Contracture of the Deltoid Muscle: Sonographic Evaluation with MRI Correlation

Transcription

1 Sonography and MRI of Deltoid Muscle Contracture Musculoskeletal Imaging Original Research Chung-Cheng Huang 1 Sheung-Fat Ko 1 Jih-Yang Ko 2 Hsuan-Ying Huang 3 Shu-Hang Ng 4 Yung-Liang Wan 4 Min-Chi Chen 5 Yu-Fan Cheng 1 Tze-Yu Lee 1 Huang C-C, Ko S-F, Ko J-Y, et al. Received June 8, 2004; accepted after revision October 15, This work was supported in part by a grant to Chung-Cheng Huang and Sheung-Fat Ko by the Chang Gung University (CMRPG 83052). 1 Department of Radiology, Chang Gung Memorial Hospital at Kaohsiung, Chang Gung University, 123 Ta-Pei Rd., Niao-Sung Hsiang, Kaohsiung Hsien 833, Taiwan. Address correspondence to S-F Ko (sfatko@adm.cgmh.org.tw). 2 Department of Orthopedic Surgery, Chang Gung Memorial Hospital at Kaohsiung, Chang Gung University, Kaohsiung Hsien, Taiwan. 3 Department of Pathology, Chang Gung Memorial Hospital at Kaohsiung, Chang Gung University, Kaohsiung Hsien, Taiwan. 4 Department of Radiology, Chang Gung Memorial Hospital at Linkou, Chang Gung University, Taoyuen Hsien, Taiwan. 5 Department of Public Health and Biostatistics, Chang Gung Memorial Hospital at Kaohsiung, Chang Gung University, Kaohsiung Hsien, Taiwan. AJR 2005; 185: X/05/ American Roentgen Ray Society Contracture of the Deltoid Muscle: Sonographic Evaluation with MRI Correlation OBJECTIVE. This article evaluates the sonographic features of deltoid contracture (DC) with MRI correlation. MATERIAL AND METHODS. Two reviewers evaluated the imaging features in 22 painful shoulders of 20 patients with a sonographic diagnosis of DC and a subsequent confirming MRI study. The sonographic and MRI findings with regard to the lesion extent (assessed by a 3-point scale: 1 = less than or equal to one third of the longitudinal deltoid length involved, 2 = greater than one third and less than or equal to two thirds involved, and 3 = greater than two thirds involved), transverse lesion morphologic appearance, and maximal transverse diameter measured were compared with kappa statistics and Wilcoxon s signed rank test, respectively. RESULTS. Compared with MRI, there were two false-positive diagnoses of DC on sonography. Among the 20 true-positive diagnoses, sonography showed good agreement with MRI in assessing the lesion extent (kappa = 0.796, p < 0.001). Three sonographic lesion morphologic patterns for hyperechoic lesions (I = with multiple < 8-mm hypoechoic spots, II = heteroechoic lesions with predominant 8 15-mm hypointense areas, and III = > 15-mm calcified nodules, respectively) showed excellent agreement with three MRI lesion patterns (I = multiple < 8-mm hypointense spots, II = predominant 8 15-mm hypointense areas, and III = > 15-mm hypointense nodules, respectively) (kappa = 0.921, p < 0.001). However, the maximum lesion diameters appeared significantly larger on sonography than on MRI (2.8 ± 0.6 cm vs 2.0 ± 0.8 cm, mean ± SD; p < 0.001), which was plausibly ascribed to the better sonographic delineation of hyperechoic immature fibrotic tissues. CONCLUSION. Sonography is helpful for evaluating DC and correlates well with MRI. ontracture of the deltoid muscle C usually is induced by repeated intramuscular injections and occasionally caused by congenital fibrosis or trauma [1 5]. Characteristic manifestations of deltoid contracture (DC) include a palpable fibrous cord within the deltoid muscle, skin dimpling overlying the fibrous cords, limitation of adduction and flexion of the glenohumeral joint, winging (lateroanterior rotation), or lateroinferior rotation of the scapula [1 5]. However, such typical clinical manifestations of DC or characteristic shoulder deformities frequently are not apparent [3 5]. Fibrous cord formation within the deltoid is the key finding for diagnosing DC. MRI has been reported as the best technique to delineate the hypointense fibrous cords in the deltoid muscle [1, 2]. However, the clinical application of sonography for the evaluation and diagnosis of DC is not well established. The purpose of this study was to evaluate the sonographic features of DC and their correlation with the findings of MRI. Materials and Methods Patients From September 1999 to October 2003, a total of 457 shoulders of 427 patients (389 patients had unilateral and 34 bilateral sonographic studies, depending on the request of the referring clinicians or surgeons) were referred to our department for sonographic studies for suspected rotator cuff tear or shoulder impingement in 393 shoulders (86%), calcific tendinitis in 37 (8.1%), nonspecific shoulder pain in 15 (3.3%), DC in 7 (1.5%), bicipital tendon rupture in 3 (0.7%) and adhesive capsulitis in 2 (0.4%). Among these 457 shoulders, 22 shoulders in 20 patients showed sonographic findings suggestive of DC. These patients were recruited for participation in this study. Seven patients were male, and 13 patients were female. Their ages ranged from 31 to 79 years (mean, 57.5 years). Fourteen left and eight right shoulders were in- 364 AJR:185, August 2005

2 Sonography and MRI of Deltoid Muscle Contracture cluded. All patients underwent subsequent shoulder MRI (as a reference standard) days after their sonographic examination. Informed consent for shoulder MRI was obtained from all patients before examination but the institutional review board of this hospital did not require approval for the retrospective review of patient records and images. All patients had anteroposterior shoulder radiographs that were retrospectively reviewed at the same time by two experienced radiologists. Sonographic Examination Sonography was performed by one of two radiologists who had 8 12 years experience with musculoskeletal imaging. Shoulder sonographic examinations were performed using an Aspen Advanced or a Sequoia 512 scanner (Acuson) with a linear 8.5-MHz transducer. The patients were requested to sit on a rotating stool with the arms hanging alongside the body. Apart from examination of the rotator cuff, routine screening of the deltoid also was performed. Anatomically, the deltoid is constructed of three portions, including the anterior, intermediate, and posterior portions, which originate from the clavicle, the acromion, and the scapular spine, respectively. All three portions insert to the deltoid tuberosity of the humerus [1, 2, 6]. Owing to different directions of the alignment of the muscle fibers, these three deltoid portions can be distinguished on sonography. To ensure a complete sonographic evaluation, all shoulders were examined with a systemic approach. At the beginning of the examination, transverse scanning of the whole length of the intermediate portion of the deltoid muscle was performed stepwise from the anterior to the posterior margin, with the sonographic probe proceeding smoothly from the origin at the lateral acromial edge down to the insertion site at the humeral tuberosity. Afterward, the whole intermediate portion was examined again stepwise from the anterior margin to the posterior margin with longitudinal scanning. The anterior and posterior portions of the deltoid then were evaluated from the origins at the clavicle and the scapular spine to the humeral tuberosity, respectively, in a similar manner, with special attention in complete coverage of the whole portion from the medial to lateral aspects. When normal fascial planes of the deltoid muscle including the intermuscular tendons, linear structures of the perimysium and epimysium were clearly identified (Fig. 1), the deltoid muscle was considered normal. When these normal landmarks could not be clearly seen, involvement of the upper, middle, and lower third of the muscles was recorded and detailed identification of abnormal findings, including the presence or absence of a masslike lesion, echogenicities, size and shape of the intralesional components, and calcifications (hyperechoic foci with acoustic shadow that subsequently could be confirmed on shoulder radiographs), was performed. Finally, at the region with the maximal discernable lesion, an image in transverse free-style extended view was obtained and the maximal transverse lesion diameter was measured. Generally, the whole deltoid could be studied thoroughly within 5 to 10 min. MRI MRI was performed on a 1.5-T system (Signa Horizon LX, GE Healthcare) equipped with a standard shoulder surface coil. The shoulder MRI protocol was identical for all patients and consisted of axial T1-weighted conventional spin-echo sequence (TR msec/te msec, /18 20), axial T2-weighted multiple planar gradient-recalled sequence ( /18 20; flip angle, 30 ), and both sequences were performed from the level of the clavicle down to the level of the deltoid insertion at the humeral tuberosity with the following parameters: matrix, ; field of view, 180 mm; section thickness, 4-mm with 1-mm gap. Coronal oblique proton-density-weighted (3,600 4,000/18 20), with axis parallel to the supraspinatus tendon, and sagittal oblique T2-weighted (3,600 4,000/60 80), with axis perpendicular to the coronal oblique axis, fast spin-echo sequences were performed with the following parameters: echo-train length, 12; matrix, ; field of view, 180 mm; section thickness, 4 mm with 1-mm gap. No IV or intraarticular gadolinium was administered. Imaging Analysis Sonographic images were randomized and reviewed simultaneously by two experienced musculoskeletal radiologists who reached a consensus agreement on all findings. The following items were assessed: the presence or absence of a masslike lesion in the deltoid muscle; the location of the lesion (anterior, intermediate, or posterior portion of the deltoid muscle); lesion extent (which was scored using a 3-point scale: 1= one third of the deltoid length from the acromion to the deltoid insertion at the humeral tuberosity, 2 = > one-third and two thirds involved, 3 = > two thirds involved, respectively); transverse intralesional morphologic appearances (intralesional hypoechoic areas or calcifications); and transverse diameter of the maximal discernable lesion. One month after the review of sonograms was completed, the MR images were randomized and then reviewed by the same two radiologists until a consensus agreement was reached for all findings. The diagnosis of DC was established by visualization of typical hypointense cords [1, 2] in the deltoid muscle on all pulse sequences. Involvement of the anterior, intermediate, and/or posterior portions of the deltoid muscle was determined. Similar to the analysis of the sonograms, the extent of the hypointense cords in the upper, middle, and lower third of the deltoid also was assessed with a 3-point scale. The cross-sectional morphologic appearances of the hypointense cord on MRI were characterized. The maximal lesion diameter measured using axial multiplanar gradient-echo and the images was recorded. In addition to scrutinizing the deltoid, other structures of the shoulder, especially the rotator cuff, also were examined. On sonography, thickening of the rotator cuff with decreased echogenicity was indicative of tendinitis. Rotator cuff with decreased echogenicity, loss of convexity of tendon to bursal interface, and focal thinning were indicative of partial thickness tear. Rotator cuff with focal tendon interruption and hypoechoic gap, tendon retraction, and uncovered humeral cartilage were indicative of full thickness tear. On MRI, rotator cuff tendinosis was characterized by intact thickened tendon with intermediate intensity on T1-weighted images and hyperintensity on T2-weighted images. Partial thickness tear was characterized by focal areas of intermediate intensity on T1-weighted images that corresponded to hyperintense fluid filling of an incomplete gap of the tendon on T2-weighted images. Full thickness tear was characterized by thickened indistinct rotator cuff tendon on T1- weighted image with prominently hyperintense fluid intensity extended through the entire thickness of the interrupted rotator cuff tendon on T2-weighted images. Eventually, all shoulder radiographs were retrospectively reviewed to compare findings with the presence or absence of deltoid calcifications detected by sonography and to detect any metallic foreign bodies that might induce artifacts on MRI. Statistical Analysis Kappa statistics were used to test the agreement of the lesion extent score between sonography and MRI. A kappa value of less than 0.20 indicated poor agreement; , fair agreement; , moderate agreement; , good agreement; and , excellent agreement. The agreement between the transverse morphologic appearances of DC demonstrated on sonography and on axial MR images also was tested by kappa statistics. Wilcoxon s signed rank test was used to compare the transverse diameter of maximal discernable lesion measured on sonography and MRI. A p value less than 0.05 was considered to indicate a statistically significant difference. The deltoid injection histories, clinical presentations, and surgical findings were reviewed by two radiologists and an orthopedic surgeon at the same time. Nine patients received surgical distal release of the DC and simultaneous resection of fibrotic AJR:185, August

3 cords was performed in four of them. Histopathologic specimens were reviewed by a radiologist and a pathologist together. Results All 20 patients had histories of repetitive intramuscular injections in the deltoid muscles but none of them was able to remember or had records available of the exact timing and number of injections. Only seven (35%) of the 20 patients were referred for shoulder sonography under an initial clinical impression of DC. Among them, two patients showed characteristic winging of the scapula, four had limited adduction of the glenohumeral joint, and one had a palpable mass in the deltoid. In another 11 patients, shoulder impingement was impressed due to pain and restricted range of shoulder motion. In the remaining two patients, one suffered from the deltoid muscle pain for 1 week after intramuscular injection and one suffered from shoulder pain for 3 weeks after shoulder sprain. Initial sonograms of these two patients showed an ill-defined hyperechoic lesion in the intermediate portion of the deltoid muscle but subsequent shoulder MRI was negative for DC. Shoulder pain in these two patients resolved within 3 weeks and 1 month, respectively, and follow-up sonography showed resumption of a normal deltoid configuration. Transient focal deltoid inflammation or myositis was diagnosed and these two cases were therefore considered as false-positive on sonography. For 20 shoulders in 18 patients, both sonography and MRI showed positive features of DC with involvement of the intermediate portion of the deltoid muscles and sparing of the anterior portion. Sonography and MRI revealed involvement of the posterior deltoid in 16 shoulders and 14 shoulders, respectively. The results of comparison of sonography and MRI assessment of the lesion extent, transverse morphologic appearance, and maximal lesion diameter are summarized in Table 1. Confident MRI diagnosis of DC could be established by coronal and sagittal oblique delineation of the whole length of the hypointense cords, which aligned along the typical course of the deltoid. In addition, none of our patients showed any metallic foreign bodies on shoulder radiographs, which allowed exclusion of metallic artifacts that might also appear hypointense on MRI. Sonography showed good agreement (kappa = 0.796, p < 0.001) with MRI in the assessment of the longitudinal extent of the DC with involvement of whole length of the deltoid in 12 shoulders (lesion score = 3), upper and middle third in five (lesion score = 2), and middle third in one (lesion score = 1). In the remaining two shoulders, sonography revealed upper two-thirds involvement while MRI showed the DC involved the whole length of the deltoid. The sonography morphologic appearances on transverse scans of the DC lesions could be categorized into three patterns. Sonographic pattern I (n = 9) denoted a hyperechoic lesion with multiple less than 8-mm intralesional hypoechoic spots (Fig. 2A). Sonographic pattern II (n = 7) denoted a lesion with heterogeneous echotexture predominantly occupied by one to several 8- to 15-mm hypoechoic areas (Fig. 3A). Sonographic pattern III (n =4) denoted a calcified nodule with size greater than 15-mm (Fig. 4A). In agreement with the sonograms, axial MR images also showed three lesion patterns. MRI pattern I (n = 10) denoted multiple less than 8-mm intradeltoid hypointense spots (Fig. 2B). MRI pattern II (n = 6) denoted a cluster of several 8- to 15-mm hypointense areas (Fig. 3B). MRI pattern III (n = 4) denoted a hypointense nodule with size greater than 15 mm (Fig. 4B). The three transverse sonography lesion morphologic patterns showed excellent agreement with three MR lesion patterns (kappa = 0.921, p < 0.001). However, the maximal lesion diameters appeared significantly larger on sonography than on MRI (mean, 2.8 ± 0.6 cm vs 2.0 ± 0.8 cm, p < 0.001, Wilcoxon s signed rank test), which was attributable to a hyperechoic rim of variable thickness on sonographic Fig year-old healthy woman volunteer. Normal sonographic free-style transverse extended view of intermediate portion of deltoid muscle with normal short linear perimysium (small arrows), intermuscular tendons (open arrows), and epimysium (large arrow) between intermediate portion and posterior portion of deltoid muscle (PM). TABLE 1: Comparison of Sonographic and MRI Lesion Extent, Morphologic Appearances, and Maximal Transverse Lesion Diameters in 20 Shoulders with Deltoid Contracture Lesion Data Sonogram (n = 20) MRI (n = 20) Agreement and p Lesion extent (lesion score) kappa = a, p < Middle third (Score = 1) n =1 n =1 Upper two thirds (Score = 2) n =7 n =5 Whole length (Score = 3) n =12 n =14 Morphologic patterns kappa = a, p <0.001 Pattern I n =9 n =10 Pattern II n =7 n =6 Pattern III n =4 n =4 Maximal diameters (cm) Range (mean ± SD) (2.8 ± 0.6) (2.0 ± 0.8) p < b Note Maximal diameter = diameter of the maximal discernable lesion measured on free-style transverse scan or shoulder MRI. a kappa = kappa statistics. b Wilcoxon s signed rank test. 366 AJR:185, August 2005

4 Sonography and MRI of Deltoid Muscle Contracture Fig year-old woman presenting with shoulder pain. A, Sonographic free-style transverse extended scan shows pattern I deltoid contracture lesion as hyperechoic lesion with blurred outline (arrows) and multiple small hypoechoic spots (open arrows) in middle third of intermediate deltoid. B, Transverse gradient-echo MR image corresponding to A reveals pattern I lesion with multiple small (< 5-mm) hypointense nodules (open arrows) in intermediate deltoid. Note presence of hyperintensities, which correspond to relatively immature fibrosis, around hypointense mature fibrotic foci. patterns I and II lesions. In addition to DC, coexistent supraspinatus tendinopathies also were identified on both sonography and MRI in 10 shoulders (complete tear in six, partial tear in three, and supraspinatus tendinitis in one). Nine of 20 shoulders were treated surgically. Four of these shoulders received concomitant repair of complete tear of the supraspinatus tendon and distal release of the DC. A fifth shoulder first was treated by repair of the supraspinatus tendon tear, and distal release of the DC was performed 3 months later due to persistent shoulder pain and limitation of motion. The remaining four shoulders without coexistent rotator cuff lesion were treated with distal release of the DC. Partial resection of the deltoid fibrous cords also was performed in four shoulders (one with sonographic pattern I, two with sonographic pattern II, and one with sonographic pattern III). Among these nine shoulders, marked clinical improvement was noted in seven while mild painful sensation during adduction of the glenohumeral joint was still noted in two. Histopathologic examinations (Fig. 3C) showed a dense fibrous portion with thickened collagen in the lesions that corresponded to the hypoechoic areas on sonograms and hypointense areas on MRI. On the other hand, the tissues at the fibromuscular junctional regions were composed of prominent capillaries and myofibroblasts while the fibrous stroma was less compact and these histopathologic findings correlated with the hyperechoic changes associated with DC, especially at the periphery of DC lesions. Prominent dystrophic calcification in the dense sclerotic fibrous stroma was noted in the specimen with sonographic pattern III. Discussion Intramuscular injection can induce local hemorrhage, ischemia, necrosis, and chemical myositis [3 5, 7, 8]. Repeated intramuscular injections, however, can lead to fibrosis of the injection sites, coalition of the adjoining perimysial and epimysial tissues, and eventually can evolve into contracture status of the muscles or injection myopathy [3, 7, 8]. In our country, intramuscular injection is a common practice in local clinics in patients with fever, pain, or infection. The gluteal, deltoid, and quadriceps muscles are the most common sites for these injections [1, 7 11] and among the three deltoid portions, the intermediate portion is the preferable site for intramuscular injection. However, the intermediate portion with short muscle fibers is most susceptible because even fibrosis of only a few short muscles easily can lead to coalition of the interspersed tendons and subsequent development of DC [2, 6, 7]. In our series, all patients had histories of repetitive injections in the deltoid muscles and DC predominantly affects the intermediate portion of the deltoid. Although typical DC may present with characteristic features such as palpable fibrous cords within the deltoid and winged scapula, this uncommon entity occasionally may be overlooked due to diversity of clinical presentations [3 5]. Among our patients, approximately two thirds manifested pain and a restricted range of shoulder motion that led to initial clinical suspicion of shoulder impingement. Subacromial impingement associated with DC has been described [9, 12]. In this study, DC was also associated with rotator cuff abnormalities in 10 of 20 shoulders. The radiographic features of DC include soft-tissue calcifications, winging of the scapula, lateral down-sloping of the acromial process, and a superior acromial enthesophyte [1, 13]. Although these indirect findings are suggestive of DC, a specific diagnosis relies on direct visualization of the fibrotic cords. MRI can offer a confident diagnosis of DC by multiplanar delineation of the hypointense fibrotic cords from the origins (clavicle, acromion, and scapular spine) to the common insertion at the humeral tuberosity [1, 2]. The location and the extent of the lesions, which are important information for surgical planning, also can be well delineated, especially A B AJR:185, August

5 A C Fig year-old woman presenting with shoulder pain and limited range of motion. A, Sonographic free-style transverse extended scan shows pattern II deltoid contracture lesion as heteroechoic lesion predominantly occupied by several 8-12-mm hypoechoic areas (open arrows) with interspersed hyperechogenicities and hyperechoic rim (arrows) in intermediate deltoid. B, Transverse gradient-echo MR image corresponding to A reveals pattern II lesion with several greater than 8-mm hypointense fibrotic areas (open arrows) in intermediate deltoid. Note presence of relatively immature fibrosis with slight hyperintensity (thin arrow) between thick hypointense fibrotic cords. C, Photomicrograph shows dense sclerotic hypocellular fibrous tissues with thickened collagenous fibers (triangles) accounting for hypoechoic area on sonogram or hypointense area on MRI. Prominent vascularities and numerous myofibroblasts within loose fibrous stroma in fibromuscular junction (open arrows) with relatively immature fibrosis accounting for hyperechogenicity on sonography and hyperintensity on MRI. (H and E; original magnification 100). on gradient-echo images, with good contrast between the hypointense cords and hyperintense soft tissues [1, 2]. Nevertheless, the routine application of MRI for screening of DC in patients with nonspecific shoulder pain is of doubtful practicality. Sonography is a useful tool for screening fibrotic soft-tissue lesions due to its relatively low cost, lack of radiation exposure, and the proximity of the lesion to the skin [14 16]. However, the diagnostic value of sonography for DC has not been well established. The results of our study indicate that sonography is feasible for the detection of clinically overlooked DC during routine shoulder examination. Our study showed that detailed transverse sonographic scanning of the three portions of the deltoid from the origins to humeral tuberosity facilitated localization of the fibrotic cords and longitudinal scanning showed good agreement with MRI in assessing the longitudinal extent of the lesions (kappa = 0.796, p < 0.001). Apart from the longitudinal extent of DC, our study also showed that three morphologic patterns of the fibrotic cords could be demonstrated on transverse sonographic scan and that these patterns were well correlated with similar findings on MRI (kappa = 0.921, p < 0.001). Pattern I denoted multiple hypoechoic or hypointense small-caliber fibrotic cords (diameter < 8 mm) and was identifiable on sonograms and MRI. This picture plausibly reflects the initial stage of small focal fibrotic foci formation within the deltoid after intramuscular injections. As the number of injections increases and/or as the lesion evolves over time, these small-caliber cords might coalesce into several 8- to 15-mm hypoechoic or hypointense areas resulting in pattern II or the lesions might develop into calcified fibrotic masses leading to pattern III. It is not surprising that the dense sclerotic fibrous tissues in DC lesions, which are composed of thickened collagen fibers and few cells, appear B 368 AJR:185, August 2005

6 Sonography and MRI of Deltoid Muscle Contracture A C Fig year-old man presenting with palpable mass in deltoid and restricted range of shoulder motion. A, Sonographic free-style transverse extended scan shows pattern III deltoid contracture lesion involving whole length of deltoid muscle and 2.5-cm calcified mass (open arrows) with obvious acoustic shadow at lower third of intermediate deltoid. B, Transverse proton density-weighted image corresponding to A reveals pattern III deltoid contracture lesion with 2.5-cm hypointense fibrotic mass (open arrow) in intermediate deltoid. C, Shoulder anteroposterior radiograph shows lateral down-sloping of acromion, lateroinferior rotation of scapula, and dense calcification (open arrow) in lower third of deltoid muscle. B hypoechoic on sonograms and hypointense on MRI [1, 2, 14 21]. Above all, the loosely used term fibrosis only vaguely characterizes the fibroblasts and collagen in the tissues and does not consider the stage of maturation. In fact, the histologic findings of early or immature fibrosis to mature fibrosis vary widely [14 17]. At one end of the spectrum, immature fibrosis is characterized by numerous fibroblasts or myofibroblasts, and scant immature collagen with thin fibers and abundant vascularities that allow passage of AJR:185, August

7 proteins and red blood cells into the extravascular spaces [17, 18]. This tissue may exhibit high signal intensity on T2-weighted MR images because of the high free water content and hypercellularity and hyperechogenicity on sonography due to immature thin collagen fibers [14 17, 19, 20]. As maturation of fibrosis progresses, increased amounts of mature collagen, decreased cellularity, and vascularity may lead to decreased echogenicity and MRI signal intensity and thus the fibrotic tissue may appear isoechoic or isointense [14, 16, 17, 19]. In this series, sonographic detection of these hyperechoic or isoechoic immature fibrotic changes, especially at the outer rim of the DC lesions, was more appreciable than the corresponding hyper- or isointense changes on MRI. This might explain why the transverse diameters of the maximal discernable lesions measured on sonograms appeared significantly larger than those on MRI. On the other end of the spectrum, hypointense or hypoechoic mature scar or fibrosis is acellular or extremely hypocellular, comprising mostly dense collagen with few vessels, and eventually may become densely calcified [1, 2, 14, 17 21]. DC can lead to decreased muscle strength in flexion, extension, and limited adduction of the shoulder. Prolonged anterosuperior translation of humeral head due to DC may lead to incongruence and osteoarthritic changes of the glenohumeral joint [3 5, 7 9, 12, 1]. Distal release and partial resection of the fibrotic masses, if necessary, result in excellent recovery [4, 5, 9 11]. Surprisingly good function and strength of the shoulder could be maintained even after resection of the affected deltoid muscle, possibly due to hypertrophy of the remaining muscles [9, 22]. However, DC may mimic or even coexist with shoulder impingement or rotator cuff abnormalities [1, 9, 12]. Our study showed that sonography is useful for detecting clinically overlooked DC and for comprehensive surgical planning, especially when rotator cuff repair and distal release of DC can be performed simultaneously [9]. On the other hand, our study harbored several limitations. First, there was selection bias as sonography was applied as the gatekeeper for subjects to enter the study. Second, the data with regard to injection type, depth, and frequency could not be documented accurately. Third, despite MRI being useful for evaluating DC [1, 2] and that we used it as a reference standard, to our knowledge, no randomized large-series study regarding the accuracy of MRI in the diagnosis of DC has been reported. Fourth, among nine of 20 shoulders that were surgically treated with distal release, only four cases underwent partial fibrous cord resection with specimens available for histopathologic correlation with imaging findings. In summary, our results show that sonography is helpful for revealing this uncommon but easily overlooked entity, especially in patients presenting with nonspecific shoulder pain and limited range of motion, and with good correlation with MRI. References 1. Chen CKH, Yeh L, Chen CT, Pan HB, Yang CF, Resnick D. Contracture of the deltoid muscle: imaging findings in 17 patients. AJR 1998; 170: Ogawa K, Takahashi M, Naniwa T. Deltoid contracture: MR imaging features. Clin Radiol 2001; 56: Shanmugasundaram TK. Post-injection fibrosis of skeletal muscle: a clinical problem: a personal series of 169 cases. Int Orthop 1980; 4: The Ad Hoc Committee of the Japanese Orthopedic Association of Muscular Contracture. Report of the diagnosis and treatment of muscular contracture [in Japanese]. Nippon Seikeigeka Gakkai Zasshi 1985; 59: Minami M, Yamasaki J, Minami A, Ishii S. A postoperative long-term study of the deltoid contracture in children. J Pediatr Orthop 1984; 4: Gray H. Anatomy of the human body, 30th American ed. Philadelphia, PA: Lea & Feibiger, 1985: Chen SS. Histopathological and histochemical studies in deltoid and gluteal contracture. Taiwan Yi Xue Hui Za Zhi 1983; 82: Chen SS, Chien CH, Yu HS. Syndrome of deltoid and/or gluteal fibrotic contracture: an injection myopathy. Acta Neurol Scand 1988; 78: Ko JY, An KN, Yamamoto R. Contracture of deltoid muscle. Results of distal release. J Bone Joint Surg Am 1998; 80: Groves RJ, Goldner JL. Contracture of the deltoid muscle in the adult after intramuscular injections. Report of three cases. J Bone Joint Surg Am 1974; 56: Chen WJ, Wu CC, Lin YH, Shih CH. Treatment of deltoid contracture in adults by distal release of the deltoid. Clin Orthop 2000; 378: Ogawa K, Inokuchi W, Naniwa T. Subacromial impingement associated with deltoid contracture. A report of two cases. J Bone Joint Surg Am 1999; 81: Ogawa K, Yoshida A, Inokuchi W. Deltoid contracture: a radiographic survey of bone and joint abnormalities. J Shoulder Elbow Surg 1999; 8: Robbin MR, Murphey MD, Temple T, Kransdorf MJ, Choi JJ. Imaging of musculoskeletal fibromatosis. RadioGraphics 2001; 21: Chan YL, Cheng JCY, Metreweli C. Ultrasonography of congenital torticollis. Pediatr Radiol 1992; 22: Ablin DS, Jain K, Howell L, West DC. Ultrasound and MR imaging of fibromatosis colli (sternomastoid tumor of infancy). Pediatr Radiol 1998; 28: Lee JKT, Glazer HS. Controversy in the MR imaging appearance of fibrosis. Radiology 1990; 177: Robbins SL, Cotran RS, Kumar V. Healing and repair. In: Pathologic basis of disease, 3rd ed. Philadelphia, PA: Saunders, 1984: Chandraratna PAN, Whittaker P, Chandraratna PM, Gallet J, Kloner RA, Hla A. Characterization of collagen by high-frequency ultrasound: evidence for different acoustic properties based on collagen fiber morphologic characteristics. Am Heart J 1997; 133: Rubenstein WA, Gray G, Auh YH, et al. CT of fibrous tissues and tumors with sonographic correlation. AJR 1986; 147: Sunderam M, McGuire MH, Schajowicz F. Soft-tissue masses: histologic basis for decreased signal (short T2) on T2-weighted MR images. AJR 1987; 148: Markhede G, Monastyrski J, Stener B. Shoulder function after deltoid muscle removal. Acta Orthop Scand 1985; 56: AJR:185, August 2005