Clinical Imaging 36 (2012) 791 796

Rotator cuff tears: association with acromion angulation on MRI

Joseph C. McGinley a, b,, Sundeep Agrawal b , Sandip Biswal b
b a Casper Medical Imaging, Casper, WY, USA Department of Radiology, Stanford University Medical Center, 300 Pasteur Drive, Stanford, CA 94305, USA

Received 12 February 2012; received in revised form 29 March 2012; accepted 4 April 2012

Abstract Objective: Using magnetic resonance imaging (MRI), evaluate the correlation of acromion angulation with thickening of the coracoacromial ligament (CAL) and narrowing of the subacromial space resulting in impingement upon the rotator cuff tendons. Materials and methods: Eighty-nine shoulder MRI studies performed on a 3T scanner were retrospectively analyzed by two blinded independent reviewers. Measurements of the acromion angle (delta angle), CAL thickness and distance between the CAL and humeral head were obtained. The data were categorized into two groups, delta angle less that and greater than 7.5. The presence or absence of full thickness (FT) or near full thickness (NFT) rotator cuff tears was noted. Results: In group 1, the acromion angle varied from -6.8 to 6.8 (1.73.5) with a CAL thickness of 0.910.20 mm and a subacromial distance of 6.470.88 mm. Group 2 acromion angle varied from 7.6 to 46.8 (18.08.1) with a CAL of 1.770.51 mm and a subacromial distance of 4.520.82 mm. The difference in CAL thickness and subacromial distance were significantly different between the two groups (Pb.001). In Group 1, 3 out of 51 patients had a FT or NFT tear of the rotator cuff compared to 20 out of 38 in Group 2 (Pb.001). There was no significant interobserver variability. Conclusion: Steep acromion angulation is associated with CAL thickening and narrowing of the subacromial space. Patients with a steep acromion angle had a statistically increased incidence of rotator cuff tears. 2012 Elsevier Inc. All rights reserved.
Keywords: Shoulder; Magnetic resonance imaging; Rotator cuff; Anatomy; Acromion

1. Introduction Subacromial impingement and degeneration of the rotator cuff (RTC) tendons is one of the most common causes of shoulder pain and disability. The incidence of shoulder pain is approximately 11.219 cases per 1000 patients per year [1]. In 1972, Neer rst described RTC impingement related to the anteroinferior portion of the acromion in three stages [2]. The process was described as a progression of disease beginning in younger patients ultimately resulting in RTC tears [2,3]. Subsequent studies go on to describe acromion
Presented at ARRS 2011, Chicago, IL. Corresponding author. Casper Medical Imaging, 419 S Washington St, Ste 101, Casper, WY 82601, USA. Tel.: +1 307 265 1620; fax: +1 307 237 1074. E-mail addresses: mcgjoe26@yahoo.com (J.C. McGinley), sagrawal@gwmail.gwu.edu (S. Agrawal), biswals@stanford.edu (S. Biswal).

shape, angulation and conguration in association with RTC pathology [46]. The anatomy and biomechanics related to subacromial impingement is a complex association between acromion angulation, the coracoacromial ligament (CAL) and the subacromial space (SAS). Previously, authors have described features of the acromion, including the type and the conguration, on both radiographic images and anatomic specimens [47]. Furthermore, variations in the anatomy and morphology of the CAL are associated with varying degrees of RTC degeneration in anatomic specimens [7,8]. The biomechanics likely relate to the steep acromion angle causing friction of the CAL along the bursal bers of the supraspinatus (SST) and infraspinatus tendons (IST). The bursal surface pressures may in turn increase tension in the undersurface bers resulting in articular surface tears of the SST and IST [9]. Most patients with persistent shoulder pain related to impingement ultimately undergo magnetic resonance

0899-7071/$ see front matter 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.clinimag.2012.04.007

792

J.C. McGinley et al. / Clinical Imaging 36 (2012) 791796

imaging (MRI) for further evaluation. Data describing the down-sloping angulation of the acromion in association with CAL thickness, SAS narrowing and ultimate RTC deterioration has not been described. The correlation of MRI ndings of acromial shape and RTC pathology has been limited due do variations in positioning, slice selection and interobserver variability [1013]. The goal of the current study was to characterize the acromion angle with respect to RTC tears and delineate a normal/abnormal range of measurements for CAL thickness and SAS distance using MRI. The methods outlined in this study attempt to provide a standard measurement and reference pattern that is reproducible and reduces interobserver variability. These measurements will provide a guideline for evaluating the SAS on MRI and may help guide the surgeon in treatment of patients with impingement.

2. Materials and methods This retrospective study was performed with institutional review board approval and a waiver of patient informed consent. 2.1. Patients A total of 89 patient exams were retrospectively reviewed. The average age was 46.316.4 years and ranged from 1580 years. Females and males represented 36 and 53 of the patients, respectively. 2.2. Acromion angulation measurement (delta angle) The acromion measurement was obtained on either T1 or PD FS coronal MRI sequences from a 3T scanner. The angle was based on the steepest point of the acromion relative to the clavicle (Fig. 1A). A line was drawn along the margin of the acromion with the reference being a horizontal line on the image. 2.3. Coracoacromial ligament thickness The CAL thickness was obtained on either T2 FS or PD FS coronal MRI sequences from a 3T scanner. The measurement was performed at the lateral margin of the acromion at the insertion of the CAL. A vertical line was drawn at the thickest point of the insertion (Fig. 2A). 2.4. Subacromial space The SAS distance was obtained on either T2 FS or PD FS coronal MRI sequences from a 3T scanner. A vertical line was drawn from the undersurface of the CAL at its insertion to the superior margin of the humeral head cartilage at the narrowest point on coronal images (Fig. 3A). In a blinded fashion, each of two reviewers independently evaluated a complete shoulder MRI exam performed on a 3T scanner on each patient. Reviewer 1 was a board certied musculoskeletal radiologist and reviewer 2 a radiology resident. The standard scanning protocol for most patients was COR T1, COR T2 FS, COR PD FS, SAG T2 FS, SAG

Fig. 1. The above images demonstrate normal and abnormal delta angle measurements. (A) Coronal T1 image of a right shoulder obtained on a 3T scanner. The white lines indicated the measurement method of the delta angle. The upper line was drawn along the undersurface of the acromion while a horizontal reference line was drawn on the lower portion of the image. The delta angle measured in this patient equaled 3. (B) Coronal PD FS image of a right shoulder obtained on a 3T scanner. The delta angle measured on this exam was abnormal equaling 22.

2.5. Study evaluation

J.C. McGinley et al. / Clinical Imaging 36 (2012) 791796

793

Fig. 2. The above images demonstrate normal and abnormal coracoacromial ligament thickness. (A) Coronal PD FS image of a right shoulder obtained on a 3T scanner. The thickness was measured at the insertion of the CAL using a vertical line. The CAL thickness in this patient measured 1.2 mm. (B) Coronal T2 FS image of a right shoulder obtained on a 3T scanner. The CAL thickness was abnormal measuring 2.6 mm.

Fig. 3. The above images demonstrate normal and abnormal subacromial space distances. A) Coronal PD FS image of a right shoulder obtained on a 3T scanner. The SAS distance was measured from the undersurface of the CAL insertion to the superior aspect of the humeral head cartilage. The SAS distance was normal measuring 6.5 mm. B) Coronal T2 FS image of a right shoulder obtained on a 3T scanner. The SAS distance was abnormal measuring 3.7 mm.

T1, and AX PD FS (slice thickness of 3 mm with a 1 mm skip and FOV of 140 cm). The reviewers each measured the delta angle, CAL thickness, SAS distance and characterized the rotator cuff tendons independently. Evaluation of the RTC was performed by specically noting pathology related to the supraspinatous tendon (SST) and infraspinatous

tendon (IST), the presence or absence of tears and the degree of tearing. A comparison of the RTC evaluation was made with the completed report provided by a board certied musculoskeletal radiologist and musculoskeletal radiology fellow. Data were acquired by each reviewer and statistical

794

J.C. McGinley et al. / Clinical Imaging 36 (2012) 791796

analysis performed to determine the delta angle associated with a signicant increased incidence of RTC tears. Once the angle was determined the average CAL thickness, and SAS distance was calculated and statistical analysis performed. 2.6. Statistical analysis The data were compared between the two groups based on the angle of the acromion associated with a signicant increased incidence of RTC tears. The acromion angle used to group the data was determined through an analysis of the acromion angulation and presence of RTC tears. The data comparison of the two groups was performed using a simple t-test analysis. Measurements were analyzed using analysis of variance for repeated measures (two times) for each group. Data was normally distributed using Wilk-Shapiro test. Interobserver agreement was determined using the coefcient as a chance-corrected measure of agreement as follows: b0 poor, 00.20 slight, 0.210.40 fair, 0.410.60 moderate, 0.610.80 substantial, and 0.811.0 almost perfect agreement [14]. 3. Results The delta angle corresponding to a signicant increase in FT or NFT RTC tears was found to be greater than 7.5 (Fig. 1B). At an angle greater than 7.5, there was a signicant increase in the incidence of full thickness and near full thickness RTC tears involving either the SST or IST (Pb.001). The data were separated into two groups: Group 1 was dened as a delta angle equal or less than 7.5and contained 51 patients. The Group 1 delta angle ranged from 6.8 to 7.5 (1.73.5). Group 2 contained all patients with a delta angle greater than 7.5 and contained 38 patients. The Group 2 delta angle ranged from 7.646.8 (18.08.1). In group 2, 20 of the 38 patients had FT or NFT tears of either the SST or IST (FT=6, NFT=14). Of the 14 NFT tears, 10 were articular surface and 4 were bursal surface tears. In group 1, 3 of the 51 patients had FT or NFT tears of either the SST or IST (FT=1, NFT=2) (Fig. 4A and B). The two NFT tears were both articular surface tears. The CAL thickness and SAS distance was compared between Groups 1 and 2, Table 1 summarizes the data. Comparison between Group 1 and Group 2 demonstrated a signicant increase in CAL thickness in Group 2 compared to Group 1 (Fig. 2, Pb.001). In Group 1, the CAL thickness measured 0.910.20 mm (0.551.4 mm). In Group 2, the CAL thickness measured 1.770.51 mm (0.822.8 mm). The SAS distance was signicantly less in Group 2 compared to Group 1 (Fig. 3, Pb.001). In Group 1, the SAS distance measured 6.470.88 mm (8.84.9 mm). In Group 2, the SAS distance measured 4.520.82 mm (63.1 mm). Evaluation of the variability between measurements of the delta angle, CAL thickness and SAS distance of the two independent reviewers demonstrated signicant agreement

Fig. 4. The above images demonstrate normal and abnormal rotator cuff tendons from Group 1 and Group 2, respectively. (A) Coronal PD FS image of a right shoulder from a 3T scanner. The image demonstrates normal signal in the RTC tendons. (B) Coronal T2 FS image of a right shoulder from a 3T scanner. The image demonstrates a FT tear in the SST with high T2 signal at its insertion.

with a kappa value of 0.81, 0.86 and 0.84 respectively. When disagreement of RTC pathology occurred, the nalized report, generated by a board certied musculoskeletal radiologist and musculoskeletal radiology fellow was used to ultimately categorize the RTC pathology. 4. Discussion The anatomy and biomechanics of the shoulder represent a complex interaction between bony structures, ligaments

J.C. McGinley et al. / Clinical Imaging 36 (2012) 791796 Table 1 Acromion angle and anatomic measurements Group 1 2 Acromion angle 6.8 to 7.5 (1.73.5) 7.646.8 (18.08.1) CAL thickness 0.910.20 mm (0.551.4 mm) 1.770.51 mm (0.822.8 mm) SAS distance 6.470.88 mm (8.84.9 mm) 4.520.82 mm (63.1 mm)

795

RTC tears 3/51 20/38

and the rotator cuff tendons [15]. Variations in the anatomic arrangement can result in functional impingement upon the rotator cuff tendons. Prior research has suggested that variation in acromion anatomy, specically its shape and angulation, and osteoarthritis can increase the risk of RTC pathology and tears [2,3,16]. Recent studies have shown a higher association of symptoms with narrowing of the acromiohumeral distance [17]. The data presented in the current studies utilizes MRI to dene the angle between the acromion and clavicle (delta angle) that is signicantly associated with RTC tears. Our data showed a delta angle greater than 7.5 was associated with a signicantly higher incidence of tears involving the SST and IST tendons. MRI of the shoulder is often the imaging modality of choice to evaluate the intrinsic anatomy of the shoulder. Previous authors have described subacromial impingement relative to the radiographic appearance of the acromion; however the current study outlines a method that reproducibly measures acromion angulation and associated pathology. The data presented further delineates the pathology associated with steep acromion angulation. In patients with a delta angle greater than 7.5, there was associated thickening of the CAL and subsequent narrowing of the SAS. Soslowsky et al. [18] demonstrated the differing anatomic and material properties of the CAL in normal shoulders compared to shoulders with RTC disease. In patients with RTC disease there was an increased crosssectional area and diminished tensile strength. Fremerey et al. [7], further described differences in CAL anatomy and material properties in patients with RTC tears with respect to age. Specically, enlargement of the lateral band was seen in older patients with RTC degeneration. Sarkar et al. [19] and Uhthoff et al. [20] demonstrated enlargement and stiffening of the CAL along with decreased collagen ber organization in patients with subacromial impingement. Our study demonstrated a signicant difference of the CAL thickness at its insertion between groups with a low delta angle and steep delta angle. There was also a signicant difference in the incidence of RTC pathology within these groups. Based on our data, using 3T MRI, a normal value of CAL thickness was found to be less than or equal to 1.5 mm. Below 1.5, there were signicantly less cases of RTC tears and a signicantly lower delta angle. Angulation of the acromion and thickening of the CAL ultimately results in narrowing of the SAS. The RTC tendons, SST and IST, traverse this region. The SST is the main tendon within the SAS; however the anterior bers of the IST also traverse this region. The SAS was dened as the region between the CAL and the cartilage of the humeral head; this

was the maximal space available for the SST tendon. Our data demonstrated in patients with steep angulation of the acromion and thickening of the CAL, there was signicant narrowing of the SAS. This group was associated with an increased incidence of RTC tears. Neer [2] initially described acromion anatomy in relation to RTC impingement. Our goal was to dene reproducible anatomic MRI measurements to help assess patients with suspected impingement. We determined a normal value of the SAS as a distance greater than or equal to 5 mm. Below this value, there was a signicant increase in the incidence or RTC tears. Our data will affect patent care by providing a standard, reproducible, measurement process using MRI to evaluate the acromion and related anatomy. A cutoff value of normal was delineated with measurements outside of this value associated with a signicantly increased incidence of RTC tears. Surgeons may use this analysis to help guide therapy in patients with shoulder pain and abnormal MRI scan demonstrated anatomy associated with impingement. Limitations associated with our methods primarily centers on the delta angle measurement. Initially several different anatomic landmarks were used as a reference point in measuring acromion angulation including the vertex of the humeral head, the glenohumeral joint, the SST tendon, the clavicle and the acromioclavicular joint. However, positioning variations in the scanner and observer variability in placing the reference line resulted in signicant differences in delta angle measurement between the two reviewers using these landmarks. A simple horizontal line was found to provide the best correlation between the data with a kappa value of 0.81 between reviewers. The horizontal line provides and easily reproducible reference to measure acromion angulation. Positioning can still vary from patient to patient; however, our data showed excellent correlation despite likely differences in position. The lack of variation may be due to similar anatomic positioning using a shoulder coil which likely limits the differences of shoulder position between patients. Additional limitations include the lack of a detailed clinical history. Several of the patients with RTC tears likely had of trauma. However one could argue the tear may not have occurred if there was not prior weakening and deterioration of the RTC tendons due to longstanding impingement.

5. Conclusion Our methods outline a process to evaluate acromion anatomy using MRI and the data denes a normal range of

796

J.C. McGinley et al. / Clinical Imaging 36 (2012) 791796 [8] Kesmezacar H, Akgun I, Ogut T, Gokay S, Uzun I. The coracoacromial ligament: the morphology and relation to rotator cuff pathology. J Shoulder Elbow Surg 2008;17:1828. [9] Lo IK, Burkhart SS. The etiology and assessment of subscapularis tendon tears: A case for subcoracoid impingement, the roller-wringer effect, and TUFF lesions of the subscapularis. Arthroscopy 2003; 19(10):114250. [10] Peh WC, Farmer TH, Totty WG. Acromial arch shape: assessment with MR imaging. Radiology 1995;195:5015. [11] Epstein RE, Schweitzer ME, Frieman BG, Fenlin JM, Mitchell DG. Hooked acromion: prevalence on MR images of painful shoulders. Radiology 1993;187:47981. [12] Haygood TM, Langlotz CT, Kneeland JB, Iannotti JP, Williams GR, Dalinka MK. Categorization of acromial shape: interobserver variability with MR imaging and conventional radiography. AJR 1994; 162:137782. [13] Mayerhoefer ME, Breitenseher MJ, Roposch A, Treitl C, Wurnig C. Comparison of MRI and conventional radiography for assessment of acromial shape. AJR 2005;184:6715. [14] Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:15974. [15] Harrison AK, Flatow EL. Subacromial impingement syndrome. J AAOS 2001;19(11):7018. [16] de Abreu MR, Chung CB, Wesselly M, Jin-Kim H, Resnick D. Acromioclavicular joint osteoarthritis: Comparison of findings derived from MR imaging and conventional radiography. Clin Imag 2005; 29(4):2737. [17] Mayerhoefer ME, Breitenseher MJ, Wurnig C, Roposch A. Shoulder impingement: Relationship of clinical symptoms and imaging criteria. Clin J of Sport Medicine 2009;19(2):839. [18] Soslowsky LJ, An CH, Johnston SP, Carpenter JE. Geometric and mechanical properties of the coracoacromial ligament and their relationship to rotator cuff disease. Clin Orthop 1994;304:107. [19] Sarkar K, Taine W, Uhthoff HK. The ultrastructure of the coracoacromial ligament in patients with chronic impingement syndrome. Clin Orthop 1990;254:4954. [20] Uhthoff HK, Hammond DI, Sarkar K, Hooper GJ, Papoff WJ. The role of the coracoacromial ligament in the impingement syndrome. A clinical, radiological and histological study. Int Orthp 1988;12:97104.

values for acromion angulation, CAL thickness and SAS distances which, when abnormal, is associated with an increased incidence of RTC tears. This study illustrates the relationship ship of subacromial narrowing with an increased incidence of near full thickness and full thickness rotator cuff tears. Our ndings may help guide therapy in patients with shoulder pain and ndings suggesting impingement. Specifically, patients presenting with non traumatic shoulder pain may benet from early physical therapy and intervention if ndings on MRI suggest impingement. Early intervention may prevent ultimate tears of the rotator cuff and could possibly reduce morbidity with aging.

References
[1] Bot SD, van der Waal JM, Terwee CB, van der Windt D, Schellevis F, Bouter L, Dekker J. Incidence and prevalence of complaints of the neck and upper extremity in general practice. Ann Rheum Dis 2005;64: 11823. [2] Neer CS. Impingement lesions. Clin Orthop 1983;173:707. [3] Neer CS, Craig EV, Fukuda H. Cuff-tear arthropathy. J Bone Joint Surg Am 1983;65:123244. [4] Bigliani LU, Morrison DS, April EW. The morphology of the acromion and its relationship to rotator cuff tears. (abstr) Orthop Trans 1986;10:228. [5] Morrison DS, Bigliani LU. Roentgenographic analysis of acromial morphology and its relationship to rotator cuff tears. Orthop Trans 1987;11:439. [6] Morrison DS, Bigliani LU. The clinical significance of variation in acromial morphology. Orthop Trans 1987;11:234. [7] Fremerey R, Bastian L, Siebert WE. The coracoacromial ligament: anatomical and biomechanical properties with respect to age and rotator cuff disease. Knee Surg, Sports Traumatol, Arthrosc 2008;8: 30913.