| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 All authors: Department of Radiology, Université Libre de Bruxelles and Erasme Hospital, Route de Lennik 808, Brussels 1170, Belgium.
Received March 1, 2005;
accepted after revision September 4, 2005.
Address correspondence to B. Stallenberg.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Bursography, sonography, and CT were performed in six cadaver shoulders and compared with anatomic sections in neutral position and while the humerus was extended and internally rotated. For the clinical study, the width of the anterior portion of the bursa was measured in both positions in both shoulders of 27 patients referred because of shoulder pain and in eight asymptomatic volunteers. Pain was coded as absent, experienced in the anteromedial portion of the shoulder, or experienced elsewhere but not anteromedially, and we compared the pain scores between shoulder positions.
RESULTS. In all cadaver shoulders, when compared with CT scans and anatomic sections, sonography showed the morphology of the bursa, its relationships with surrounding structures, and morphologic changes associated with position. In volunteers, the mean width of the bursa was 0.74 ± 0.05 and 0.93 ± 0.09 mm (p = 0.013), respectively, in neutral and stress position. In patients, the same values were 0.70 ± 0.07 and 0.81 ± 0.14 mm (p = 0.286) in the asymptomatic side and 1.20 ± 0.11 and 1.75 ± 0.23 mm (p < 0.001) in the symptomatic side, respectively. The bursa was wider in patients experiencing pain anteromedially than in those who experienced pain elsewhere and volunteers (p = 0.002 and < 0.001, respectively), and the bursa was wider in symptomatic shoulders than in asymptomatic shoulders (p < 0.001).
CONCLUSION. Widening of the anterior portion of the subacromial-subdeltoid bursa is associated with anteromedial shoulder pain and the clinical syndrome of coracoid impingement.
Keywords: anatomy • coracoid impingement • shoulder • sonography • subacromial-subdeltoid bursa
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In our clinical practice, we have observed at sonography that patients complaining of anteromedial shoulder pain frequently had a widened anterior portion of the subacromial-subdeltoid bursa. Several studies have suggested the involvement of this bursa in shoulder pain experienced by patients with impingement syndrome [2, 3], but quantitative data are still not available. Coracoid impingement is a cause of anteromedial shoulder pain. The aim of the present study was to verify the hypothesis that increased widening of the anterior portion of this bursa could be associated with anteromedial shoulder pain and with the symptoms of coracoid impingement. Because these patients do not usually undergo surgery for such disorders—preventing us from obtaining surgical confirmation—we first conducted a cadaver study to compare the morphology of the anterior portion of the subacromial-subdeltoid bursa between sonograms and CT scans. Thereafter, we conducted a clinical study to compare the width of the subacromial-subdeltoid bursa measured on sonography between painful and nonpainful shoulders.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cadaver Study
Specimens—Six shoulders were obtained from three human
cadavers (one woman and two men) with age at death ranging from 72 to 79
years. The medical history of these cadavers was not available. Four fresh
shoulders were harvested from two nonembalmed cadavers, and two shoulders were
harvested from one embalmed cadaver. The specimens were deep-frozen at
-40°C immediately after death. All specimens were allowed to thaw for 24
hours at room temperature before imaging. To preserve soft tissues, the
specimens consisted of the cephalic part of the thorax cut at the level of the
seventh thoracic vertebra and the proximal portion of the arm through the
distal third of the humeral diaphysis. To move the arm, a screw was placed
through the proximal third of the humeral diaphysis of nonembalmed
specimens.
Bursography—While the cadaver specimens were placed in the supine position, a 19-gauge (1.1 x 40.0 mm) needle was inserted obliquely through the skin with an anterior approach under fluoroscopic guidance [6]. The needle was inserted until the subacromial bursa was entered. In the nonembalmed shoulders, 8-15 mL of iodinated contrast material (3.2 mg I/10 mL of ioxaglate meglumine [Hexabrix 320, Schering]) was injected in the subacromial bursa. In the embalmed shoulders, after injection of a small amount of iodinated contrast material to verify the position of the needle within the subacromial bursa, 9-12 mL of a 20% concentrated solution of gelatin was injected into the bursa until resistance to further injection was felt by the operator.
Neutral position of the shoulder was determined on the anteroposterior fluoroscopic view by positioning the bicipital groove at two thirds laterally between the tip of the coracoid process medially and the great tuberosity laterally. After removing the needle, the pressure was maintained at the point of puncture for 2 minutes to prevent leakage of contrast material or gelatin into the subcutaneous soft tissues.
Sonography—Sonography was performed immediately after bursography by a musculoskeletal radiologist with 15 years of experience who used a broadband linear array transducer with frequencies of 5.0-12.0 MHz on a commercially available sonography unit (HDI 3000 system, ATL). Grayscale images of the bursa were obtained in the transverse plane at the level of the tip of the coracoid process and approximately 2 cm above and below its tip. A line was drawn on the skin, parallel to the transducer, to indicate the transverse plane for the subsequent CT examination and additional anatomic slices. In the four nonembalmed shoulders, the bursa was imaged while the humerus was in neutral position and while it was simultaneously extended (i.e., posteriorly elevated) and internally rotated, as recommended by Burns and Whipple [7]. We used this position to test the coracoid impingement instead of the one proposed by Gerber et al. [2] because it allows visualization of the anterior part of the shoulder on sonography.
CT—Within 20 minutes after bursography, CT images were obtained using a commercially available single-detector scanner (Somatom Plus, Siemens Medical Solutions). CT scans were obtained in a plane parallel to the line previously drawn on the skin. Data were obtained with a 2-mm collimation at 140 kVp and 189 mAs. The table feed was 2 mm per 1 sec of scanner rotation, resulting in a pitch of 1.0. From the row data, 2-mm-thick axial sections were reconstructed with a 2-mm increment. In the four nonembalmed shoulders, the bursa was imaged while the humerus was in neutral position and while it was simultaneously extended and internally rotated.
Anatomy and comparisons with sonography and CT—After sonography and CT, all cadaver shoulders were frozen at -40°C for 48 hours. Shoulders were sectioned with a band saw into 5-mm-thick transverse sections parallel to the line drawn on the skin. Two radiologists compared by consensus the anatomy of the bursa from the bicipital groove to the tip of the coracoid process between sonography images, CT scans, and anatomic sections.
Clinical Study
Subject population—Between September and December 2002,
sonography of both shoulders was performed in 108 consecutive unselected
patients referred by orthopedic surgeons and rheumatologists for imaging of
the shoulder because of shoulder pain. Among these patients, 64 presented with
unilateral symptoms and the other shoulder was asymptomatic with no history of
trauma. Of these 64 patients, those with a full-thickness tear of the rotator
cuff on the asymptomatic side were excluded because that injury can be seen in
asymptomatic patients and can induce an increase in the thickness of the
synovium, liquid accumulation in the subacromial bursa, or both
[8]. In addition, patients with
an abnormal long head biceps tendon, rotator interval, or subscapularis tendon
in the symptomatic side were also excluded because those abnormalities can be
associated with anteromedial pain
[5,
9]. Thirty-seven patients were
excluded. Our study patients thus consisted of the 27 remaining patients (nine
men, 18 women; mean age, 44 years; age range, 20-79 years).
In addition, 12 asymptomatic volunteers with no history of shoulder symptoms or shoulder trauma underwent sonography. These volunteers were not a sample of the general population; they were recruited from among hospital employees. Eight of these 12 volunteers (three women, five men; mean age, 40 years; age range, 24-59 years) had normal sonography findings bilaterally. These eight volunteers were included in the present study. Thus, our study sample consisted of 43 asymptomatic and 27 painful shoulders.
When our study was performed, the policy of the Ethics Committee at the University Erasme Hospital was that it was the responsibility of the investigators involved in a study to determine whether ethics committee review and approval were needed. Because the present sonography study was part of the routine workup of patients referred by physicians for clinical suspicion of rotator cuff tear, the ethics committee did not require submission of our study protocol; however, we did obtain written informed consent from each volunteer after explaining the study rationale and procedures.
Sonography—Sonography images were obtained on the same machine and by the same radiologist as in the cadaver study. The rotator cuff, the long head biceps tendon, and the rotator interval were imaged while the humerus was in neutral position, externally rotated, and extended and internally rotated at the maximum possible for the subject. In neutral position and while the humerus was extended and internally rotated, the anteroposterior width of the bursa was measured in the transverse plane at the level of the tip of the coracoid process. In neutral position, the widest part of the bursa between the tip of the coracoid process and the lateral and proximal portion of the bicipital groove was measured. Because the bicipital groove was no longer visible while the humerus was extended and internally rotated, the previously measured interval distance was reported from the tip of the coracoid process and the widest part of the bursa within this interval was measured. Thus, at both positions, the same landmarks were considered, and all measurements were obtained in the same transverse plane at the level of the tip of the coracoid process. The transducer pressure was minimized to avoid compression of the bursa. In addition, type 1 calcification within the supraspinatus and subscapularis tendons was coded as present or absent [10].
With their shoulder in neutral position, the patients and volunteers were asked to indicate where they experienced pain. We divided our study sample into three groups according to pain coded as absent (group 1), experienced in the anteromedial part of the shoulder (group 2), or experienced elsewhere but not in the anteromedial part of the shoulder (group 3). While patients and volunteers extended and rotated the humerus internally, we asked them to compare the pain in this position with that in neutral position. We then divided our study sample into three groups according to the modifications in the pain with the movement of the shoulder as still absent (group A), equal or decreased (group B), or increased (group C) in the anteromedial part of the shoulder.
|
|
|
Statistical Methods
Results are expressed as mean ± SEM. Means were compared using the
Student's t test for paired or independent samples in cases of
comparisons between two groups and using a one-way analysis of variance
procedure followed with a multiple comparison procedure of Tukey in cases of
comparisons between three groups.
Statistical significance for all tests was set at a p value of less than 0.05. The statistical software used was SPSS (release 11.0, Statistical Package for the Social Sciences) for Windows (Microsoft).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
Clinical Study
In patients and in volunteers, the subacromial-subdeltoid bursa appeared on
sonography as hypoechoic tissue between highly reflective lines. Measurements
of the bursa in both positions are summarized in
Figure 3. In both positions,
the width did not differ significantly between volunteers and the asymptomatic
side of patients (p = 0.703 and 0.540, respectively). In both
positions, the width was significantly lower in volunteers than in the
symptomatic side of patients (p < 0.001 and < 0.002,
respectively) and in the asymptomatic than in the symptomatic side of patients
(p < 0.001).
|
In neutral position, the mean width of the bursa was 0.71 ± 0.04 mm in group 1, 1.48 ± 0.13 mm in group 2, and 0.95 ± 0.14 mm in group 3. Differences in width reached statistical significance between groups 1 and 2 (p < 0.001) and between groups 2 and 3 (p = 0.002), but not between groups 1 and 3 (p = 0.125). While the humerus was extended and internally rotated, the mean width of the bursa was 0.81 ± 0.08 mm in group A, 1.96 ± 0.46 mm in group B, and 2.73 ± 0.19 mm in group C. Differences in width reached statistical significance between groups A and B (p < 0.001) and between groups A and C (p < 0.001), but not between groups B and C (p = 0.072). Measurements and statistical comparisons are summarized in Figures 4 and 5. An illustrative case is shown in Figures 6A and 6B. In the symptomatic side, three patients had a full-thickness tear, and four patients had supraspinatus tendinopathy without a full-thickness tear.
|
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our anatomic study shows that bulging and folding of the anterior part of the subacromial-subdeltoid bursa facing the coracoid process occur while the humerus is simultaneously extended and internally rotated. This bursa is firmly adherent to the superior aspect of the rotator cuff and to the under aspect of the coracoacromial ligament and the acromion, but it is less adherent to the under aspect of the coracoid process, the deltoid muscle, and the rotator interval [11, 12]. When the humerus moves, the difference in adherence between the bursa and the surrounding structures leads to its folding and bulging on the tip of the coracoid process and the most anterior portion of the acromio-clavicular ligament [3], similarly to the bulging of the subacromial-subdeltoid bursa observed on the outer edge of the acromion [13, 14].
Goldthwait [15] has shown that morphologic abnormalities of the coracoid process can induce the compression of the coracoid bursa between this process and the humeral head. The present study suggests that the anterior portion of the subacromial-subdeltoid bursa can also be involved in this impingement. On the other hand, previous anatomic studies have shown that communications may exist between the subcoracoid portion of the subacromial-subdeltoid bursa and the subcoracoid bursa [16], suggesting that in such individuals, both bursae could be considered as one structure. In our anatomic study, such a communication was detected in one cadaver shoulder. Using sonography, we were nevertheless not able to detect such communications in living subjects because the medial portion of this bursa extends posteriorly to the coracoid process and is thus not visible because of acoustic shadowing produced by the coracoid process.
As reported by van Holsbeeck and Strouse [17], the subacromial-subdeltoid bursa appears as a complex structure consisting of a hypoechoic-anechoic area between two layers of fat with a maximum width of 2 mm, with the hypoechoic surface area increasing with the amount of fluid within the bursa. Fluid appears anechoic, but it may be undetectable if squeezed away by the pressure exerted with the ultrasound probe [17, 18]. In asymptomatic shoulders of our patients and in the shoulders of volunteers, the width of the hypoechoic area was approximately 0.7 mm, and even with minimal transducer pressure, no fluid was detectable within the anterior portion of the bursa facing the tip of the coracoid process. In patients with impingement syndrome, fluid was not detectable in front of the tip of the coracoid process. This could be explained by synovial thickening without any fluid collection or by fluid accumulation in portions of the bursa not investigated in the present study, particularly while the humerus was extended and internally rotated [17, 18].
Our clinical study shows that increased widening of the subacromial-subdeltoid bursa is associated with anteromedial shoulder pain. The subacromial-subdeltoid bursa is a major focus of pain because it is anatomically vulnerable to friction leading to subsequent abnormalities [8]. Neer and Welsh [8] have proposed three clinical and surgical stages in impingement lesions: first, edema and hemorrhage within the bursa; second, widening and fibrosis of the bursa and tendinitis; and third, bone spurs and tendon rupture. All these conditions are associated with widening of the bursa. Moreover, the inflammatory response leads to synovial proliferation, and its severity is correlated with the intensity of pain [19-21]. In our patients who were experiencing pain in the anteromedial portion of the shoulder, synovium of the corresponding portion of the bursa was thicker than in patients with pain elsewhere in the shoulder. In these patients, another portion of the bursa could be the widest, but we did not search for it in the present study, which focused on the anterior portion of the subacromial-subdeltoid bursa.
Our study also shows that while the humerus is simultaneously extended and internally rotated, the widening of the bursa is more increased in symptomatic shoulders than in nonsymptomatic ones. Among symptomatic patients, the widening of the bursa tended to be higher in those who experience increased pain during the maneuver than in those who did not, but this difference did not reach statistical significance. Interestingly, in patients who had anteromedial pain in neutral position, this pain increased when the humerus was extended and internally rotated. Conversely, the more the bursa was widened anteriorly in neutral position, the more its widening induced by simultaneous extension and internal rotation of the humerus was increased. As shown in our anatomic study, widening of the bursa may be due to deformation and bulging when a shoulder moves. On the other hand, the anteromedial shoulder pain could also involve the coracoacromial ligament because sensitive neural proliferation has been reported in the bursal tissue surrounding this ligament in patients with rotator cuff tear [22]. Stretching this ligament is related to impingement, and surgical removal of both the coracoid process and the adjacent part of the coracoacromial ligament relieves pain [1, 5, 23]. Both stretching the ligament and the subacromial-subdeltoid bursa and bulging of this bursa around the tip of the coracoid process while extending and internally rotating the humerus could thus explain pain experienced when the shoulder is in this position.
Morphologic changes in the coracoid process are rare but may produce coracoid impingement [1, 24]. Enlargement of soft tissues between the coracoid process and the glenohumeral joint by structures such as a ganglion cyst and subscapularis calcifications may also produce coracoid impingement [1, 2, 25]. In the present study, we did not find such structures or calcifications, but we detected a calcification larger than 1 cm in diameter within the supraspinatus tendon in two patients. In those two patients, the bursa was widened in both positions and the shoulder was more painful when the arm was extended and internally rotated than when it was in neutral position. Burns and Whipple [7] have reported a contact between the supraspinatus tendon at its junction with the coracoacromial ligament and the coracoid process in this position. As suggested by our study, a calcification that enlarges the supraspinatus tendon can be associated with anteromedial pain, particularly in a stress position.
Our study may have some limitations. First, we did not use a visual analog scale to grade pain, preventing us from correlating the widening of the bursa and the intensity of pain. Second, because we did not collect arthroscopic or surgical data, we were not able to verify the relationship, as proposed by Neer and Welsh [8], Ptasznik [13], and Neer [26], between the widening of the bursa and clinical, surgical, and histologic stages. Third, a specific impingement test might have increased the diagnostic value if pain relief had been induced by injection of 10 mL of 1% lidocaine into the subcoracoid space [23]. Such a test was not performed because our recruitment process was based on outpatients referred by orthopedic surgeons and rheumatologists, thus preventing us from obtaining consistent follow-up information. Fourth, measurements were not repeated either within or between observers.
In conclusion, the present study shows that widening of the anterior portion of the subacromial-subdeltoid bursa measured on sonography is associated with anteromedial shoulder pain and could be an additional cause of symptoms of coracoid impingement.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
