| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Service Radiologie, Polyclinique Atlantique, Rue Claude Bernard, BP419, Saint
Herblain, Cedex 44819, France.
2 Service Orthopedie, Polyclinique Atlantique, Saint Herblain, Cedex 44819,
France.
Received November 14, 2004;
accepted after revision March 23, 2005.
Address correspondence to O. P. Krief
(okrief{at}yahoo.fr).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. One thousand seventy-nine consecutive patients referred for shoulder MRI were asked to complete the L'Insalata SAQ. Results from the L'Insalata SAQ and MRI were cross-tabulated and analyzed with multivariable linear regression.
RESULTS. No statistical relationship could be found between the level of pain, impairment, and disability as reported on the L'Insalata SAQ and the location and size of full-thickness tears of the rotator cuff as observed on MRI. Pain and disability are significantly linked to the presence of supraspinatus tendon lesions and the presence of bursitis, but these factors contribute little to the symptoms. Patients with biceps tendinopathy did not experience increased pain when compared with patients without biceps tendinopathy or with biceps tendon rupture.
CONCLUSION. No statistical relationship was found between the level of pain and disability and the size and location of full-thickness tears of the rotator cuff.
Keywords: abnormalities • bursitis • MR arthrography • MRI • musculoskeletal imaging • rotator cuff • shoulder • sports medicine • surgery
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Finding a correlation between symptoms and images is a challenging task and is essential to ensure that the imaging findings explain the symptoms and can be used to adjust the therapy, especially if the option of surgery is retained. Previous imaging or surgical studies have shown the high prevalence of rotator cuff tears in an asymptomatic population [6-11]. Therefore, misuse or misinterpretation of clinical data without imaging or of imaging without clinical data may not lead to accurate management of patients with rotator cuff disease [6].
The purpose of our study was to compare the level of shoulder pain and disability at rest and in everyday activities indicated by the L'Insalata Self-Administered Questionnaire (SAQ) with the location and extent of rotator cuff lesions observed on MRI.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
L'Insalata SAQ
On arrival at the MR unit, all patients referred for shoulder MRI were
asked to complete the L'Insalata SAQ and a consent form for the examination
and study. The L'Insalata SAQ can be used to assess symptoms in different
domains of pain, daily activities, work, satisfaction, and areas for
improvement; 18 questions are scored from 1 to 5, two questions discuss the
general level of satisfaction and areas in which the patient wishes to
improve, and the global assessment of impairment is assessed on a visual
analog scale [12]. Four of the
questions refer directly to pain (pain at rest, during activities, and at
night; and the frequency of severe pain during the previous months). The next
six questions concern the patient's ability to perform personal or household
activities (global disability; limitations in putting on and removing a
sweater, combing hair, reaching overhead shelves, washing lower back, and
lifting a bag of groceries). The next three questions relate to recreational
and athletic activities (overall limitation, limitation in throwing a ball,
and limitations in the most regularly practiced sport). The next five
questions refer to work (work status and disability induced by the shoulder
pain or weakness, frequency of inability to work because of shoulder pain,
impairment at work linked to shoulder pain, frequency of reduced workday due
to shoulder pain). For each patient the total score of the 18 questions was
recorded, the mean value for each domain of pain and disability was entered on
a scale from zero to 10, and the overall disability was evaluated on the
visual analog scale from zero to 10.
The mean time for a patient to complete the SAQ was less than 5 min and never exceeded 10 min. When necessary, a secretary assisted a patient in completing the SAQ and checked for omitted fields. The results of the SAQ, age, sex, dominant arm, occupation, worker's compensation status, and MRI findings were cross-tabulated for statistical analysis.
MR Protocol
Nine hundred thirty-nine examinations were performed using an indirect
arthrographic method with an IV injection of 15 mL of gadopentetate
dimeglumine (Omniscan, Amersham) 15 min before the examination with gentle
active mobilization of the upper limb.
Direct MR arthrography was performed in 140 patients: 26 patients for suspicion of labral tear; 18 patients for postsurgical examination, five of whom had undergone rotator cuff repair; 57 patients with a discrepancy between the previous MR findings and the clinical findings, 42 of whom had undergone a previous MR examination at another institution; and 17 patients for suspicion of posterosuperior glenoid impingement. The remainder of the MR arthrography examinations was performed due to a specific professional or sports-related request. MR arthrography was performed with injection under fluoroscopic control of 12 mL of a gadopentetate dimeglumine preparation for intraarticular use (25 µmol/mL, Artirem, Guerbet) mixed with 1 mL of 1% lidocaine and iodinated contrast material.
All shoulder MR scans were obtained on a 1.5-T unit (Intera, Philips Medical Systems) using a 3-inch (8-cm) surface receive-only coil. Patients were placed in a slight lateral decubitus position with the examined shoulder and scapula horizontal on the MRI table to minimize motion artifacts and the arm by their side in a neutral position or in mild external rotation.
The MRI scanning protocol was identical for direct and indirect arthrography and included the following: oblique coronal, sagittal, and axial fat-suppressed proton density-weighted sequences (TR/TE, 1,500/25; field of view, 120 mm; 4-mm-thick slices with 0.4-mm gap and 320 x 512 matrix; 4 excitations); and coronal and sagittal T2-weighted fast spin-echo sequences (1,800/100; field of view, 100 mm; 4-mm-thick slices with 0.4-mm intersection gap and 354 x 512 matrix; 4 excitations). The total scanning time was less than 15 min.
MR Analysis
All the MR examinations were prospectively analyzed by the same
musculoskeletal radiologist. The radiologist was not blinded to the L'Insalata
SAQ results. Each rotator cuff tendon was evaluated according to a grading
system adapted for imaging from Ellman's arthroscopic classification system
[13]. This classification
system indicates which surface of the tendon is involved and grades the
severity of the tear according to its depth: grade 0, homogeneous signal and
regular margins of tendon; grade 1, articular or bursal surface side lesion
involving less than a quarter of the tendon thickness; grade 2,
partial-thickness tear involving less than half the thickness of the tendon;
and grade 3, partial-thickness tear involving more than one half of the
thickness of the tendon with tenuous continuity but without full-thickness
tear. In the presence of a full-thickness tear with fluid signal intensity
extending from the bursal to the articular side lesion of the rotator cuff,
the size of the tear was assessed in the coronal and sagittal planes except
isolated full-thickness tears of the subscapularis tendon, which were measured
in the coronal and axial planes. When full-thickness tears were associated
with trophic alterations of the proximal tendon or with having high and
irregular intensity of the tendon stump, both the size of the full-thickness
tear and the size of the whole tendon with the abnormalities were
measured.
The biceps tendon lesions were classified as grade zero when the tendon had a normal shape and signal in all planes, grade 1 when shape abnormalities or signal abnormalities without discontinuity were present, and grade 2 when a discontinuity of the tendon was observed.
Statistical Analysis
The statistical analysis was performed with software (version 2.03
SigmaStat, Systat Software Inc.) using a multivariable linear regression and
backward stepwise method. The SigmaStat software allows the use of a specific
missing value code to handle the presence of eventual missing data on an SAQ.
The dependent variable for the regression model was the L'Insalata SAQ
results, with the overall and mean scores for each domain of pain and
disability (pain, disability in everyday life, disability in work and sports
activities) and the mean score for disability as assessed on the visual analog
scale. The independent variables included in the model were patient age, sex,
occupation, affected shoulder side, and dominant arm side; existence of a work
compensation claim; and MR results. The MR variables included in the model
were the Ellman stage of the tendon, from zero to 3, for each tendon of the
rotator cuff; the sizes of the full-thickness tears in the sagittal and
coronal planes; the stage, from zero to 2, of the biceps tendon; and the
presence of subacromial bursitis. In each step, the normality test was
assessed and the variable with the smallest contribution to the model was
removed as long as the p value was greater than 0.05. For each
regression analysis and for each domain, the p values, the
R2 and adjusted R2, the standard error
of the estimate (SEE), and the residual were reported. The power for each test
was assessed with an alpha value of 0.05.
In the first step, the statistical analysis was performed for all patients. Then, the regression analysis was performed after having excluded patients with previous surgery or patients with an abnormality or disease other than rotator cuff tendinopathy that could interfere with rotator cuff symptoms, even when concomitant rotator cuff tendinopathy was present (Table 1). To avoid a selection bias that could arise from excluding patients with abnormalities or diseases other than rotator cuff disease, the statistical analysis was performed for both populations. Two hundred nineteen patients were excluded from this second statistical analysis: 13 patients with previous subacromial decompression, five with associated surgical repair of the rotator cuff, 31 with calcifications greater than 5 mm, 87 with clinical or MRI findings suggestive of capsulitis (painful and restricted passive and active range of motion or capsular thickening > 5 mm with enhancement on MRI), nine with glenohumeral instability, 29 with important signs of acromioclavicular osteoarthritis (painful joint at palpation and positive cross-arm test), 20 with posttraumatic bone bruises, 19 with chondral glenohumeral lesions observed on MRI, three with signs of neoplastic diseases (one case of multiple myeloma, one metastasis, and one aneurysmal bone cyst), two with avascular necrosis of the humeral head, and one patient with scapular winging linked to long thoracic nerve palsy. Patients with mild acromioclavicular osteoarthritis associated with rotator cuff lesions were not excluded.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MR Results
The MR results are summarized in Table
2. In 389 patients, a normal rotator cuff was displayed on MRI,
with regular smooth border tendons and homogeneous low signal intensity on all
sequences. Four hundred fifty-one patients presented with a partial-thickness
tear, including 286 grade 1 tears on the Ellman classification system, 115
grade 2 tears, and 50 grade 3 tears.
|
Two hundred thirty-nine patients displayed full-thickness tears of the rotator cuff on MRI. All but 17 tears involved the supraspinatus tendon. Of those 17 remaining tears, 16 involved the subscapularis tendon and one, the infraspinatus tendon in a patient with a history of previous calcification.
One hundred twenty patients displayed an isolated tear of the supraspinatus tendon; 60 patients had tears involving both the supraspinatus and infraspinatus tendons; 13 patients had tears involving the supraspinatus and subscapularis tendons; and 29 patients had tears involving the supraspinatus, infraspinatus, and subscapularis tendons.
The mean size of the rotator cuff tear was 18 mm, with an SD of 11 mm, and the largest tear observed measured 66 mm.
Statistical Analysis
The dependent variable could be predicted from a linear combination of the
independent variables before and after exclusion of patients with a history of
surgery or with abnormalities or diseases other than rotator cuff
tendinopathy. The global disability was statistically linked to
partial-thickness tears involving the superficial and deep surfaces of the
supraspinatus tendon (p < 0.01), to the presence of bursitis
(p = 0.01), and to the age of the patient and the side of the
dominant arm (p = 0.04) (Table
3). Pain and disability in all domains were positively correlated
to age (p = 0.01), despite lower demand with aging. Although
significant results were obtained for the presence of supraspinatus tendon
lesions and bursitis, the contribution of these factors to the symptoms was
low, with an R2 of 0.350 and adjusted
R2 of 0.337 even with a limited number of independent
variables. This means that the contribution of the selected variables to the
model was about one third and that there must be other factors not included in
the study that could explain the symptoms.
|
The global disability was not statistically linked to any abnormality of the subscapularis (p = 0.25) or infraspinatus (p = 0.44) tendon, biceps tendinopathy or rupture (p = 0.5), or patient sex or work status (p = 0.1). The presence, size, and location of full-thickness tears of the rotator cuff did not influence the level of disability (p = 0.1) or the level of pain (p = 0.09). Patients with biceps tendinopathy in the presence of a rotator cuff tear were not associated with increased pain or disability. Patients with a discontinuous biceps tendon in the presence of a rotator cuff tear did not present with a lower level of pain than patients with biceps tendinopathy without rupture. Figure 1 displays a scatterplot graph for overall disability versus maximal size of the rotator cuff tears, and Figure 2 shows a scatterplot graph for the results of the L'Insalata SAQ in the pain domain (average level of pain at rest, during effort, and at night; and maximal pain) versus maximal size of the rotator cuff tears.
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
Our study confirms previous results regarding the discrepancy between MRI findings and symptoms related to rotator cuff disease. Our statistical analysis did not reveal any relationships between the location or the size of rotator cuff tears and the intensity of pain and the level of disability perceived by the patient. In their MRI study, Miniaci et al. [10] found rotator cuff lesions to be present in asymptomatic professional throwing athletes without any decrease in the level of their performances. In their sonography studies, Milgrom et al. [8] found a prevalence of rotator cuff tears of more than 50% in patients older than 70 years, and Tempelhof et al. [11] found that more than 20% of asymptomatic volunteers above the age of 80 years presented with rotator cuff tears larger than 5 cm. Both groups of researchers concluded that rotator cuff lesions may be regarded as a natural correlate of aging. It is more likely to find a rotator cuff abnormality on an MR examination in an asymptomatic elderly population than in young patients with a painful shoulder.
It is not known what makes some tears symptomatic and others asymptomatic [24]. Many tears do not interfere with normal function [7, 24, 25] (Figs. 3A and 3B). Yamaguchi et al. [26] did not find any relationship between the evolution of rotator cuff size and the occurrence of symptoms in a longitudinal sonography follow-up of initially asymptomatic rotator cuff tears. Some of the tears may even become silent with increasing size [26], whereas initially silent tears may become symptomatic, which makes the surgical decision complex [27]. Goodman et al. [6] and Sher et al. [7] concluded that the surgical decision should not rely only on imaging. The age of the patient is thus an essential element in the therapeutic decision: Orthopedic surgeons suggest repairing the rotator cuff in young patients even when the level of disability remains moderate because of their unpredictable evolution.
From a fluoroscopic study, Burkhart [25] states that normal function is possible as long as the posterior aspect of the rotator cuff is preserved to such a degree that the force couple in the transverse plane is maintained. A rindlike margin resistant to tearing limits the extension of the tear and acts as a suspension bridge, thereby limiting its biomechanical consequences. Burkhart's theory thus assumes that there is a threshold effect, with small painless lesions, and that pain and disability appear only once the rotator cable has been ruptured. Burkhart's theory could explain why some full-thickness tears may present with a negative Jobe test, but we did not find any threshold effect for symptoms relating to the tear size. Some small rim-rent tears may be painful, whereas some large tears, completely painless and even compatible with normal and painless functional activity (Fig. 4). In a review of 80 patient files, Patte and Goutallier [28] found that anterior rotator cuff lesions were more frequently revealed by pain and posterior rotator cuff lesions, by weakness. We did not find any relationship between the type of symptom and the location of rotator cuff lesions as observed on MRI. Patients with full-thickness tears involving the infraspinatus tendon did not display any increased disability compared with patients with full-thickness tears sparing the infraspinatus tendon.
Several factors other than rotator cuff lesions may be involved in shoulder pain and disability. Bursitis is frequently associated with rotator cuff lesions, and we found a significant increase in shoulder pain and perception of disability with patients who presented with bursitis. Pain may also be related to the presence of labral capsular ligamentous complex or cartilaginous lesions. Labral lesions are often secondary to glenohumeral joint instability [29], but may be also age-dependant degenerative lesions in conjunction with glenohumeral chondropathy [30]. Cartilage lesions have been reported to be present on arthroscopy in up to 29% of patients referred for subacromial impingement, and these lesions are underestimated on MRI [31]. Occult labral or cartilage lesions could partially explain the discrepancy between pain and disability and the importance of rotator cuff lesions observed in our study. Lastly, there may simply be no relationship between rotator cuff tear size and the inflammatory reaction responsible for the pain and the disability, like lower back pain intensity is unrelated to the size of disk herniations. This could explain the fluctuant character of shoulder disability due to incremental progression of cuff tears, with possible increasing tear size while the pain may regress.
These results raise the problem of the place for imaging in the management of rotator cuff lesions. Because there is no clear relationship between symptoms and lesions, some surgical teams suggest that presurgical imaging may not in any way affect the management of rotator cuff lesions [32]. However, preoperative knowledge of the size and shape of the full-thickness rotator cuff tear is important for patient counseling because tear size affects the choice of surgery and its functional outcome [4, 5, 33] and it is not possible to assess the size and the precise position of the ends of the remaining tendons on the basis of the clinical examination [3]. Moreover, fatty degeneration of the muscles of the rotator cuff assessed on MRI may not be depicted on clinical examination and may not be reversed by surgery regardless of the tear size. Imaging should therefore be included in the preoperative assessment.
Our study has several limitations. There is a selection bias for patients referred for shoulder MRI. Neither patients who respond favorably to nonoperative therapy nor those with straightforward clinical diagnosis of very large tears obviating MRI or with decreased acromiohumeral distance in the case of large tears are usually referred for MRI and, therefore, are not candidates for surgical repair of the rotator cuff. Moreover, the time lapse between the onset of symptoms and the MRI examination depended on the appointment delay for MRI, which is approximately 4-6 weeks in our unit. Any change in the length of this delay would certainly affect the results of the SAQ at the time of the examination. Therefore, the results of the L'Insalata SAQ and the comparison between symptoms and images could not fit the overall population. Another limitation is the absence of a gold standard for the results of the MRI examinations; however, to avoid selection bias, we did not include the surgical results even for the patients who did benefit from a surgical program. Finally, we did not include the glenoid labral lesions as a variable in our statistical analysis because the different clinical tests and L'Insalata SAQ are inappropriate for the assessment of disability linked to glenohumeral joint instability, and only a few of our patients underwent MR arthrography.
In conclusion, the level of disability is significantly linked to the presence of supraspinatus tendon lesions or bursitis, but the contribution of these factors to the presence of symptoms remains low. There is no clear statistical relationship between the level of pain and disability and the location and extent of full-thickness tears of the rotator cuff as observed on MRI. This discrepancy could be related to the secondary inflammatory reaction, which may not be proportional to the size of the rotator cuff lesions, thus explaining the fluctuant character of pain and disability. Despite the absence of correlation between the size of the rotator cuff tears and the level of disability, MRI provides important data that may affect the management of rotator cuff lesions and should be performed before rehabilitation or surgery.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
