Compiled By: Matthew P. Chanin, M.D.
History: Fifty One year old Male with right shoulder pain, bursitis, and decreased range of motion for 5 months without known injury. No surgical or cancer history, no diabetes or rheumatoid arthritis.
Exam: Oblique Coronal T1 FSE (TR/TE: 550/9.864 ms), T2 FSE (TR/TE: 4400/.864 ms), T2 FAT SAT (TR/TE: 4550/87.104 ms); Oblique Sagittal T1 SE (TR/TE: 616.664/8.848 ms), T2 FAT SAT (TR/TE: 4650/89.536 ms); Axial T1 SE (TR/TE: 400/11 ms), T2* GRE (TR/TE: 517/22 flip angle 20 degrees). Acquisition matrix 512 x 512 with 3 mm slice thickness and 4 mm separation on axial sequences, 4 mm thickness and separation for oblique sagittal and coronal sequences. GE Genesis Signa 1.5T.
Findings: Intermediate T1 and T2 signal within a non-distended axillary recess with T2 signal less than muscle, but greater than simple fluid. Indurated inferior glenohumoral ligament and peri-capsular soft tissues. Thickened coraco-humeral ligament.
Diagnosis: Adhesive capsulitis.
Discussion: Adhesive capsulitis or "frozen shoulder" is a syndrome characterized by nonspecific shoulder pain and limitation of motion, typically affecting the non-dominant shoulder of women aged 40 to 70 years and may be idiopathic, preceded by trauma, or associated with conditions such as diabetes mellitus, Dupuytren disease, or cardiac surgery (1, 4, 5, 6).
The term frozen shoulder was used in 1934 by Codman, who described a clinical syndrome of slow pain onset, inability to sleep on the affected arm, and restriction of both active and passive elevation and external rotation (3). In 1945 Neviaser coined adhesive capsulitis and described the pathology, and later arthrographic findings including small joint capsule volume (5-10ml vs. normal volume of 16-35ml), resistance to instillation of contrast, absent axillary and subscapularis recesses, and poor filling of the biceps tendon sheath (2).
MRI's emergence as an important component in the diagnostic algorithm for evaluation of the painful shoulder is well documented. Familiarity with the MRI presentation of adhesive capsulitis aids in its differentiation from other pathologies in the differential diagosis, including rotator cuff tear, labral injuries, biceps pulley injuries, brachial plexus pathology, entrapment neuropathy, arthritis, and neoplasia.
In 1995 Emig et. al. concluded that capsular and synovial thickness of greater than 4 mm can suggest the diagnosis with reasonable sensitivity and excellent specificity, while articular volume, coracohumeral ligament thickness, and pathologic findings in the rotator interval were not significantly different between diseased and control shoulders (7). More recently Mengiardi et. al. found obliteration of the subcoracoid fat triangle and thickening of the coacohumeral ligament and capsule at the rotator cuff interval as most reliable, however with low specificity. They also confirmed that those with adhesive capsulitis have a low volume joint capacity, though the etiology of this finding remains indeterminate and can be complicated by variability in the success of arthrographic technique. Inferior glenohumeral ligament thickening was not confirmed as a reliable finding during their MR arthrographic study (8).
Most shoulder MRI is performed without direct or indirect arthrographic technique. Therefore findings most suggestive of adhesive capsulitis, with acceptable sensitivity, include intermediate T1 and T2 signal intensity within a low volume axillary recess, though this consideration must fit the clinical history and patient demographics. Helpful confirmatory findings include a peri-articular soft tissue "glow" best appreciated on oblique sagittal water weighted sequences and "squeezing" of articular fluid into the subcoracoid recess and peri-bicipital sheath, in the absence of axillary recess distension.
Images :
Oblique Coronal T2 FSE (TR/TE: 4400/87.936 ms). Intermediate signal within thickened though non-distended axillary recess.
