoid multiple suture technique (Caspari’s method) and followed-up for more than 2 years. A retrospective analysis of the clinical outcome was performed to determine the factors related to poor results. The mean age at operation was 21 years (range, 13 to 50 years) and the mean follow-up period was 40 months (range, 24 to 70 months). According to the status of the ligament-labrum complex and the glenoid bone defect, the Bankart lesions were classified into five types arthroscopically. There were 21 shoulders of type 1, 33 shoulders of type 2, 22 shoulders of type 3, and 6 shoulders of type 5. Twenty-four of the patients played contact sports before the operation. The clinical outcome was assessed by Rowe’s criteria (1978). To analyze the factors related to a poor outcome, a multivariate analysis was done to assess the influence of 12 clinical factors (age at operation, age at first dislocation, sex, dominant side, disease duration, number of dislocations, sporting activity before operation, inferior joint laxity, thickness of the ligament-labrum complex, type of Bankart lesion, number of sutures, and method of suture fixation). Fifty-five of 82 patients had an excellent outcome, 14 had a good result, and 13 had a poor result. According to postoperative instability, redislocation was seen in 13 patients (16%), resubluxation in 2 patients (2%), with a recurrence rate of 18%. The mean limitation of external rotation at 90° abduction was 6.0° (range, 0° to 30°), and there was a 10° loss of external rotation in 10 patients. The factors significantly related to recurrence were a type 3 Bankart lesion, playing contact sports preoperatively, a thin ligament-labrum complex, and repair with less than four sutures. In conclusion, a 18% rate of recurrence is not acceptable. To obtain a better clinical outcome, very careful selection of patients for this technique is necessary. Our analysis of the factors related to a poor outcome may help to decide what the proper indications are for this technique. Key Words: Arthroscopic Bankart suture repair—Anterior shoulder instability—Causes of a recurrence.
A
rthroscopic surgery of the shoulder allows the surgeon to treat patients who have glenohumeral instability with less morbidity than with an open
From the Department of Orthopaedic Surgery, Osaka KoseiNenkin Hospital, Osaka, Japan. An abstract of this article was presented at the 63rd Annual Meeting of the American Academy of Orthopaedic Surgeons, Atlanta, Georgia, February, 1996. Address correspondence and reprint requests to Kenji Hayashida, M.D., Ph.D., 540 Buckingham Rd, Apt 1113, Richardson, TX 75081, U.S.A. r 1998 by the Arthroscopy Association of North America 0749-8063/98/1403-1724$3.00/0
procedure. The early literature on arthroscopic Bankart suture repair for anterior glenohumeral instability was favorable.1-6 These reports encouraged many surgeons to adopt this technique, and we also started to treat patients with traumatic anterior shoulder instability using the arthroscopic multiple suture technique. However, less successful results have recently been reported in the literature, so that caution in the use of arthroscopic stabilization is recommended, and there have been many studies on the relationship between clinical features and the clinical outcome.7-11 The purposes of the present study are to report our clinical results obtained with this technique, to analyze the
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 14, No 3 (April), 1998: pp 295–301
295
296
K. HAYASHIDA ET AL.
causes of a poor outcome, and to determine the appropriate candidates for this operation. MATERIALS AND METHODS Patients From 1989 to 1994, 163 patients with traumatic anterior shoulder instability underwent surgery. Seventy-four patients who had large bone defects of the anteroinferior glenoid were treated by an open Bankart procedure augmented with Bristow’s method, and 89 patients without bone defects or with small bone defects were treated by the arthroscopic transglenoid multiple suture technique (Caspari’s method). Eightytwo of the 89 patients (92%) could be followed-up for more than 2 years and were reviewed retrospectively. All of them had surgery on one side. Sixty-three were men and 19 were women. The mean age at operation was 21 years (range, 13 to 50 years), and the mean follow-up period was 40 months (range, 24 to 70 months). Fifty-six of the 82 patients had experienced frank dislocation, and 26 patients had complained of instability; 79 showed a positive anterior apprehension sign in a crank test, and 3 only had pain with the shoulder in abduction and external rotation at a preoperative clinical examination. Fifty-two patients had surgery on the dominant side, and 30 on the nondominant side. The mean duration of shoulder problems was 35 months (range, 2 to 180 months), and the mean age of first dislocation was 18 years (range, 11 to 35 years). Regarding the number of dislocations, 2 patients had suffered only one dislocation, 29 patients had two to four dislocations, and 51 patients had more than four dislocations. Inferior laxity to the extent of half of the humeral head was seen in 20 shoulders, and laxity greater than half of a humeral head was seen in 2 shoulders, whereas the other shoulders had no significant inferior laxity. Before injury, 24 of 82 patients played contact sports (e.g., American football, rugby, judo) at the competition level in college or high school. Arthroscopic Bankart Classification We classified the Bankart lesions into 5 types based on the arthroscopic findings.12 Type 1 was labral detachment with a well-developed glenohumeral ligament (Fig 1). Type 2 was labral detachment with a poorly-developed glenohumeral ligament (Fig 2). Type 3 was a ligamentous tear with labral disruption (Fig 3). Type 4 was a ligamentous disruption with a bony defect of the glenoid (Fig 4). Type 5 was a slack glenohumeral ligament without a classic Bankart le-
FIGURE 1. Type 1: labral detachment, broad middle and anteriorinferior glenohumeral ligament, and no inferior foramen. BL, Bankart lesion; GF, glenoid fossa; HH, humeral head; LHB, long head of biceps tendon; SBS, subscapularis tendon; SGHL, superior glenohumeral ligament; MGHL, middle glenohumeral ligament; AIGHL, anterior inferior glenohumeral ligament.
sion (Fig 5). Twenty-one shoulders were classified as Type 1, 33 shoulders were Type 2, 22 shoulders were Type 3, and 6 shoulders were Type 5. There were no Type 4 shoulders in this study because a Type 4 Bankart lesion was outside our indications for arthroscopic repair. Operative Technique Under general anesthesia, the patient was placed in the lateral decubitus position. With the arm in 60° of
ARTHROSCOPIC BANKART SUTURE REPAIR
297
ligament/labrum complex by the suture punch technique. The mean number of sutures used was 7 (range, 4 to 9 sutures). A drill pin was inserted posteriorly through the glenoid neck and the sutures in the IGHL/LC were passed posteriorly with the pin. Then each suture was manually adjusted until the tension was as equal as possible. Finally, the sutures were fixed to the fascia of the infraspinatus in 22 patients, or on the posterior bone surface of the scapula using a poly-L-lactic acid button in 60 patients.12 Thickness of the IGHL/LC The thickness of the IGHL/LC was measured arthroscopically in 70 patients who were classified into 3
FIGURE 2. Type 2: labral detachment, narrow and thin middle glenohumeral ligament, narrow anterior-inferior glenohumeral ligament, and inferior foramen.
abduction and 20° of forward flexion, 3 to 4 kg of skin traction was applied. We used both an anterior portal and a posterior portal for Caspari’s technique.2 After classification of the Bankart lesion, debridement was performed and the lesion was released and mobilized. Then we abraded the anterior surface of the glenoid neck. Before suturing the labrum, we measured the thickness of the inferior glenohumeral ligament/ labrum complex (IGHL/LC) at 4 o’clock with a probe that had a 5-mm long hook (Fig 6). Multiple 1-0 or 2-0 PDS II sutures were inserted into the glenohumeral
FIGURE 3. Type 3: ligamentous tear with labral disruption, shrunk inferior glenohumeral ligament and wide inferior foramen.
298
K. HAYASHIDA ET AL. RESULTS Clinical Outcome The clinical outcome was excellent in 55 patients (67%), good in 14 patients (17%), and poor in 13 patients (16%), according to Rowe’s rating scale (1978).14 Regarding postoperative instability, all 13 patients (16%) with poor results had redislocation. Two patients who were classified as good on Rowe’s scale had no apparent redislocation, but felt mild discomfort. In the clinical examination at the followup, a half of humeral head shift in the abduction inferior stability test was recognized in these 2 patients, and they were classified as subluxation cases.
FIGURE 4. Type 4: severe ligamentous disruption and a large bony defect of the glenoid.
groups: ⬍3-mm thick (n ⫽ 16), 3- to 5-mm thick (n ⫽ 36), and ⬎5-mm thick (n ⫽ 18). Postoperative Regimen The arm was immobilized in a Velpeau bandage for 3 weeks and then in a sling for another 3 weeks. Exercises to extend the range of motion were started 3 weeks after operation. A complete return to normal daily activities was allowed 3 months after surgery, with ordinary sporting activity being allowed at 6 months and contact sport at 9 months. Statistical Analysis Multivariate analysis was used to investigate the factors related to a poor clinical outcome.
FIGURE 5. Type 5: slack glenohumeral ligament without a classic Bankart lesion.
ARTHROSCOPIC BANKART SUTURE REPAIR
299
TABLE 2. Bankart Classification Classification
Poor Outcome
Type 1 Type 2 Type 3 Type 5
FIGURE 6. The thickness of the IGHL/LC at 4 o’clock was measured with a probe that had a 5-mm long hook.
Thus, the recurrence rate of this study was 18% (15 of 82 patients). In 90° abduction, the mean restriction of external rotation was 6.0° (0° to 30°), and restriction of more than 10° was seen in 10 shoulders (12%). Transient damage to the infraspinatus branch of the suprascapular nerve was detected in 2 patients postoperatively, but no other complications were recognized. Analysis of Factors Related to Poor Results Fifteen patients with redislocation and resubluxation were defined as the patients with poor clinical outcome, and multivariate analysis was used to assess 12 clinical parameters (age at operation, sex, dominant side, duration of disease, number of dislocations, age at first dislocation, preinjury sporting activity, inferior joint laxity, type of Bankart lesion, thickness of the IGHL/LC, number of sutures, and method of suture fixation) to investigate the factors related to a poor outcome. The items that were significantly related to a poor result were the preinjury sporting activity, the type of Bankart lesion, the thickness of the IGHL/LC, and the number of sutures (P ⬍ .05). Regarding preinjury sporting activity (Table 1), the percentage of poor results in the patients who were engaged in contact sports (29%) was higher than in the patients who did not play contact sports (14%). According to the Bankart classification (Table 2), the percentage of poor results in the patients with type
2/21 shoulders (9%) 7/33 shoulders (21%) 6/22 shoulders (27%) 0/6 shoulders (0%)
1 (9%) and type 5 (0%) lesions was low, whereas it was high (27%) in the patients with type 3 lesions. The patients with an IGHL/LC more than 5 mm thick at 4 o’clock had a good outcome, whereas the patients with an IGHL/LC less than 3 mm thick tended to have poor result (31%) (Table 3). The patients with 5 sutures or fewer had more chance of redislocation (36%) than those with 6 or more sutures (15%) (Table 4). DISCUSSION According to our clinical results, Caspari’s technique can be expected to achieve a good range of motion postoperatively, but an 18% rate of recurrence (redislocation and resubluxation) is not acceptable. There have been many reports on the clinical outcome after arthroscopic suture repair of Bankart lesions. The early reports were hopeful, because the success rate was more than 90%.1-6 However, recent authors have reported a high rate of poor results (nearly 50%),7,8,10 although our data were not so bad. On the basis of our experience of arthroscopic Bankart repair with metal staples, we thought that patients with a bony defect of glenoid should be excluded from this type of Bankart repair, so the patients in this series had no large bony defects. This may be one of the reasons that we achieved better results than those recently reported by others. The second reason for our good outcome was the use of the Caspari method. With the Caspari method, it is easier to advance the IGHL superiorly compared with other arthroscopic Bankart repair techniques, and we could make a good capsular shift in this series. But the outcome was still not good compared with the clinical results of open Bankart repair.13-16 It has been emphasized by many authors that correct patient selection is important to improve the results of TABLE 3. Thickness of the Anterior IGHL (n ⫽ 70)
TABLE 1. Preinjury Sporting Activity Thickness Activity
Poor Outcome
Contact sports Noncontact sports
7/24 shoulders (29%) 8/58 shoulders (14%)
⬍3 mm 3-5 mm ⬎5 mm
Poor Outcome 5/16 shoulders (31%) 5/36 shoulders (14%) 1/18 shoulders (5%)
300
K. HAYASHIDA ET AL. TABLE 4. Number of Sutures
Number
Poor Outcome
ⱕ5 sutures ⱖ6 sutures
4/11 shoulders (36%) 11/71 shoulders (15%)
arthroscopic Bankart repair. To obtain proper selection criteria for Caspari’s technique, we investigated the causes of a poor outcome using multivariate analysis. The age at operation, sex, dominant side, duration of disease, number of dislocations, age at first dislocation, inferior laxity, and method of suture fixation did not influence the clinical results. Patients who played contact sports had a high rate of recurrence. This was probably because almost all of them wanted to return to contact sports again and thus had a high risk of reinjury. In this study, 24 patients played contact sports before injury, and 20 of 24 tried to return the same contact sports. There were 7 recurrences in this group, and 5 of them had redislocation during the contact sporting activity. Accordingly, they do not seem to be such good candidates for the arthroscopic Bankart repair. Patients with type 1 Bankart lesions were expected to have a good outcome, whereas patients with type 3 Bankart lesions were anticipated to have a poor result. In type 1, there is a well-developed glenohumeral ligament and the ligament/labrum complex is preserved, so we can expect a good anterior buttressing effect after Bankart repair. On the other hand, the type 3 lesion is associated with a poor glenohumeral ligament with labral disruption, so the anterior buttressing effect will be weak. These factors might explain the good results of type 1 lesions and the poor results of type 3 lesions. In this study, we also achieved good results in the patients with type 5 Bankart lesions. In type 5, there is no classical Bankart lesion and patients usually have a thin, stretched IGHL/LC, so poor results might be expected. Recently, Torchia, et al.17 reported the long-term clinical outcome of the Caspari method, and stated that the patients without a Bankart lesion could expect a better outcome than those with a discreet Bankart lesion. They explained that preservation of the labrum and a precise capsular shortening were cause of success. In all our type 5 cases, the labrums were preserved in good condition, and more capsular shift might be made to get good capsular tension, which might be one of the reasons for good outcome. The thickness of the IGHL/LC at 4 o’clock was significantly related to the clinical results. In the Bankart operation, refreshing the bone surface and
positioning the IGHL/LC at a good location are the first steps of the repair process. To complete a Bankart repair, it is important to obtain firm attachment of the IGHL/LC to the glenoid with enough strength to prevent the humeral head escaping from the glenoid anteriorly. In all of our recurrent cases, retear at the repair site was recognized at the time of second operation or on double-contrast computed tomographic arthrography, indicating that the cause of redislocation was insufficient healing of the repair site. A thick IGHL/LC has two advantages. One is a wide contact area facing the glenoid, which promotes good healing, and the other is a good suture anchorage from which it is difficult to tear out the threads. Therefore, the patients with a thick IGHL/LC had a good result. The number of sutures also had a significant relationship with the outcome. Maintaining contact between the IGHL/LC and the anterior surface of the glenoid until tissue healing is important for the success of this operation, so it seems to be reasonable to use more sutures to get a good repair. Based on our results, patients who do not play contact sports, who have a type 1 Bankart lesion and a thick IGHL/LC, and who undergo repair with numerous sutures may be expected to achieve a good result from the arthroscopic Bankart repair and are good candidates for this operation. On the other hand, patients who play contact sports, who have a type 3 Bankart lesion and a thin IGHL/LC, and who undergo repair with only a few sutures may anticipate a poor result, and should be treated using other procedures. According to other recent reports, patients with a bony Bankart lesion,7 who play contact sports,8,9 who have postoperative immobilization for less than 6 weeks,10 or whose labrum is stabilized with a small number of sutures11 are likely to have a poor result. These reports are compatible with our findings, and these analyses seem to clarify the good candidates for the arthroscopic Bankart repair using the Caspari technique. To obtain better results, Caspari’s method was also modified recently. Marcacci et al.18 changed the suture site from the fascia to the spine of the scapula and achieved a stable fixation of sutures with good clinical results. We have also modified the method of suture fixation by using a biodegradable poly-L-lactic acid button,12 and used the new anchoring system in 60 patients in this study. But the percentage of poor outcome of the new system was 18% (11 of 60) and that of the original Caspari method was also 18% (4 of 22), so we could not find any efficacy in changing the anchoring system in this study.
ARTHROSCOPIC BANKART SUTURE REPAIR Initial reports on arthroscopic Bankart repair were encouraging, but recent results have been disappointing. However, many patients have been treated with this technique and those with a good clinical outcome do not feel apprehension, have little limitation of motion, and are satisfied with the small scar. Thus, we should not abolish this technique but should instead make efforts to establish the proper indications, and to modify the operation to obtain better clinical results comparable with those achieved using open methods. REFERENCES 1. Morgan CD, Bodenstab AB. Arthroscopic Bankart suture repair: Technique and early results. Arthroscopy 1987;3:111122. 2. Caspari RB. Arthroscopic reconstruction for shoulder instability. Tech Orthop 1988;3:59-66. 3. Benedetto KP, Glotzer W. Arthroscopic Bankart procedure by suture technique: Indications, technique, and results. Arthroscopy 1992;8:111-115. 4. Goldberg BJ, Nirschl RP, McConnel JP, Pettrone FA. Arthroscopic transglenoid suture capsulolabral repairs: Preliminary results. Am J Sports Med 1993;21:656-665. 5. Green MR, Christensen KP. Arthroscopic versus open Bankart procedures: A comparison of early morbidity and complications. Arthroscopy 1993;9:371-374. 6. Tauro JC, Cater FM. Arthroscopic capsular advancement for anterior and anterior-inferior shoulder instability: A preliminary report. Arthroscopy 1994;10:513-517.
301
7. Walch G, Boileau P, Levigne C, Mandrino A, Neyret P, Donell S. Arthroscopic stabilization for recurrent anterior shoulder dislocation: Results of cases. Arthroscopy 1995;11:173-179. 8. Grana WA, Buckley PD, Yates CK. Arthroscopic Bankart suture repair. Am J Sports Med 1993;21:348-352. 9. Youssef JA, Carr CF, Walther CE, Murphy JM. Arthroscopic Bankart suture repair for recurrent traumatic unidirectional anterior shoulder dislocation. Arthroscopy 1995;11:561-563. 10. Mologne TS, Lapoint JM, Morin WD, Zilberfarb J, O’Brien JT. Arthroscopic anterior labral reconstruction using a transglenoid suture technique. Am J Sport Med 1996;24:268-274. 11. Landsiedl F. Arthroscopic therapy of recurrent anterior luxation of the shoulder by capsular repair. Arthroscopy 1992;8:296304. 12. Yoneda M, Hayashida K, Izawa K, Shimada K, Shino K. A simple and secure anchoring system for Caspari’s transglenoid multiple suture technique using a biodegradable poly-L-lactic acid button. Arthroscopy 1996;12:293-299. 13. Rowe CR, Zarins B. Recurrent transient subluxation of the shoulder. J Bone Joint Surg Am 1981;63:863-872. 14. Bankart ASB. Recurrent or habitual dislocation of the shoulder joint. BMJ 1923;2:1132-1133. 15. Rowe CR, Zarins B. Recurrent anterior dislocations after surgical repair. J Bone Joint Surg Am 1984;66:159-168. 16. Matsen FA, Thomas SC, Rockwood FA III, eds. The shoulder. Ed 1, Vol 1. Philadelphia: WB Saunders, 1990:547-551. 17. Torchia ME, Caspari RB, Assemlmeier MA, Beach WR, Gayari M. Arthroscopic transglenoid multiple suture repair: 2 to 8 year results in 150 shoulders. Arthroscopy 1997;13:609619. 18. Marcacci M, Zaffagnini S, Petitto A, Neri MP, Iacono F, Visani A. Arthroscopic management of recurrent anterior dislocation of the shoulder: Analysis of technical modifications on the Caspari procedure. Arthroscopy 1996;12:144-149.
Summary: Eighty-two patients with traumatic anterior shoulder instability were treated with an arthroscopic transglenoid multiple suture technique (Caspari’s method) and followed-up for more than 2 years. A retrospective analysis of the clinical outcome was performed to determine the factors related to poor results. The mean age at operation was 21 years (range, 13 to 50 years) and the mean follow-up period was 40 months (range, 24 to 70 months). According to the status of the ligament-labrum complex and the glenoid bone defect, the Bankart lesions were classified into five types arthroscopically. There were 21 shoulders of type 1, 33 shoulders of type 2, 22 shoulders of type 3, and 6 shoulders of type 5. Twenty-four of the patients played contact sports before the operation. The clinical outcome was assessed by Rowe’s criteria (1978). To analyze the factors related to a poor outcome, a multivariate analysis was done to assess the influence of 12 clinical factors (age at operation, age at first dislocation, sex, dominant side, disease duration, number of dislocations, sporting activity before operation, inferior joint laxity, thickness of the ligament-labrum complex, type of Bankart lesion, number of sutures, and method of suture fixation). Fifty-five of 82 patients had an excellent outcome, 14 had a good result, and 13 had a poor result. According to postoperative instability, redislocation was seen in 13 patients (16%), resubluxation in 2 patients (2%), with a recurrence rate of 18%. The mean limitation of external rotation at 90° abduction was 6.0° (range, 0° to 30°), and there was a 10° loss of external rotation in 10 patients. The factors significantly related to recurrence were a type 3 Bankart lesion, playing contact sports preoperatively, a thin ligament-labrum complex, and repair with less than four sutures. In conclusion, a 18% rate of recurrence is not acceptable. To obtain a better clinical outcome, very careful selection of patients for this technique is necessary. Our analysis of the factors related to a poor outcome may help to decide what the proper indications are for this technique. Key Words: Arthroscopic Bankart suture repair—Anterior shoulder instability—Causes of a recurrence.
A
rthroscopic surgery of the shoulder allows the surgeon to treat patients who have glenohumeral instability with less morbidity than with an open
From the Department of Orthopaedic Surgery, Osaka KoseiNenkin Hospital, Osaka, Japan. An abstract of this article was presented at the 63rd Annual Meeting of the American Academy of Orthopaedic Surgeons, Atlanta, Georgia, February, 1996. Address correspondence and reprint requests to Kenji Hayashida, M.D., Ph.D., 540 Buckingham Rd, Apt 1113, Richardson, TX 75081, U.S.A. r 1998 by the Arthroscopy Association of North America 0749-8063/98/1403-1724$3.00/0
procedure. The early literature on arthroscopic Bankart suture repair for anterior glenohumeral instability was favorable.1-6 These reports encouraged many surgeons to adopt this technique, and we also started to treat patients with traumatic anterior shoulder instability using the arthroscopic multiple suture technique. However, less successful results have recently been reported in the literature, so that caution in the use of arthroscopic stabilization is recommended, and there have been many studies on the relationship between clinical features and the clinical outcome.7-11 The purposes of the present study are to report our clinical results obtained with this technique, to analyze the
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 14, No 3 (April), 1998: pp 295–301
295
296
K. HAYASHIDA ET AL.
causes of a poor outcome, and to determine the appropriate candidates for this operation. MATERIALS AND METHODS Patients From 1989 to 1994, 163 patients with traumatic anterior shoulder instability underwent surgery. Seventy-four patients who had large bone defects of the anteroinferior glenoid were treated by an open Bankart procedure augmented with Bristow’s method, and 89 patients without bone defects or with small bone defects were treated by the arthroscopic transglenoid multiple suture technique (Caspari’s method). Eightytwo of the 89 patients (92%) could be followed-up for more than 2 years and were reviewed retrospectively. All of them had surgery on one side. Sixty-three were men and 19 were women. The mean age at operation was 21 years (range, 13 to 50 years), and the mean follow-up period was 40 months (range, 24 to 70 months). Fifty-six of the 82 patients had experienced frank dislocation, and 26 patients had complained of instability; 79 showed a positive anterior apprehension sign in a crank test, and 3 only had pain with the shoulder in abduction and external rotation at a preoperative clinical examination. Fifty-two patients had surgery on the dominant side, and 30 on the nondominant side. The mean duration of shoulder problems was 35 months (range, 2 to 180 months), and the mean age of first dislocation was 18 years (range, 11 to 35 years). Regarding the number of dislocations, 2 patients had suffered only one dislocation, 29 patients had two to four dislocations, and 51 patients had more than four dislocations. Inferior laxity to the extent of half of the humeral head was seen in 20 shoulders, and laxity greater than half of a humeral head was seen in 2 shoulders, whereas the other shoulders had no significant inferior laxity. Before injury, 24 of 82 patients played contact sports (e.g., American football, rugby, judo) at the competition level in college or high school. Arthroscopic Bankart Classification We classified the Bankart lesions into 5 types based on the arthroscopic findings.12 Type 1 was labral detachment with a well-developed glenohumeral ligament (Fig 1). Type 2 was labral detachment with a poorly-developed glenohumeral ligament (Fig 2). Type 3 was a ligamentous tear with labral disruption (Fig 3). Type 4 was a ligamentous disruption with a bony defect of the glenoid (Fig 4). Type 5 was a slack glenohumeral ligament without a classic Bankart le-
FIGURE 1. Type 1: labral detachment, broad middle and anteriorinferior glenohumeral ligament, and no inferior foramen. BL, Bankart lesion; GF, glenoid fossa; HH, humeral head; LHB, long head of biceps tendon; SBS, subscapularis tendon; SGHL, superior glenohumeral ligament; MGHL, middle glenohumeral ligament; AIGHL, anterior inferior glenohumeral ligament.
sion (Fig 5). Twenty-one shoulders were classified as Type 1, 33 shoulders were Type 2, 22 shoulders were Type 3, and 6 shoulders were Type 5. There were no Type 4 shoulders in this study because a Type 4 Bankart lesion was outside our indications for arthroscopic repair. Operative Technique Under general anesthesia, the patient was placed in the lateral decubitus position. With the arm in 60° of
ARTHROSCOPIC BANKART SUTURE REPAIR
297
ligament/labrum complex by the suture punch technique. The mean number of sutures used was 7 (range, 4 to 9 sutures). A drill pin was inserted posteriorly through the glenoid neck and the sutures in the IGHL/LC were passed posteriorly with the pin. Then each suture was manually adjusted until the tension was as equal as possible. Finally, the sutures were fixed to the fascia of the infraspinatus in 22 patients, or on the posterior bone surface of the scapula using a poly-L-lactic acid button in 60 patients.12 Thickness of the IGHL/LC The thickness of the IGHL/LC was measured arthroscopically in 70 patients who were classified into 3
FIGURE 2. Type 2: labral detachment, narrow and thin middle glenohumeral ligament, narrow anterior-inferior glenohumeral ligament, and inferior foramen.
abduction and 20° of forward flexion, 3 to 4 kg of skin traction was applied. We used both an anterior portal and a posterior portal for Caspari’s technique.2 After classification of the Bankart lesion, debridement was performed and the lesion was released and mobilized. Then we abraded the anterior surface of the glenoid neck. Before suturing the labrum, we measured the thickness of the inferior glenohumeral ligament/ labrum complex (IGHL/LC) at 4 o’clock with a probe that had a 5-mm long hook (Fig 6). Multiple 1-0 or 2-0 PDS II sutures were inserted into the glenohumeral
FIGURE 3. Type 3: ligamentous tear with labral disruption, shrunk inferior glenohumeral ligament and wide inferior foramen.
298
K. HAYASHIDA ET AL. RESULTS Clinical Outcome The clinical outcome was excellent in 55 patients (67%), good in 14 patients (17%), and poor in 13 patients (16%), according to Rowe’s rating scale (1978).14 Regarding postoperative instability, all 13 patients (16%) with poor results had redislocation. Two patients who were classified as good on Rowe’s scale had no apparent redislocation, but felt mild discomfort. In the clinical examination at the followup, a half of humeral head shift in the abduction inferior stability test was recognized in these 2 patients, and they were classified as subluxation cases.
FIGURE 4. Type 4: severe ligamentous disruption and a large bony defect of the glenoid.
groups: ⬍3-mm thick (n ⫽ 16), 3- to 5-mm thick (n ⫽ 36), and ⬎5-mm thick (n ⫽ 18). Postoperative Regimen The arm was immobilized in a Velpeau bandage for 3 weeks and then in a sling for another 3 weeks. Exercises to extend the range of motion were started 3 weeks after operation. A complete return to normal daily activities was allowed 3 months after surgery, with ordinary sporting activity being allowed at 6 months and contact sport at 9 months. Statistical Analysis Multivariate analysis was used to investigate the factors related to a poor clinical outcome.
FIGURE 5. Type 5: slack glenohumeral ligament without a classic Bankart lesion.
ARTHROSCOPIC BANKART SUTURE REPAIR
299
TABLE 2. Bankart Classification Classification
Poor Outcome
Type 1 Type 2 Type 3 Type 5
FIGURE 6. The thickness of the IGHL/LC at 4 o’clock was measured with a probe that had a 5-mm long hook.
Thus, the recurrence rate of this study was 18% (15 of 82 patients). In 90° abduction, the mean restriction of external rotation was 6.0° (0° to 30°), and restriction of more than 10° was seen in 10 shoulders (12%). Transient damage to the infraspinatus branch of the suprascapular nerve was detected in 2 patients postoperatively, but no other complications were recognized. Analysis of Factors Related to Poor Results Fifteen patients with redislocation and resubluxation were defined as the patients with poor clinical outcome, and multivariate analysis was used to assess 12 clinical parameters (age at operation, sex, dominant side, duration of disease, number of dislocations, age at first dislocation, preinjury sporting activity, inferior joint laxity, type of Bankart lesion, thickness of the IGHL/LC, number of sutures, and method of suture fixation) to investigate the factors related to a poor outcome. The items that were significantly related to a poor result were the preinjury sporting activity, the type of Bankart lesion, the thickness of the IGHL/LC, and the number of sutures (P ⬍ .05). Regarding preinjury sporting activity (Table 1), the percentage of poor results in the patients who were engaged in contact sports (29%) was higher than in the patients who did not play contact sports (14%). According to the Bankart classification (Table 2), the percentage of poor results in the patients with type
2/21 shoulders (9%) 7/33 shoulders (21%) 6/22 shoulders (27%) 0/6 shoulders (0%)
1 (9%) and type 5 (0%) lesions was low, whereas it was high (27%) in the patients with type 3 lesions. The patients with an IGHL/LC more than 5 mm thick at 4 o’clock had a good outcome, whereas the patients with an IGHL/LC less than 3 mm thick tended to have poor result (31%) (Table 3). The patients with 5 sutures or fewer had more chance of redislocation (36%) than those with 6 or more sutures (15%) (Table 4). DISCUSSION According to our clinical results, Caspari’s technique can be expected to achieve a good range of motion postoperatively, but an 18% rate of recurrence (redislocation and resubluxation) is not acceptable. There have been many reports on the clinical outcome after arthroscopic suture repair of Bankart lesions. The early reports were hopeful, because the success rate was more than 90%.1-6 However, recent authors have reported a high rate of poor results (nearly 50%),7,8,10 although our data were not so bad. On the basis of our experience of arthroscopic Bankart repair with metal staples, we thought that patients with a bony defect of glenoid should be excluded from this type of Bankart repair, so the patients in this series had no large bony defects. This may be one of the reasons that we achieved better results than those recently reported by others. The second reason for our good outcome was the use of the Caspari method. With the Caspari method, it is easier to advance the IGHL superiorly compared with other arthroscopic Bankart repair techniques, and we could make a good capsular shift in this series. But the outcome was still not good compared with the clinical results of open Bankart repair.13-16 It has been emphasized by many authors that correct patient selection is important to improve the results of TABLE 3. Thickness of the Anterior IGHL (n ⫽ 70)
TABLE 1. Preinjury Sporting Activity Thickness Activity
Poor Outcome
Contact sports Noncontact sports
7/24 shoulders (29%) 8/58 shoulders (14%)
⬍3 mm 3-5 mm ⬎5 mm
Poor Outcome 5/16 shoulders (31%) 5/36 shoulders (14%) 1/18 shoulders (5%)
300
K. HAYASHIDA ET AL. TABLE 4. Number of Sutures
Number
Poor Outcome
ⱕ5 sutures ⱖ6 sutures
4/11 shoulders (36%) 11/71 shoulders (15%)
arthroscopic Bankart repair. To obtain proper selection criteria for Caspari’s technique, we investigated the causes of a poor outcome using multivariate analysis. The age at operation, sex, dominant side, duration of disease, number of dislocations, age at first dislocation, inferior laxity, and method of suture fixation did not influence the clinical results. Patients who played contact sports had a high rate of recurrence. This was probably because almost all of them wanted to return to contact sports again and thus had a high risk of reinjury. In this study, 24 patients played contact sports before injury, and 20 of 24 tried to return the same contact sports. There were 7 recurrences in this group, and 5 of them had redislocation during the contact sporting activity. Accordingly, they do not seem to be such good candidates for the arthroscopic Bankart repair. Patients with type 1 Bankart lesions were expected to have a good outcome, whereas patients with type 3 Bankart lesions were anticipated to have a poor result. In type 1, there is a well-developed glenohumeral ligament and the ligament/labrum complex is preserved, so we can expect a good anterior buttressing effect after Bankart repair. On the other hand, the type 3 lesion is associated with a poor glenohumeral ligament with labral disruption, so the anterior buttressing effect will be weak. These factors might explain the good results of type 1 lesions and the poor results of type 3 lesions. In this study, we also achieved good results in the patients with type 5 Bankart lesions. In type 5, there is no classical Bankart lesion and patients usually have a thin, stretched IGHL/LC, so poor results might be expected. Recently, Torchia, et al.17 reported the long-term clinical outcome of the Caspari method, and stated that the patients without a Bankart lesion could expect a better outcome than those with a discreet Bankart lesion. They explained that preservation of the labrum and a precise capsular shortening were cause of success. In all our type 5 cases, the labrums were preserved in good condition, and more capsular shift might be made to get good capsular tension, which might be one of the reasons for good outcome. The thickness of the IGHL/LC at 4 o’clock was significantly related to the clinical results. In the Bankart operation, refreshing the bone surface and
positioning the IGHL/LC at a good location are the first steps of the repair process. To complete a Bankart repair, it is important to obtain firm attachment of the IGHL/LC to the glenoid with enough strength to prevent the humeral head escaping from the glenoid anteriorly. In all of our recurrent cases, retear at the repair site was recognized at the time of second operation or on double-contrast computed tomographic arthrography, indicating that the cause of redislocation was insufficient healing of the repair site. A thick IGHL/LC has two advantages. One is a wide contact area facing the glenoid, which promotes good healing, and the other is a good suture anchorage from which it is difficult to tear out the threads. Therefore, the patients with a thick IGHL/LC had a good result. The number of sutures also had a significant relationship with the outcome. Maintaining contact between the IGHL/LC and the anterior surface of the glenoid until tissue healing is important for the success of this operation, so it seems to be reasonable to use more sutures to get a good repair. Based on our results, patients who do not play contact sports, who have a type 1 Bankart lesion and a thick IGHL/LC, and who undergo repair with numerous sutures may be expected to achieve a good result from the arthroscopic Bankart repair and are good candidates for this operation. On the other hand, patients who play contact sports, who have a type 3 Bankart lesion and a thin IGHL/LC, and who undergo repair with only a few sutures may anticipate a poor result, and should be treated using other procedures. According to other recent reports, patients with a bony Bankart lesion,7 who play contact sports,8,9 who have postoperative immobilization for less than 6 weeks,10 or whose labrum is stabilized with a small number of sutures11 are likely to have a poor result. These reports are compatible with our findings, and these analyses seem to clarify the good candidates for the arthroscopic Bankart repair using the Caspari technique. To obtain better results, Caspari’s method was also modified recently. Marcacci et al.18 changed the suture site from the fascia to the spine of the scapula and achieved a stable fixation of sutures with good clinical results. We have also modified the method of suture fixation by using a biodegradable poly-L-lactic acid button,12 and used the new anchoring system in 60 patients in this study. But the percentage of poor outcome of the new system was 18% (11 of 60) and that of the original Caspari method was also 18% (4 of 22), so we could not find any efficacy in changing the anchoring system in this study.
ARTHROSCOPIC BANKART SUTURE REPAIR Initial reports on arthroscopic Bankart repair were encouraging, but recent results have been disappointing. However, many patients have been treated with this technique and those with a good clinical outcome do not feel apprehension, have little limitation of motion, and are satisfied with the small scar. Thus, we should not abolish this technique but should instead make efforts to establish the proper indications, and to modify the operation to obtain better clinical results comparable with those achieved using open methods. REFERENCES 1. Morgan CD, Bodenstab AB. Arthroscopic Bankart suture repair: Technique and early results. Arthroscopy 1987;3:111122. 2. Caspari RB. Arthroscopic reconstruction for shoulder instability. Tech Orthop 1988;3:59-66. 3. Benedetto KP, Glotzer W. Arthroscopic Bankart procedure by suture technique: Indications, technique, and results. Arthroscopy 1992;8:111-115. 4. Goldberg BJ, Nirschl RP, McConnel JP, Pettrone FA. Arthroscopic transglenoid suture capsulolabral repairs: Preliminary results. Am J Sports Med 1993;21:656-665. 5. Green MR, Christensen KP. Arthroscopic versus open Bankart procedures: A comparison of early morbidity and complications. Arthroscopy 1993;9:371-374. 6. Tauro JC, Cater FM. Arthroscopic capsular advancement for anterior and anterior-inferior shoulder instability: A preliminary report. Arthroscopy 1994;10:513-517.
301
7. Walch G, Boileau P, Levigne C, Mandrino A, Neyret P, Donell S. Arthroscopic stabilization for recurrent anterior shoulder dislocation: Results of cases. Arthroscopy 1995;11:173-179. 8. Grana WA, Buckley PD, Yates CK. Arthroscopic Bankart suture repair. Am J Sports Med 1993;21:348-352. 9. Youssef JA, Carr CF, Walther CE, Murphy JM. Arthroscopic Bankart suture repair for recurrent traumatic unidirectional anterior shoulder dislocation. Arthroscopy 1995;11:561-563. 10. Mologne TS, Lapoint JM, Morin WD, Zilberfarb J, O’Brien JT. Arthroscopic anterior labral reconstruction using a transglenoid suture technique. Am J Sport Med 1996;24:268-274. 11. Landsiedl F. Arthroscopic therapy of recurrent anterior luxation of the shoulder by capsular repair. Arthroscopy 1992;8:296304. 12. Yoneda M, Hayashida K, Izawa K, Shimada K, Shino K. A simple and secure anchoring system for Caspari’s transglenoid multiple suture technique using a biodegradable poly-L-lactic acid button. Arthroscopy 1996;12:293-299. 13. Rowe CR, Zarins B. Recurrent transient subluxation of the shoulder. J Bone Joint Surg Am 1981;63:863-872. 14. Bankart ASB. Recurrent or habitual dislocation of the shoulder joint. BMJ 1923;2:1132-1133. 15. Rowe CR, Zarins B. Recurrent anterior dislocations after surgical repair. J Bone Joint Surg Am 1984;66:159-168. 16. Matsen FA, Thomas SC, Rockwood FA III, eds. The shoulder. Ed 1, Vol 1. Philadelphia: WB Saunders, 1990:547-551. 17. Torchia ME, Caspari RB, Assemlmeier MA, Beach WR, Gayari M. Arthroscopic transglenoid multiple suture repair: 2 to 8 year results in 150 shoulders. Arthroscopy 1997;13:609619. 18. Marcacci M, Zaffagnini S, Petitto A, Neri MP, Iacono F, Visani A. Arthroscopic management of recurrent anterior dislocation of the shoulder: Analysis of technical modifications on the Caspari procedure. Arthroscopy 1996;12:144-149.