Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print May 18, 2006, 10.1148/radiol.2393050453
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2393050453v1
240/1/152    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Herold, T.
Right arrow Articles by Paetzel, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Herold, T.
Right arrow Articles by Paetzel, C.
(Radiology 2006;240:152-160.)
© RSNA, 2006

Musculoskeletal Imaging

Indirect MR Arthrography of the Shoulder: Use of Abduction and External Rotation to Detect Full- and Partial-Thickness Tears of the Supraspinatus Tendon1

Thomas Herold, MD, Maike Bachthaler, MD, Okka W. Hamer, MD, Reiner Hente, MD, Stefan Feuerbach, MD, Claudia Fellner, PhD, Michael Strotzer, MD, Markus Lenhart, MD and Christian Paetzel, MD

1 From the Department of Diagnostic Radiology, University Hospital of Regensburg, Franz-Josef-Strauss-Allee 11, D-93042 Regensburg, Germany. Received March 18, 2005; revision requested May 12; revision received May 27; accepted June 21; final version accepted August 11. Address correspondence to T.H. (e-mail: thomas.herold{at}klinik.uni-regensburg.de).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Purpose: To evaluate prospectively the accuracy of indirect magnetic resonance (MR) arthrography for supraspinatus tendon tears during neutral positioning or abduction and external rotation (ABER) and neutral positioning.

Materials and Methods: Informed consent was obtained in all patients, and the study was approved by the institutional review board. Indirect MR arthrography of the shoulder was performed in 51 symptomatic patients (14 female, 37 male; mean age, 47 years) in the neutral position (set 1) and in the neutral and ABER positions (set 2). Two readers independently interpreted both sets, and diagnoses were compared with arthroscopic findings. Diagnostic accuracy was calculated, and 95% confidence intervals were used to detect significant differences between sets. Diagnostic confidence was recorded by using a three-level confidence score. Differences between sets were evaluated by using the Wilcoxon signed rank test. Interobserver agreement was determined separately for each set and for all diagnoses, full-thickness tears, and partial-thickness tears.

Results: For full-thickness tears, there was no benefit to reading set 2. For reader 1, sensitivity and specificity were 95% and 100%, respectively, for set 1 and 100% and 100%, respectively, for set 2. For reader 2, sensitivity and specificity were 80% and 100%, respectively, for set 1 and 100% and 100%, respectively, for set 2. For partial-thickness tears, sensitivity was significantly higher after reading set 2. For reader 1, sensitivity and specificity were 71% and 88%, respectively, for set 1 and 93% and 100%, respectively, for set 2. For reader 2, sensitivity and specificity were 50% and 88%, respectively, for set 1 and 86% and 94%, respectively, for set 2. For both readers, diagnostic confidence for partial-thickness tears was significantly higher after reading set 2. After the interpretation of set 2, {kappa} values increased from 0.35 to 1.00 for full-thickness tears and from 0.12 to 0.63 for partial-thickness tears.

Conclusion: Indirect MR arthrography with supplementary images obtained with patients in the ABER position significantly improved sensitivity and increased diagnostic confidence for partial-thickness tears of the supraspinatus tendon. Interobserver agreement was improved for both full- and partial-thickness tears.

© RSNA, 2006


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
A tear of the rotator cuff is one of the most frequent injuries of the shoulder. These tears result from sports injuries in young patients and from degeneration or overload in elderly patients (1–3). With an increasing demand in activity, there is a tendency to perform arthroscopy or surgery on the rotator cuff earlier—even in elderly patients—because repair of even partial-thickness tears may alleviate symptoms (2,4–6). Therefore, noninvasive imaging techniques have gained importance for confirming the presence and type of tear. Direct magnetic resonance (MR) arthrography has proved to be more sensitive than conventional MR imaging in the diagnosis and characterization of tears of the supraspinatus tendon (7–12). The intraarticular injection of saline or paramagnetic contrast material, however, is invasive and must be performed with fluoroscopic or ultrasonographic (US) guidance (7,13). It has recently been shown that the intravenous administration of gadopentetate dimeglumine enhances the joint cavity and thus indirectly produces an arthrographic effect (14–16). This technique, which is known as indirect MR arthrography, has shown a sensitivity of up to 100% in the detection of full-thickness tears of the rotator cuff when arthroscopy or surgery was used as the reference standard (3,17–20).

The correct diagnosis of partial-thickness tears, however, remains difficult, regardless of whether unenhanced MR imaging or MR arthrography is performed (8–10,18,20). Hence, some authors have suggested that supplementary MR arthrography performed with the patient's arm in abduction and external rotation (ABER) can improve the diagnosis of partial-thickness tears of the rotator cuff (21–23). To date, indirect MR arthrography combined with the ABER position has been performed only for anterior shoulder instability (24), and direct MR arthrography has been performed in all published series in which the ABER position was used to detect rotator cuff tears.

Thus, the purpose of our study was to evaluate prospectively the accuracy of indirect MR arthrography in demonstrating supraspinatus tendon tears during neutral positioning or ABER and neutral positioning, with arthroscopy used as the reference standard.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Patients
From October 2003 to October 2004, 67 symptomatic patients who were referred to our institution for MR arthrography of the shoulder were consecutively enrolled in this prospective study. Arthroscopy was planned in all patients and was indicated on the basis of clinical assessment. Sixteen patients (five women and 11 men) were excluded on the basis of the following criteria: (a) MR imaging was not feasible because of a cardiac pacemaker (n = 4) or severe claustrophobia (n = 3) and (b) patients were unable to keep their symptomatic arm in the ABER position due to severe pain (n = 6) or noncompliance (n = 3). In the remaining 51 patients, indirect MR arthrography with the additional ABER position was feasible. These patients (14 female and 37 male; age range, 16–69 years; mean age, 47 years) comprised the final study group. The right shoulder was involved in 32 patients, and the left shoulder was involved in 19 patients. A history of trauma was reported in 17 (33%) of 51 patients. Fourteen (27%) of 51 patients had previous luxation, and 36 (71%) presented with clinical signs of impingement. None of the patients had previously undergone surgery in the symptomatic shoulder. Written informed consent was obtained from all patients or their guardians, and our study was approved by the institutional review board.

Imaging Technique
All MR examinations were performed with a 1.5-T unit (Magnetom Symphony; Siemens Medical Solutions, Erlangen, Germany). Imaging was performed with a gradient strength of 20 mT/m, rise time of 0.4 msec, and slew rate of 50 (T · m–1)/sec. Indirect MR arthrography of the shoulder was initiated after intravenous injection of gadopentetate dimeglumine (Magnevist, Schering, Berlin, Germany; vial concentration, 0.5 mmol/L; applied dose, 0.1 mmol per kilogram body weight) into an antecubital vein. Immediately after injection, patients were instructed to move the injured shoulder for 15 minutes (rotation, abduction, and adduction).

Fifteen minutes after the injection of gadopentetate dimeglumine, the patients were positioned in the MR unit (head first and supine). Imaging was performed with patients in the ABER position as described by Tirman et al (21). Patients were instructed to place the hand of the affected arm behind the head, with the elbow flexed. A medium-sized (17-cm) flexible surface coil was placed on the axilla. A T1-weighted coronal localizer sequence (15/5 [repetition time msec/echo time msec], 8-mm section thickness, 0.16-mm intersection gap, 500-mm field of view, and one signal acquired) was performed. Frequency-selective fat-saturated T1-weighted oblique transverse spin-echo MR images were obtained along the axis of the humerus (1011/20, 3-mm section thickness, 0.3-mm intersection gap, 180-mm field of view, 307 x 512 matrix, and two signals acquired).

Next, patients were instructed to place their arm in the neutral position, and the flexed surface coil was repositioned to guarantee optimal coverage of the shoulder. While patients were in the neutral position, a transverse localizer sequence (15/5, 8-mm section thickness, 0.16-mm intersection gap, 500-mm field of view, and one signal acquired) was performed. A T1-weighted frequency-selective fat-saturated spin-echo MR imaging sequence (1011/20, 3-mm section thickness, 0.3-mm intersection gap, 180-mm field of view, 307 x 512 matrix, and two signals acquired) was performed in the coronal oblique (parallel to the long axis of the supraspinatus tendon) and transverse planes.

For the assessment of fatty infiltration and atrophy of the supraspinatus muscle, a T1-weighted spin-echo MR imaging sequence (720/20, 3-mm section thickness, 0.3-mm intersection gap, 180-mm field of view, and 307 x 512 matrix) was performed in the sagittal oblique plane (perpendicular to the long axis of the supraspinatus tendon) without fat saturation. For optimal visualization of extraarticular disease and to reduce misinterpretation caused by the magic angle phenomenon, an additional coronal oblique T2-weighted fast spin-echo MR imaging sequence (3000/119, 3-mm section thickness, 0.3-mm intersection gap, 180-mm field of view, 180 x 256 matrix, turbo factor of seven, and two signals acquired) was performed.

The ABER position necessitated an additional table time of 9 minutes. Acquisition time was 5 minutes 45 seconds, and positioning and repositioning of the shoulder necessitated an average of 3 minutes 15 seconds. No complications occurred after indirect MR arthrography.

Image Analysis and Comparison with Arthroscopy
Images were divided into two sets; one set (set 1) included all images obtained with patients in the neutral position and the other set (set 2) included all images from set 1 and all images obtained with patients in the ABER position. Two board-certified radiologists (T.H. and M.L., with 6 and 8 years of experience in musculoskeletal radiology, respectively) analyzed the images independently and prospectively. The readers were blinded to clinical data and to the results of arthroscopy. First, the readers analyzed the images from set 1. Images from set 2 were analyzed 3 weeks later. Images were presented in random order, and readers were blinded to patient data. Images were evaluated by using a picture archiving and communication system (Magic View 300; Siemens) with black and white high-luminosity monitors (Sienet; Siemens) that had a resolution of 1280 x 1024 pixels.

The findings at indirect MR arthrography were compared with those at arthroscopy by one radiologist (T.H.) and one orthopedic surgeon (R.H.) after sets 1 and 2 were evaluated. The time between MR imaging and arthroscopy ranged from 25 to 175 days (mean, 37 days). One board-certified orthopedic surgeon (R.H.) with 14 years of experience in the surgical treatment of the shoulder and 8 years of experience in arthroscopy of the shoulder performed all arthroscopic examinations and surgical procedures. During arthroscopy, all patients underwent exploration of the bursal and glenohumeral compartments. There were no complications caused by arthroscopy.

The interpretation of both sets of MR images was performed according to previously described criteria, which included alterations in signal intensity and morphologic changes of the supraspinatus tendon (8–10,19,21,22,25–30). The readers were instructed to classify their findings as no tear, full-thickness tear, or partial-thickness tear.

For the purpose of this study, a normal supraspinatus tendon or a diagnosis of tendinosis or degeneration was classified as no tear. For both sets of images, a diagnosis of no tear was made if the tendon had homogeneous low signal intensity on all images. In case of ill-defined areas of high signal intensity on fat-saturated T1-weighted MR arthrograms and T2-weighted MR images, the high signal intensity was not to be associated with contour abnormalities, enlargement of the tendon, element interruption, or retraction.

A full-thickness tear was diagnosed if there was high signal intensity involving the entire thickness and tendinous discontinuity with or without retraction of the musculotendinous junction. Bursal joint fluid was not essential for the diagnosis of a full-thickness tear (8,31).

A partial-thickness tear of the tendon was diagnosed when a focus of high signal intensity involving either the bursal or the articular surface of the tendon was apparent with surface abnormalities but without complete discontinuity of the tendon on T1- and T2-weighted MR images. An intratendinous partial-thickness tear was diagnosed when this well-defined area of high signal intensity was completely located within the tendon and did not extend to the bursal or articular surface. For statistical analysis, no attempt was made to differentiate between partial tears located on the bursal or articular surface and intratendinous tears.

Readers also recorded their degree of diagnostic confidence by using a three-level confidence score (0, diagnosis unsure; 1, diagnosis probable; and 2, diagnosis sure). The confidence score was determined separately for each image set and for each reader. We did not evaluate the infraspinatus, subscapularis, or long biceps tendon in this study, although a complete interpretation of the MR examinations was performed for patient care.

Statistical Analysis
By using arthroscopic results as the standard of reference, we calculated the sensitivity, specificity, and accuracy of indirect MR arthrography for the presence of a tear (full- or partial-thickness tear) compared with the presence of no tear for each reader and each image set. The tear category was then subdivided into full- and partial-thickness tears, and the sensitivity, specificity, and accuracy were again calculated relative to the diagnosis of no tear. A statistically significant difference between the two sets was suggested if the 95% confidence intervals did not overlap. For each reader, differences in confidence ratings between sets 1 and 2 were evaluated by using the Wilcoxon signed rank test. A P value of .05 or less was considered indicative of a statistically significant difference. The Cohen {kappa} value for agreement between the two readers was calculated for each set and for all diagnoses, full-thickness tears, and partial-thickness tears. Interobserver agreement was rated as poor for a {kappa} value of less than or equal to 0.2, fair for a {kappa} value of 0.21–0.40, moderate for a {kappa} value of 0.41–0.60, good for a {kappa} value of 0.61–0.80, and excellent for a {kappa} value of 0.81–1.00 (32). Data entry procedures and statistical analyses were performed by using a statistical software program (SPSS, version 12; SPSS, Chicago, Ill).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Arthroscopy
Arthroscopy revealed tears of the supraspinatus tendon in 34 (67%) of 51 patients. Partial-thickness tears were diagnosed in 14 (27%) of 51 patients, and full-thickness tears were diagnosed in 20 (39%) of 51 patients (Table 1). The 14 partial-thickness tears were further classified with arthroscopy as being bursal-side tears (four [8%] of 51 patients), joint-side tears (eight [16%] of 51 patients), and intratendinous tears (two [4%] of 51 patients). In 11 patients, the supraspinatus tear extended into the infraspinatus tendon. Seven patients had a superior labral anteroposterior lesion as an associated abnormality, eight had Bankart lesions, six had Hill-Sachs deformities, and four had a posterior labral tear.


View this table:
[in this window]
[in a new window]

  		Table 1. Arthroscopic Findings for the Supraspinatus Tendon

 
Sensitivity and Specificity
With regard to the diagnosis of a full-thickness tear, indirect MR arthrography with patients in the neutral position (set 1) had a high diagnostic accuracy (Table 2). There were no false-positive findings—that is, neither reader overestimated tears (Table 1). After reading set 1, readers 1 and 2 interpreted one and four full-thickness tears, respectively, as partial-thickness tears (false-negative findings). After review of the supplementary images obtained with patients in the ABER position (set 2), neither reader had any false-negative findings (Table 1, Fig 1). Within the category of full-thickness tears, overlapping 95% confidence intervals were not indicative of a statistically significant difference between sensitivity and specificity for the two sets.


View this table:
[in this window]
[in a new window]

  		Table 2. Diagnostic Accuracy in the Evaluation of the Supraspinatus Tendon

 

Figure 1
View larger version (144K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a: Full-thickness tear of the supraspinatus tendon in 47-year-old man. (a) Coronal oblique fat-saturated T1-weighted contrast material–enhanced indirect MR arthrogram (1011/20) obtained with patient in the neutral position shows incomplete element interruption (long arrow) in bursal-side fibers of supraspinatus tendon. On the joint-side surface, tendon seems to be intact (short arrows). Both readers assessed this lesion as a bursal-side partial-thickness tear. (b) Corresponding image obtained with patient in the ABER position (1011/20) demonstrates complete discontinuity of the tendon, which was filled by enhanced granulated tissue (arrows). Full-thickness tear was confirmed at arthroscopy.

 

Figure 1
View larger version (159K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b: Full-thickness tear of the supraspinatus tendon in 47-year-old man. (a) Coronal oblique fat-saturated T1-weighted contrast material–enhanced indirect MR arthrogram (1011/20) obtained with patient in the neutral position shows incomplete element interruption (long arrow) in bursal-side fibers of supraspinatus tendon. On the joint-side surface, tendon seems to be intact (short arrows). Both readers assessed this lesion as a bursal-side partial-thickness tear. (b) Corresponding image obtained with patient in the ABER position (1011/20) demonstrates complete discontinuity of the tendon, which was filled by enhanced granulated tissue (arrows). Full-thickness tear was confirmed at arthroscopy.

 
The diagnostic accuracy of indirect MR arthrography with patients in the neutral position (set 1) was lower for the detection of partial-thickness tears than for the detection of full-thickness tears (Table 2). Of the 14 partial-thickness tears, 10 were correctly diagnosed by reader 1 and eight were correctly diagnosed by reader 2 (Table 1, Fig 2). Four partial-thickness tears were misdiagnosed as no tear by reader 1, and five partial-thickness tears were misdiagnosed as no tear by reader 2 (false-negative findings).


Figure 2
View larger version (152K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a: Joint-side partial-thickness tear of the supraspinatus tendon in 37-year-old man. (a) Coronal oblique fat-saturated T1-weighted contrast-enhanced indirect MR arthrogram (1011/20) obtained with patient in the neutral position demonstrates element interruption (arrow) of articular-side fiber of tendon. (b) Fat-saturated T1-weighted contrast-enhanced indirect MR arthrogram (1011/20) obtained with patient in the ABER position shows partial-thickness tear (arrow).

 

Figure 2
View larger version (158K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b: Joint-side partial-thickness tear of the supraspinatus tendon in 37-year-old man. (a) Coronal oblique fat-saturated T1-weighted contrast-enhanced indirect MR arthrogram (1011/20) obtained with patient in the neutral position demonstrates element interruption (arrow) of articular-side fiber of tendon. (b) Fat-saturated T1-weighted contrast-enhanced indirect MR arthrogram (1011/20) obtained with patient in the ABER position shows partial-thickness tear (arrow).

 
In detail, readers underestimated three bursal-side partial-thickness tears (one by reader 1 and two by reader 2) (Fig 3), one joint-side partial-thickness tear (Fig 4), and two intratendinous partial-thickness tears (Fig 5). Interpretation of supplementary images obtained with patients in the ABER position (set 2) reduced the rate of false-negative interpretations, with only one bursal-side partial-thickness tear (reader 2) and one intratendinous tear (both readers) being wrongly categorized as no tear.


Figure 3
View larger version (180K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a: Bursal-side partial-thickness tear of the supraspinatus tendon in 43-year-old woman. (a) Coronal oblique fat-saturated T1-weighted contrast-enhanced indirect MR arthrogram (1011/20) obtained with patient in the neutral position does not depict tear because neither suspected morphologic changes nor signal intensity alterations are visible (arrow). (b) Fat-saturated T1-weighted contrast-enhanced indirect MR arthrogram (1011/20) obtained with patient in the ABER position shows how tendon kinking causes the spreading of fibers and correctly demonstrates bursal-side partial-thickness tear (long arrow) with horizontal component (short arrow).

 

Figure 3
View larger version (144K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b: Bursal-side partial-thickness tear of the supraspinatus tendon in 43-year-old woman. (a) Coronal oblique fat-saturated T1-weighted contrast-enhanced indirect MR arthrogram (1011/20) obtained with patient in the neutral position does not depict tear because neither suspected morphologic changes nor signal intensity alterations are visible (arrow). (b) Fat-saturated T1-weighted contrast-enhanced indirect MR arthrogram (1011/20) obtained with patient in the ABER position shows how tendon kinking causes the spreading of fibers and correctly demonstrates bursal-side partial-thickness tear (long arrow) with horizontal component (short arrow).

 

Figure 4
View larger version (170K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4a: Joint-side partial-thickness tear of the supraspinatus tendon in 55-year-old woman. (a) Coronal oblique contrast-enhanced fat-saturated T1-weighted indirect MR arthrogram (1011/20) obtained with patient in the neutral position shows low-signal-intensity supraspinatus tendon (arrow), with no alteration in signal intensity. Both readers classified this as no tear. (b) Corresponding image (1011/20) obtained with patient in the ABER position depicts small flap (arrow) on joint-side area of tendon. Joint-side partial-thickness tear was correctly diagnosed by both readers.

 

Figure 4
View larger version (174K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4b: Joint-side partial-thickness tear of the supraspinatus tendon in 55-year-old woman. (a) Coronal oblique contrast-enhanced fat-saturated T1-weighted indirect MR arthrogram (1011/20) obtained with patient in the neutral position shows low-signal-intensity supraspinatus tendon (arrow), with no alteration in signal intensity. Both readers classified this as no tear. (b) Corresponding image (1011/20) obtained with patient in the ABER position depicts small flap (arrow) on joint-side area of tendon. Joint-side partial-thickness tear was correctly diagnosed by both readers.

 

Figure 5
View larger version (159K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5a: Intratendinous partial-thickness tear of the supraspinatus tendon in 45-year-old man. (a) Coronal oblique fat-saturated T1-weighted contrast-enhanced indirect MR arthrogram (1011/20) obtained with patient in the neutral position shows supraspinatus tendon (arrow) with normal signal intensity and without morphologic changes. (b) Fat-saturated T1-weighted contrast-enhanced indirect MR arthrogram (1011/20) obtained with patient in the ABER position demonstrates no surface abnormalities but depicts linear area of high signal intensity (arrow) within tendon and small space between fibers. Both readers correctly interpreted these signal intensity and morphologic alterations as indicative of an intratendinous partial-thickness tear.

 

Figure 5
View larger version (188K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5b: Intratendinous partial-thickness tear of the supraspinatus tendon in 45-year-old man. (a) Coronal oblique fat-saturated T1-weighted contrast-enhanced indirect MR arthrogram (1011/20) obtained with patient in the neutral position shows supraspinatus tendon (arrow) with normal signal intensity and without morphologic changes. (b) Fat-saturated T1-weighted contrast-enhanced indirect MR arthrogram (1011/20) obtained with patient in the ABER position demonstrates no surface abnormalities but depicts linear area of high signal intensity (arrow) within tendon and small space between fibers. Both readers correctly interpreted these signal intensity and morphologic alterations as indicative of an intratendinous partial-thickness tear.

 
For images obtained with patients in the neutral position (set 1), both readers classified two tears as partial-thickness tears; these tears were diagnosed as no tear at arthroscopy (false-positive findings). Supplementary images obtained with patients in the ABER position helped reduce the number of false-positive diagnoses. Hence, with set 2, sensitivity and specificity for the detection of a partial-thickness tear were 93% and 100%, respectively, for reader 1 and 86% and 94%, respectively, for reader 2.

For partial-thickness tears, the 95% confidence intervals did not overlap; thus, the additional information derived with the use of the ABER position significantly improved sensitivity in the detection of a partial-thickness tear of the supraspinatus tendon. For partial-thickness tears, specificity also increased from 88% to 94%–100% after the reading of set 2 (Table 2). This difference, however, was not statistically significant.

Diagnostic Confidence
Independent of the image set, diagnostic confidence was lower for the diagnosis of partial tears than for the diagnosis of no tear (Table 3). As for the diagnosis of no tear by both readers, no difference was seen in confidence score by using indirect MR arthrography with or without patients in the ABER position (P = .157 and P = .317, respectively). For the detection of a full-thickness tear, a higher level of confidence was achieved by using images obtained with patients in the ABER position for reader 1 only (P = .046). With regard to the diagnosis of a partial-thickness tear, the confidence level was significantly higher for both readers when evaluating indirect MR arthrograms obtained with patients in the ABER position (P = .034 and P = .046 for readers 1 and 2, respectively).


View this table:
[in this window]
[in a new window]

  		Table 3. Diagnostic Confidence for Tears of the Supraspinatus Tendon

 
Interobserver Agreement
For all diagnoses, interobserver agreement was good ({kappa} = 0.73) after the interpretation of set 1 and excellent ({kappa} = 0.94) after the interpretation of set 2. For full-thickness tears, agreement between the two readers was fair ({kappa} = 0.35) after the interpretation of set 1 and excellent ({kappa} = 0.94) after the interpretation of set 2. For partial-thickness tears, interobserver agreement increased from poor ({kappa} = 0.12) after the interpretation of set 1 and good ({kappa} = 0.63) after the interpretation of set 2.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
 References
 
The differentiation between full- and partial-thickness tears of the supraspinatus tendon is of great clinical importance because treatment options can be different (2,4–6,23,33). Previously published data for the detection of full-thickness tears (8–10,12,17,19,20,25,28,34,35) show good diagnostic accuracy for conventional MR imaging, as well as for direct and indirect MR arthrography. Reported sensitivities and specificities ranged from 81%–100% and 77%–100%, respectively. In our study, indirect MR arthrography with patients in the neutral position showed comparable results to the published data, with sensitivities of 80%–95% and a specificity of 100% for the presence of a full-thickness tear. The indirect technique is a noninvasive procedure that also enhances the joint cavity and offers logistical advantages because fluoroscopic guidance is not required (13).

With indirect MR arthrography, contrast enhancement is not affected by compartment anatomy, but every joint space is enhanced. Consequently, contrast material in the subacromial-subdeltoid bursa cannot be considered an indirect sign of a full-thickness tear as it is during direct MR arthrography (13). In our study, this did not create a disadvantage for assessing large full-thickness tears of the supraspinatus tendon because morphologic criteria were obvious in all cases. According to the literature (10,13,28,35), small full-thickness tears do not always show fluid in the subacromial-subdeltoid bursa because the edges are filled with synovia or fibrous tissue, which indicates that this sign is not reliable for small tears.

In our study, reader 2 showed a high rate of false-negative findings for small full-thickness tears, classifying them as partial-thickness tears. There is a continuous spectrum from a high-grade partial-thickness tear to a small complete tear (5). This fact made diagnosis difficult when patients were in the neutral position, but the interpretation of additional images obtained with patients in the ABER position reduced the rate of false-negative findings. Because the 95% confidence intervals overlapped, the improvement in sensitivity from 80% to 100% was not statistically significant. Nevertheless, for reader 2, diagnostic confidence for full-thickness tears increased significantly after reading images obtained with patients in the ABER position.

Interobserver agreement for full-thickness tears was fair by using indirect MR arthrography with patients in the neutral position. This was mainly caused the high rate of false-negative findings for reader 2. In set 2, agreement between the readers was excellent due to improvement in the diagnostic accuracy of reader 2. We believe that images obtained with patients in the ABER position enable readers to make a correct diagnosis fairly independent of experience level after a brief learning period. Our data help confirm that additional images with patients in the ABER position are useful for making a reliable diagnosis of a full-thickness tear.

The diagnosis of a partial-thickness tear is clearly more difficult than that of a full-thickness tear, and clinical tests, US, and unenhanced MR imaging have been shown to be inaccurate for preoperative diagnosis (2,4,21,28,34–39). Although intraarticular injection of gadopentetate dimeglumine has improved the diagnostic accuracy of partial-thickness tears, it still does not enable a reliable diagnosis in all cases (7–9,12). Reported sensitivities showed a range of between 38% and 100%, with specificities of between 84% and 100% (8–10,12). In our study, we calculated sensitivities of 50%–71% for indirect MR arthrography with patients in the neutral position. In contrast to previously published data that had higher sensitivities, our data included bursal-side partial-thickness tears because, in all our patients, both surfaces of the rotator cuff were inspected during arthroscopy.

One advantage of the indirect technique is that the subacromial-subdeltoid bursa is always enhanced. This makes it easier to delineate bursal-side partial tears (3,17). Nevertheless, three of four bursal-side partial tears were underestimated (one by reader 1 and two by reader 2) on images obtained with patients in the neutral position as a result of additional bursitis. Two intratendinous partial-thickness tears could not be adequately detected during indirect MR arthrography with patients in the neutral position, and both were assessed as no tear.

Tirman et al (21) first presumed that direct MR arthrography with patients in the ABER position would help improve the diagnosis of partial tears. Their findings were confirmed in a retrospective analysis by Roger et al (22), who used the direct technique and reported an increase in sensitivity from 83% to 100% after reading images obtained with patients in the ABER position. Our data also show that indirect MR arthrography performed with patients in the ABER position in addition to the neutral position significantly improved sensitivity in the diagnosis of a partial-thickness tear (from 50%–71% to 86%–93%).

One reason why the ABER position increases sensitivity is because it decreases tension on the posterior portion of the capsule and rotator cuff. In this way, tendons are kinked, and it is possible to depict joint-side undersurface tears without effacement by the humeral head (21,22). Moreover, the spreading of fibers enables intravasation of the enhanced fluid, and, combined with bursal enhancement of the indirect technique, the ABER view also helps improve the depiction of bursal-side partial-thickness tears of the supraspinatus tendon. We agree with Lee and Lee (23), who stated that all these mechanisms are responsible for better delineation of horizontal components of a partial tears, as well as of intratendinous tears that have no abnormality along the surface of the tendon.

The main problem with indirect MR arthrography is the correct interpretation of an area of increased signal intensity within the tendon, which can be caused by tendinopathy or enhancement of fibrovascular tissue in a partial tear. Additional morphologic criteria—namely, a surface abnormality or element interruption—are not reliable (14,16,19,20). In our study, the number of partial tears that was overestimated with indirect MR arthrography was negligible. Both readers interpreted two large foci of tendinopathy as partial-thickness tears. Again, this rate could be further reduced by using the ABER position, which increased specificity from 88% to 94%–100%. The overlapping of the 95% confidence intervals, however, suggests that such differences were only indicative of a tendency and were not statistically significant.

For partial-thickness tears, the agreement between the two readers was poor ({kappa} = 0.12) for indirect MR arthrography with patients in the neutral position. This was mainly caused by the fact that both readers underestimated different partial-thickness tears as no tear and agreed on only six diagnoses for partial-thickness tears. To our knowledge, no data exist with regard to interobserver agreement in the diagnosis of partial-thickness tears with use of MR arthrography. Nonetheless, our {kappa} values for set 1 were similar to those reported in two previously published retrospective analyses on unenhanced MR sequences (34,35). In the studies of Robertson et al (34) and Balich et al (35), four and five readers, respectively, evaluated unenhanced MR images of symptomatic patients with arthroscopic correlation, and the {kappa} values for partial-thickness tears ranged from 0.13 to 0.44. In both series, Cohen {kappa} values showed a positive correlation with the reader's experience. In our group, we calculated good interobserver agreement ({kappa} = 0.63) after the interpretation of images obtained with patients in the ABER position, which suggests that the reading of additional images obtained with use of this position leads to more consistent results.

A potential limitation of our study is selection bias because we included only those patients who later underwent surgery. This fact could help explain the high number of pathologic findings revealed during arthroscopy. Another selection bias could have resulted from the exclusion of patients who were unable to keep their arm in the ABER position. The position, however, was tolerated in 51 of 60 patients, including the elderly. There were also limitations due to the delay between arthroscopy and MR imaging (range, 25–175 days; mean, 37 days). It is possible that some disease progression may have occurred during this time, resulting in an artificially reduced sensitivity.

In our study, we focused on the evaluation of tears of the supraspinatus tendon. Use of the ABER position, however, helps display the infraspinatus tendon in a more cross-sectional plane and thus enables the separate assessment of the supraspinatus and infraspinatus tendon surfaces (21,23). Further studies are necessary to evaluate the benefits of the ABER position for the detection and characterization of abnormalities of the infraspinatus tendon.

In conclusion, indirect MR arthrography performed with patients in the neutral position and supplementary ABER position significantly improved sensitivity and increased confidence in the diagnosis of partial-thickness tears of the supraspinatus tendon. Interobserver agreement also increased and was excellent for full-thickness tears and good for partial-thickness tears.


   ADVANCES IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
References
 


   FOOTNOTES
 

Abbreviations: ABER = abduction and external rotation

Authors stated no financial relationship to disclose.

Author contributions: Guarantor of integrity of entire study, T.H.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, T.H., O.W.H., C.F., M.L.; clinical studies, T.H., M.B., O.W.H., R.H., S.F., C.F., M.L., C.P.; statistical analysis, T.H., M.B., R.H., S.F., C.F., M.S., M.L., C.P.; and manuscript editing, T.H., O.W.H., R.H., S.F., C.F., M.S., M.L., C.P.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 

  1. Uri DS. MR imaging of shoulder impingement and rotator cuff disease. Radiol Clin North Am 1997;35(1):77–96.[Medline]
  2. Fukuda H, Hmada K, Nakajima T, Yamada N, Tomonaga A, Goto M. Partial-thickness tears of the rotator cuff: a clinicopathological review based on 66 surgically verified cases. Int Orthop 1996;20:257–265.[CrossRef][Medline]
  3. Schroder RJ, Bostanjoglo M, Kaeaeb M, et al. Accuracy of routine MRI in lesions of the supraspinatus tendon: comparison with surgical findings [in German]. Rofo 2003;175:920–928.[Medline]
  4. Fukuda H. The management of partial-thickness tears of the rotator cuff. J Bone Joint Surg Br 2003;85:3–11.
  5. Snyder SJ, Pachelli AF, Pizzo WD, Friedman MJ, Ferkel RD, Pattee G. Partial thickness rotator cuff tears: results of arthroscopic treatment. Arthroscopy 1991;7:1–7.[Medline]
  6. Waibl B, Buess E. Partial-thickness articular surface supraspinatus tears: a new transtendon suture technique. Arthroscopy 2005;21(3):376–381.[Medline]
  7. Flannigan B, Kursunoglu-Brahme S, Snyder S, Karzel R, Del Pizzo W, Resnick D. MR arthrography of the shoulder: comparison with conventional MR imaging. AJR Am J Roentgenol 1990;155:829–832.[Abstract/Free Full Text]
  8. Hodler J, Kursunoglu-Brahme S, Snyder SJ, et al. Rotator cuff disease: assessment with MR arthrography versus standard MR imaging in 36 patients with arthroscopic confirmation. Radiology 1992;182:431–436.[Abstract/Free Full Text]
  9. Palmer WE, Brown JH, Rosenthal DI. Rotator cuff: evaluation with fat-suppressed MR arthrography. Radiology 1993;188:683–687.[Abstract/Free Full Text]
  10. Quinn SF, Sheley RC, Demlow TA, Szumowski J. Rotator cuff tendon tears: evaluation with fat-suppressed MR imaging with arthroscopic correlation in 100 patients. Radiology 1995;195:497–501.[Abstract/Free Full Text]
  11. Magee T, Williams D, Mani N. Shoulder MR arthrography. AJR Am J Roentgenol 2004;183:969–974.[Abstract/Free Full Text]
  12. Meister K, Thesing J, Montgomery WJ, Indelicato PA, Walczak S, Fontenot W. MR arthrography of partial thickness tears of the undersurface of the rotator cuff: an arthroscopic correlation. Skeletal Radiol 2004;33:136–141.[CrossRef][Medline]
  13. Steinbach LS, Palmer WE, Schweitzer ME. MR arthrography. RadioGraphics 2002;22:1223–1246.[Abstract/Free Full Text]
  14. Winalski CS, Aliabadi P, Wright RJ, Shortkroff S, Sledge CB, Weissman BN. Enhancement of joint fluid with intravenously administered gadopentetate dimeglumine: technique, rationale, and implications. Radiology 1993;187:179–185.[Abstract/Free Full Text]
  15. Drape JL, Thelen P, Gay-Depassier P, Silbermann O, Benacerraf R. Intraarticular diffusion of Gd-DOTA after intravenous injection in the knee: MR imaging evaluation. Radiology 1993;188:227–234.[Abstract/Free Full Text]
  16. Vahlensieck M, Peterfy CG, Wischer T, et al. Indirect MR arthrography: optimization and clinical applications. Radiology 1996;200(1):249–254.[Abstract/Free Full Text]
  17. Vahlensieck M, Sommer T. Indirect MR arthrography of the shoulder. Radiologe 1996;36:960–965.[CrossRef][Medline]
  18. Allmann KH, Schaefer O, Hauer M, et al. Indirect MR arthrography of the unexercised glenohumeral joint in patients with rotator cuff tears. Invest Radiol 1999;34(6):435–440.[CrossRef][Medline]
  19. Rudolph J, Lorenz M, Schroder R, Sudkamp NP, Felix R, Maurer J. Indirect MR arthrography in the diagnosis of rotator cuff lesions. Rofo 2000;172:686–691.[Medline]
  20. Yagci B, Manisali M, Yilmaz E, et al. Indirect MR arthrography of the shoulder in detection of rotator cuff ruptures. Eur Radiol 2001;11:258–262.[CrossRef][Medline]
  21. Tirman PF, Bost FW, Steinbach LS, et al. MR arthrographic depiction of tears of the rotator cuff: benefit of abduction and external rotation of the arm. Radiology 1994;192:851–856.[Abstract/Free Full Text]
  22. Roger B, Skaf A, Lektrakul N, Yeh LR, Resnick D. Imaging findings in the dominant shoulder of throwing athletes: comparison of radiography, arthrography, CT arthrography, and MR arthrography with arthroscopic correlation. AJR Am J Roentgenol 1999;172:1371–1380.[Abstract/Free Full Text]
  23. Lee SY, Lee JK. Horizontal component of partial-thickness tears of the rotator cuff: imaging characteristics and comparison of ABER view with oblique coronal view at MR arthrography—initial results. Radiology 2002;224:470–476.[Abstract/Free Full Text]
  24. Wintzell G, Larsson H, Larsson S. Indirect MR arthrography of anterior shoulder instability in the ABER and the apprehension test position: a prospective comparative study of two different shoulder positions during MRI using intravenous gadodiamide contrast for enhancement of the fluid. Skeletal Radiol 1998;27:488–494.[CrossRef][Medline]
  25. Iannotti JP, Zlatkin MB, Esterhai JL, Kressel HY, Dalinka MK, Spindler KP. Magnetic resonance imaging of the shoulder: sensitivity, specificity, and predictive value. J Bone Joint Surg Am 1991;73:17–29.[Abstract/Free Full Text]
  26. Kjellin I, Ho CP, Cervilla V, et al. Alterations in the supraspinatus tendon at MR imaging: correlation with histopathologic findings in cadavers. Radiology 1991;181:837–841.[Abstract/Free Full Text]
  27. Vahlensieck M, Pollack M, Lang P, Grampp S, Genant HK. Two segments of the supraspinous muscle: cause of high signal intensity at MR imaging. Radiology 1993;186:449–454.[Abstract/Free Full Text]
  28. Reinus WR, Shady KL, Mirowitz SA, Totty WG. MR diagnosis of rotator cuff tears of the shoulder: value of using T2-weighted fat-saturated images. AJR Am J Roentgenol 1995;164:1451–1455.[Abstract/Free Full Text]
  29. Singson RD, Hoang T, Dan S, Friedman M. MR evaluation of rotator cuff pathology using T2-weighted fast spin-echo technique with and without fat suppression. AJR Am J Roentgenol 1996;166:1061–1065.[Abstract/Free Full Text]
  30. Vahlensieck M. MRI of the shoulder. Eur Radiol 2000;10:242–249.[CrossRef][Medline]
  31. Schweitzer ME, Magbalon MJ, Fenlin JM, Frieman BG, Ehrlich S, Epstein RE. Effusion criteria and clinical importance of glenohumeral joint fluid: MR imaging evaluation. Radiology 1995;194:821–824.[Abstract/Free Full Text]
  32. Seigel DG, Podgor MJ, Remaley NA. Acceptable values of kappa for comparison of two groups. Am J Epidemiol 1992;135:571–578.[Abstract/Free Full Text]
  33. Zanetti M, Jost B, Lustenberger A, Hodler J. Clinical impact of MR arthrography of the shoulder. Acta Radiol 1999;40:296–302.[Medline]
  34. Robertson PL, Schweitzer ME, Mitchell DG, et al. Rotator cuff disorders: interobserver and intraobserver variation in diagnosis with MR imaging. Radiology 1995;194:831–835.[Abstract/Free Full Text]
  35. Balich SM, Sheley RC, Brown TR, Sauser DD, Quinn SF. MR imaging of the rotator cuff tendon: interobserver agreement and analysis of interpretive errors. Radiology 1997;204:191–194.[Abstract/Free Full Text]
  36. Teefey SA, Rubin DA, Middleton WD, Hildebolt CF, Leibold RA, Yamaguchi K. Detection and quantification of rotator cuff tears: comparison of ultrasonographic, magnetic resonance imaging, and arthroscopic findings in seventy-one consecutive cases. J Bone Joint Surg Am 2004;86:708–716.[Abstract/Free Full Text]
  37. Strobel K, Zanetti M, Nagy L, Hodler J. Suspected rotator cuff lesions: tissue harmonic imaging versus conventional US of the shoulder. Radiology 2004;230:243–249.[Abstract/Free Full Text]
  38. Jacobson JA, Lancaster S, Prasad A, van Holsbeeck MT, Craig JG, Kolowich P. Full-thickness and partial-thickness supraspinatus tendon tears: value of US signs in diagnosis. Radiology 2004;230:234–242.[Abstract/Free Full Text]
  39. Traughber PD, Goodwin T. Shoulder MRI: arthroscopic correlation with emphasis on partial tears. J Comput Assist Tomogr 1992;16(1):129–133.[Medline]




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2393050453v1
240/1/152    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Herold, T.
Right arrow Articles by Paetzel, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Herold, T.
Right arrow Articles by Paetzel, C.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE