| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Musculoskeletal Imaging |
1 From the Department of Radiology, Showa University Fujigaoka Hospital, 1-30 Fujigaoka, Aoba-ku, Yokohama, Kanagawa-ken 227-8501, Japan (H.S., H.K.); and the Department of Orthopedics, Showa University Fujigaoka Rehabilitation Hospital, Japan (K.S., K.M., H.T.). Received June 28, 2001; revision requested August 16; revision received November 15; accepted January 7, 2002. Address correspondence to H.S. (e-mail: sugimo-h@sannet.ne.jp).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Thirty patients (25 men, five women; mean age, 28 years) who had undergone suture-anchor Bankart repair of one shoulder underwent MR arthrography before second-look arthroscopy. Ninety-eight anchors were used for the sutures. MR arthrographic diagnosis of anatomic reestablishment of the capsulolabral complex was correlated with arthroscopic findings. Contingency table analysis was performed to determine the relationship between MR arthrographic findings and arthroscopic findings.
RESULTS: MR findings of reattachment of the capsulolabral complex were in agreement with arthroscopic findings in 93 anchor points (accuracy, 93 of 98 anchor points; 95%). In 28 shoulders, oblique transverse images obtained with the shoulder in the abduction and external rotation position showed that the anterior band of the inferior glenohumeral ligament (AIGHL) abutted the humeral head and that reattachment of the AIGHL to the glenoid rim was seamless. Arthroscopy revealed satisfactory reestablishment of the capsulolabral complex in these shoulders. In the remaining two shoulders, a pool of contrast material was seen between the AIGHL and humeral head and a "divot" was detected at the point of reattachment of the AIGHL to the glenoid rim. Arthroscopy revealed unsatisfactory reestablishment of the capsulolabral complex. MR arthrographic findings of reattachment of the AIGHL were significantly associated with arthroscopic findings of reestablishment of the capsulolabral complex (P < .01).
CONCLUSION: MR arthrography can be reliably used for the postoperative assessment of suture-anchor Bankart repair.
© RSNA, 2002
Index terms: Arthroscopy • Shoulder, injuries, 414.4842 • Shoulder, MR, 414.12143 • Shoulder, surgery, 414.45
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Although the suture-anchor Bankart procedure yields relatively good results in the majority of patients (1,2), some experience instability or even dislocation after the repair and require a second operation. In current practice, a decision by an orthopedic surgeon to operate on a shoulder again is usually made on the basis of the patient’s clinical history and findings at physical examination and conventional radiography. Although magnetic resonance (MR) arthrography is gaining widespread use in this clinical setting (5), scientific reporting of MR arthrographic findings in patients treated with suture-anchor Bankart repair has been limited (6).
The purpose of this study was to determine the ability of MR arthrography to depict the anatomic reestablishment of the capsulolabral complex after suture-anchor Bankart repair.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Surgical Procedure
Unlike classic Bankart repair, which involves drilling holes in the margin of the glenoid rim, suture-anchor Bankart repair involves placement of anchors along the articular margin of the glenoid rim; the detached capsulolabral complex is sutured to these anchors (Fig 1) (2,4). In eight shoulders, titanium suture-anchors (Mitek GII; Ethicon, Somerville, NJ) were used. In the other 22 shoulders, bioabsorbable polylactic acid suture-anchors (Panalok 3.5; Ethicon) were used. A total of 98 anchors were used in the repair procedures: two in two shoulders, three in 18 shoulders, and four in 10 shoulders. The postoperative course of 29 patients was uneventful. One patient experienced instability of the shoulder after the repair. No patient had recurrence of shoulder subluxation or dislocation.
|
MR arthrography was performed at 1.0 T (Signa; GE Medical Systems, Milwaukee, Wis). The patients’ arms were in a neutral position. T1-weighted spin-echo fat-suppressed imaging was performed in the transverse, oblique coronal, and oblique sagittal planes with general-purpose dual coils wrapped around the affected shoulder. The following parameters were used in all imaging planes: repetition time msec/echo time msec, 440/9; two signals acquired; field of view, 16 cm; imaging matrix, 256 x 160; section thickness, 4 mm; and no intersection gap. Oblique transverse images were then obtained with the shoulders in the abduction and external rotation (ABER) position. The patients were instructed to assume the ABER position by placing the hand of the affected extremity posterior to the contralateral aspect of the head or neck with the elbow flexed (7). T1-weighted spin-echo fat-suppressed oblique transverse images were obtained. All MR imaging examinations were adequately performed.
One radiologist (H.S.) initially interpreted all the images and reported the findings before second-look arthroscopy was performed. Because the success of a Bankart repair is determined by assessing the degree of functional reestablishment of the capsulolabral complex (1–3), we assessed this degree in the following three ways.
First, we assessed whether the suture at each anchor point was intact. The presence of low-signal-intensity structures, which were considered to be the reattached capsulolabral complex, adjacent to the glenoid rim was evaluated at each anchor point on the oblique sagittal images. When a low-signal-intensity structure was seen abutting the glenoid rim, the glenoid rim was considered to be covered by the reattached capsulolabral complex. To enable the precise correlation of MR arthrographic findings with arthroscopic findings, the position of each anchor track was recorded according to the analogue clock notation system. Further, we assessed whether intraarticular contrast material had extravasated through the suture in each imaging plane.
Second, we assessed whether functional reestablishment of the capsulolabral complex was complete. Because investigations have shown that oblique transverse images obtained with the patient in the ABER position are most useful for evaluation of the anatomic integrity of the AIGHL (8–11), reattachment of the AIGHL to the glenoid rim was evaluated with these images.
Third, we assessed whether the anterior capsular attachment was related to the tightness of the anterior recess. Because the anterior capsular insertion seen on transverse images is reported to be related to joint stability (12), reattachment of the anterior capsule was also evaluated. The capsular reattachment was described in relation to its proximity to the glenoid rim according to a previously published classification system for anterior capsules (12). Type I capsular insertions occur directly adjacent to the glenoid labrum. Type II capsular insertions occur on the anterior glenoid rim in the space between the base of the labrum and the area 1 cm medial to the base of the labrum. Type III insertions occur in a location medial to that of a type II insertion.
Second-Look Arthroscopy
All patients underwent second-look arthroscopy for evaluation of their postoperative status. The arthroscopic procedure included the creation of a minor incision that allowed screws to be removed from the osteotomy site of the coracoid process. Experienced staff orthopedic surgeons (K.S., K.M., H.T.) performed all arthroscopic examinations. They were not blinded to the results of MR arthrography. The average interval between MR arthrography and second-look arthroscopy was 11 days (range, 1–31 days). The procedures were performed while the patients underwent 5 kg of vertical and 3 kg of longitudinal traction in the lateral decubitus position. Standard posterior shoulder arthroscopic portals were created, and diagnostic arthroscopic examinations were performed. Specific findings regarding the reattachment of the capsulolabral complex to the glenoid rim were recorded, including presence of failed sutures, tension of the AIGHL during 90° abduction of the shoulder, and tightness of the anterior recess.
Comparison of MR Arthrography and Arthroscopy
After the completion of 30 arthroscopic examinations, one radiologist (H.S.) tabulated all MR findings for comparison with arthroscopic findings. The arthroscopic findings were used as the standard of reference against which a radiologist (H.S.) and an orthopedic surgeon (K.S.) reassessed the MR arthrographic findings. Multiple digitally recorded arthroscopic images were available for review. Further, the orthopedic surgeon reviewed videotapes recorded during the arthroscopic examination to reevaluate tension of the AIGHL with the shoulder in the 90° abduction position. During the review of the videotapes, arthroscopic diagnosis of a failed suture was made when the capsulolabral complex was detached at previously reattached points or when loosening of sutures was identified at the anchor points. Whether sutures had failed was recorded for each anchor point.
Functional reestablishment of the capsulolabral complex was graded as good, fair, or poor according to the following criteria. When the AIGHL was firmly reattached to the glenoid rim and became tight during 90° abduction of the shoulder, functional reestablishment of the capsulolabral complex was judged as good. When the AIGHL was reattached to the glenoid rim but did not become tight during 90° abduction of the shoulder, functional reestablishment of the capsulolabral complex was judged as fair. When the AIGHL was reattached medial to the glenoid rim and did not become tight during 90° abduction of the shoulder, functional reestablishment of the capsulolabral complex was judged as poor. Functional reestablishment of the capsulolabral complex was clinically considered satisfactory when arthroscopic observation revealed good or fair reestablishment and was considered unsatisfactory when arthroscopic observation revealed poor reestablishment.
Reattachment of the anterior capsule was judged as tight when no pocket was found at the glenoid margin from the view of the posterior arthroscopic portal. Reattachment of the anterior capsule was judged as redundant when a pocket was observed at the glenoid margin from the view of the posterior arthroscopic portal.
Contingency table analysis was performed to evaluate the relationship between MR arthrographic findings of the reattachment of the AIGHL and arthroscopic findings of reestablishment of the capsulolabral complex. The same analysis was performed to evaluate the relationship between visualization of the anterior capsular attachment at MR arthrography and visualization of tightness of the anterior recess at arthroscopy. In these statistical analyses, P values less than .05 were considered significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
On oblique sagittal images, the reattached capsulolabral complex appeared as a low-signal-intensity structure adjacent to the glenoid rim at 75 anchor points (Fig 2). MR arthrographic findings of the reattached capsulolabral complex were in agreement with arthroscopic observations of 70 anchor points in 26 shoulders (Fig 2). In the remaining four shoulders, arthroscopy revealed five failed sutures (one failed suture in three shoulders and two failed sutures in one shoulder), despite the presence of a low-signal-intensity structure abutting the glenoid rim on MR arthrograms. All failed sutures that could not be detected at MR arthrography were bioabsorbable polylactic acid sutures. Although on MR arthrograms of these cases, the capsulolabral complex appeared to be sutured to the glenoid rim at the anchor points, arthroscopy revealed that it was actually detached.
|
|
|
|
|
|
Functional Reestablishment of Capsulolabral Complex
On the basis of oblique transverse MR arthrograms obtained with the shoulder in the ABER position, reattachment of the AIGHL to the glenoid rim was categorized into one of three groups: good, fair, and poor (Table 1). In the first group (n = 19), the AIGHL abutted the humeral head, covering its entire anterior margin (Fig 5). The transition from the AIGHL to the labrum was seamless. In the second group (n = 9), a small pool of contrast material was visible between the humeral head and the AIGHL, with a seamless transition from the AIGHL to the glenoid rim. In these two groups, reestablishment of the capsulolabral complex was considered to be good (n = 26) or fair (n = 2) at arthroscopy (Fig 5). In the third group (n = 2), a pool of contrast material was seen between the AIGHL and the humeral head at MR arthrography, and a "divot" occurred at the junction of the AIGHL to the glenoid rim (Fig 6). The reattachment of the AIGHL was a few millimeters from the glenoid rim. At arthroscopy, reattachment of the AIGHL was detected medial to the glenoid rim, and no appreciable tension of the AIGHL was observed during 90° external rotation of the shoulder (Fig 6). In one patient who experienced instability of the shoulder after repair, arthroscopy revealed marked synovial proliferation in the joint, a finding indicative of chronic damage due to persistent instability. Findings of reattachment of the AIGHL at MR arthrography were significantly associated with findings of reestablishment of the capsulolabral complex at arthroscopy (P < .01).
|
|
|
|
|
In four shoulders, oblique transverse images obtained with the patients in the ABER position revealed posterior subluxation of the humeral head (Fig 7). In these shoulders, the AIGHL was straightened and appeared tight. Arthroscopy revealed that reattachment of the capsulolabral complex to the glenoid rim was good.
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
In postoperative evaluation of shoulders that have been repaired with the Bankart procedure, depiction of reattachment of the labrum and the AIGHL is the most important factor at imaging that indicates functional reestablishment of the capsulolabral complex. Because MR arthrography of the shoulder is becoming widespread as a method for assessing a detached anterior labrum and torn AIGHL (9,11), this technique also appears to be useful for evaluating the shoulder joint after the Bankart procedure. Our present results confirm this assumption.
Several types of anchors are available for suture-anchor Bankart repair. In our study, the insertion site of a titanium anchor at the glenoid rim could be identified easily because of a ferromagnetic artifact. The drill-hole tracks for bioabsorbable polylactic acid anchors were somewhat more difficult to identify because of the absence of ferromagnetic artifacts. Although these drill holes are eventually refilled with normal marrow, the tracks could be seen as parallel low-signal-intensity lines in the glenoid rim when MR arthrography was performed within 400 days after surgery.
We used the analogue clock notation system to describe the position of each anchor on oblique sagittal images. Because both the radiologist and the orthopedist could determine the exact location of each anchor with this system, it appears to be effective for investigating possible correlations between MR arthrographic findings and arthroscopic findings. Furthermore, exposure of the glenoid rim and extravasation of contrast material through a failed suture were easily detected on oblique sagittal MR images. In our study, MR arthrographic findings were 93% accurate for the detection of intact or failed sutures. Therefore, we propose that the oblique sagittal plane should be included in the imaging protocol at MR arthrography of shoulders that have been repaired with the Bankart procedure, to permit evaluation of the suture status at each anchor point.
We must address two issues regarding the MR arthrographic findings of sutures. First, all failed sutures missed at MR arthrography were bioabsorbable polylactic acid sutures. However, the specific suture type appears to be unconnected to whether a suture is missed at MR arthrography, because both suture types are similar in terms of their strength and biomechanical properties. Second, to achieve good clinical results after Bankart repair, all sutures do not necessarily need to be intact. The sutures at the 1- and 2-o’clock positions often do not contribute to preventing subluxation after Bankart repair, and clinical results were often satisfactory even when sutures at these points failed. Therefore, although MR arthrographic depiction of sutures may not be 100% accurate, this technique seems appropriate for assessing the postoperative status of suture anchors.
Tirman and colleagues (7,16) introduced the ABER position for the depiction of undersurface tears of the infraspinatus tendon. Since then, authors of several studies have found that the ABER position offers advantages in MR arthrography of the shoulder (9,11,17). These studies showed that oblique transverse images obtained with patients in the ABER position are more sensitive for the detection of anterior labral detachment and torn AIGHL than are images obtained in other planes (9,11). Because maximum tension is applied to the AIGHL when patients are in the ABER position, reattachment of the AIGHL to the glenoid rim should also be depicted effectively when patients are imaged in this position. Our study confirmed this assumption. Findings of reattachment of the AIGHL to the glenoid rim at MR arthrography of the shoulder performed with patients in the ABER position were in agreement with findings of functional reestablishment of the capsulolabral complex at arthroscopy. Therefore, oblique transverse images obtained with patients in the ABER position should enable effective assessment of the postoperative status of shoulders after suture-anchor Bankart repair.
No association is apparent between tension of the AIGHL at arthroscopy and posterior subluxation of the humeral head when shoulders are in the ABER position. Because loss of external rotation after Bankart repair is relatively common (1), this MR arthrographic finding may represent relative tightening of the anterior capsule, which may prevent full external rotation after repair. Most important, our results indicate that posterior subluxation of the humeral head and straightening of the AIGHL when it is in the ABER position are not necessarily signs of anatomically insufficient reestablishment of the capsulolabral complex, although they could be indicative of a relative tightening of the AIGHL.
For optimal positioning of suture anchors, holes should be drilled in the glenoid rim at the junction of the articular cartilage and the bony rim (1,2). Thus, the anterior insertion of the capsule should be a type I insertion when the procedure is performed optimally. However, half of our patients had type II insertions. In a shoulder that has not undergone surgery, the insertion of the anterior capsule has different appearances depending on the degree of shoulder rotation (18). Because the type of anterior insertion of a reattached capsule in our study had no relationship to redundancy of the anterior recess noted at arthroscopy, the appearance of anterior capsular reattachment on transverse MR arthrograms may be due to a difference in the degree of shoulder rotation at the time of imaging.
There were several limitations to our study. First, only one reader evaluated the MR arthrograms. Interobserver agreement of MR findings in patients with suture-anchor Bankart repair should be evaluated in a study with multiple readers. Second, the study included only two patients with symptoms related to a loose AIGHL. Therefore, the MR findings related to poor functional reestablishment of the capsulolabral complex may not be reliable. Third, because all orthopedic surgeons knew the MR arthrographic findings at the time of second-look arthroscopy, their impressions at arthroscopy may have been biased. However, because the surgeons had been performing second-look arthroscopy without the aid of MR arthrographic findings at our institution for a long time, we consider the MR findings unlikely to have affected their interpretation of findings at shoulder arthroscopy.
In conclusion, we believe MR arthrography may be used instead of second-look arthroscopy to assess the postoperative condition of the shoulder after suture-anchor Bankart repair. Although the surgeons at our institution still sometimes use second-look arthroscopy, they understand the accuracy of MR arthrography and now primarily use it. Oblique sagittal MR images are valuable for assessing whether the capsulolabral complex is sutured or detached at each anchor point. Reattachment of the AIGHL to the glenoid rim can be determined confidently with an oblique transverse MR arthrogram obtained with the patient in the ABER position.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, H.S., K.S.; study concepts, H.S., H.T.; study design, H.S., K.S.; literature research, H.S.; clinical studies, all authors; data acquisition and analysis/interpretation, H.S., K.S.; statistical analysis, H.S.; manuscript preparation, H.S., K.S.; manuscript definition of intellectual content, H.S.; manuscript editing, H.S., H.T.; manuscript revision/review, H.S., K.S., H.T.; manuscript final version approval, H.S., H.T.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. V. R. Mohana-Borges, C. B. Chung, and D. Resnick MR Imaging and MR Arthrography of the Postoperative Shoulder: Spectrum of Normal and Abnormal Findings RadioGraphics, January 1, 2004; 24(1): 69 - 85. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
