A Sonographic Technique to Evaluate the ... - Wiley Online Library

34 downloads 0 Views 1MB Size Report
aspect of the proximal ulna on the sublime tubercle of the coronoid process1,8–10 and at the medial ulnar collateral ridge, which was named by Farrow et al.10.
3403jum363-518 copy_Layout 1 2/20/15 3:14 PM Page 377

ORIGINAL RESEARCH

A Sonographic Technique to Evaluate the Anterior Bundle of the Ulnar Collateral Ligament of the Elbow Imaging Features and Anatomic Correlation Fernando B. M. D. Ferreira, MD, Eloy D. A. Fernandes, MD, PhD, Flavio D. Silva, MD, Magno C. Vieira, MD, MSc, Andrea Puchnick, BSc, Artur R. C. Fernandes, MD, PhD

Article includes CME test

Objectives—The aim of this study was to test a sonographic technique used to view the anterior bundle of the ulnar collateral ligament (UCL), describe its sonographic characteristics in healthy volunteers, and verify these characteristics by determining interobserver variability and their correlations in cadavers. Methods—Sonographic studies of the anterior bundle of the UCL were performed on 48 elbows of asymptomatic healthy volunteers. The participants were examined by 3 experts, who identified the insertion sites of the anterior bundle and subjectively evaluated its echogenicity and echo texture. A sonographic examination of the anterior bundle of the UCL in a cadaveric elbow was performed, and the same aspects were evaluated. Results—In all cases, the anterior bundle of the UCL appeared as a triangular structure in the coronal plane and had a hyperechoic homogeneous echo texture in most of these cases. The cadaveric elbow had the same sonographic characteristics as the volunteers. Conclusions—As shown by examining the interobserver variability and determining the correlation with cadaveric tissue, sonography proved to be a reliable tool for evaluating the normal aspects of the anterior bundle of the UCL. Key Words—collateral ligaments; elbow; musculoskeletal ultrasound; observer variation; sonography; ulna

Received May 6, 2014, from the Department of Diagnostic Imaging, Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, Brazil. Revision requested May 22, 2014. Revised manuscript accepted for publication July 2, 2014. Address correspondence to Eloy D. A. Fernandes, MD, PhD, Department of Diagnostic Imaging, Escola Paulista de Medicina, Universidade Federal de São Paulo, Rua Napoleão de Barros 800, 04024-002 São Paulo–SP, Brazil. E-mail: [email protected] Abbreviations

ICC, intraclass correlation coefficient; UCL, ulnar collateral ligament doi:10.7863/ultra.34.3.377

T

he ulnar collateral ligament (UCL) of the elbow consists of 3 components: the anterior, posterior, and transverse bundles. The anterior bundle of the UCL strengthens the joint capsule and is the main stabilizer against valgus stress on the elbow, especially in flexing movements between 30° and 120°. The anterior bundle is at risk of injury in the overhead throwing athlete because it is the major limiting factor of valgus force across the elbow. These injuries are common in specific populations of athletes (overhead performing athletes), thus indicating the necessity to study this anatomic structure.1,2 Anatomic characteristics on sonography are fundamental for proper evaluation of the anterior bundle and distinguishing normal findings from pathologic ones. Precise characterization of the ligament is important because injuries of the UCL can cause substantial pain and disability, particularly in baseball pitchers and athletes who play tennis and golf.3

©2015 by the American Institute of Ultrasound in Medicine | J Ultrasound Med 2015; 34:377–384 | 0278-4297 | www.aium.org

3403jum363-518 copy_Layout 1 2/20/15 3:14 PM Page 378

Ferreira et al—Sonography of the Anterior Bundle of the Ulnar Collateral Ligament

Magnetic resonance imaging and magnetic resonance arthrography have been successfully implemented in diagnosing UCL abnormalities3–6; however, the procedures are time-consuming and expensive. The elbow is easily accessible by sonography.7 Sonographic evaluations tend to have greater availability, allow multiplanar evaluation, and are cost-effective. The disadvantages of this method include its operator dependence and steep learning curve. Nonetheless, in the hands of experienced professionals and with the use of the proper technique, sonography is a valuable tool in the diagnosis of elbow abnormalities.7 Based on superficial anatomic dissection studies of the elbow, the anterior bundle of the UCL has a rectangular shape. Proximally, it inserts at the inferior portion of the medial epicondyle, and distally it inserts along the medial aspect of the proximal ulna on the sublime tubercle of the coronoid process1,8–10 and at the medial ulnar collateral ridge, which was named by Farrow et al.10 Among the published studies regarding the use of sonography for the anterior bundle of the UCL, some researchers reported that the shape of the ligament was triangular,11 whereas others claimed that it was cordlike.7 There is also some disagreement regarding methods used for measuring the structure when using sonography and cadaveric tissues.8,9,11 Therefore, the objective of this study was to test a standardized evaluation technique for the anterior bundle of the UCL in an attempt to describe its sonographic characteristics (thickness measurements and qualitative characteristics) in healthy volunteers not engaged in any professional sports. An additional goal was to verify the interobserver variability and correlations of the identified characteristics in cadaveric tissues using the same measurement method for the sonographic and anatomic studies.

The examinations were performed with Sonoline Antares ultrasound equipment (Siemens Medical Solutions, Malvern, PA) equipped with high-resolution broadband linear transducers that had a frequency range of 8 to 14 MHz. The parameters were optimized for musculoskeletal analysis with good near-field focus, and they had been specifically developed by 3 radiologists to study superficial structures. One of the radiologists had 15 years of experience in musculoskeletal sonography; another was a fourthyear radiology resident; and the third was a second-year radiology resident. Both radiology residents were previously trained to correctly evaluate the anterior bundle of the UCL. The examiners performed the sonographic examinations and measurements independently and were blinded to the results of the other examiners. For the examinations, a technique similar to that described by Ward et al11 was used. The participants remained seated in front of the examiner with the forearm in a neutral, flexed position to approximately 80° to 90° and with slight external rotation. The transducer was placed parallel to the ground on the bent elbow (Figure 1) and moved caudally to observe the highest point in the coronoid process of the ulna as well as its medial surface, which Figure 1. Position and examination technique used to obtain a sonogram of the anterior bundle of the UCL. The patient is sitting in front of the examiner, and her forearm is in a neutral and flexed position to approximately 90° with slight lateral rotation (external). The transducer is positioned on the flexed elbow and moved caudally to observe the highest point in the coronoid process of the ulna and its medial surface, which corresponds to the sublime tubercle, where the distal insertion of the ligament can be found. In the proximal region, the transducer is directed to the inferior surface of the medial epicondyle of the humerus.

Materials and Methods Elbow sonographic studies were conducted to evaluate the anterior bundle of the UCL in 48 elbows of 24 healthy asymptomatic volunteers recruited by hospital staff who were not engaged in any professional sports and who had shown no history of trauma, degenerative processes, infections, or rheumatic diseases. Of the 24 volunteers, 16 were male (66.7%), and 8 were female (33.3%), with a mean age of 43.7 years (range, 25–82 years), and all of them were right-handed. The study was approved by the institutional Research Ethics Committee, and all participants signed a written informed consent form. Each of the 48 elbows was examined 3 times, once per examiner, resulting in a total of 144 measurements.

378

J Ultrasound Med 2015; 34:377–384

3403jum363-518 copy_Layout 1 2/20/15 3:14 PM Page 379

Ferreira et al—Sonography of the Anterior Bundle of the Ulnar Collateral Ligament

corresponds to the sublime tubercle, where the proximal portion of the distal insertion of the ligament can be found. In the proximal region, the transducer was directed toward the bottom surface of the medial epicondyle of the humerus, thus obtaining the cutting plane of the anterior bundle of the UCL. The examiners identified the bony structures and the anterior bundle (bottom surface of the medial epicondyle and sublime tubercle) and conducted a long longitudinal ligament plane (coronal plane) assessment to subjectively identify the echographic pattern in the proximal, medial, and distal thirds, characterizing them as hyperechoic or iso/ hypoechoic in comparison to the adjacent musculature. Hyperechoic patterns were defined as internal echoes of the ligament that were more intense than the muscle; isoechogenicity was defined as echoes that were similar to the muscle; and hypoechoic patterns were defined as echoes that were less intense than the echoes from the muscle planes. As a qualitative assessment, the echo texture of the ligament was characterized as homogeneous or heterogeneous. The echo texture was characterized as homogeneous when echoes within the ligament were of similar intensity and had a fibrillar pattern, whereas a heterogeneous pattern consisted of echoes of different intensity within the ligament. The ligament was also assessed for calcifications and whether it was surrounded by fluid. The degree of confidence with which the ligament and its boundaries was identified was scored as 100%, 75%, 50%, 25%, or 0%, corresponding to certain identification, near-certain identification, doubtful identification, possible identification, or impossible identification, respectively. As an objective parameter, the examiners measured the thickness of the anterior bundle 3 times (coronal plane) using a strict protocol (Figure 2). The first measurement point was located in the proximal third, with the insertion of the ligament fibers in the medial epicondyle; the second was located in the medial third; and the third was located on the distal third, immediately proximal and adjacent to the humeroulnar joint. The second part of the study included a sonographic examination of a cadaveric left elbow, which had been fixed in 10% formaldehyde, using a portable SonoSite 180 Plus ultrasound system (FUJIFILM Sonosite, Inc, Bothell, WA) equipped with a linear transducer with a frequency range of 8 to 10 MHz. The sonographic technique for the cadaveric study was the same as the one used for the volunteers. The echogenicity relative to the adjacent muscle, fibrillar pattern (homogeneous or heterogeneous), and anterior bundle shape were evaluated in both the cadaver and volunteers. In addition, the ligament thickness of the cadaveric

J Ultrasound Med 2015; 34:377–384

specimen was measured at the same points as in the volunteers. The confidence level for the ligament and the definition of its boundaries was rated. With a 25 × 7 (25 × 0.70-mm) needle and sonographic guidance, approximately 0.5 mL of methylene blue dye was injected into the cadaveric structure that was previously identified by sonography as the anterior bundle of the UCL. After the injection, the elbow was dissected by an experienced anatomist to confirm that the viewed structure was in fact the anterior bundle and to evaluate its morphologic characteristics (shape and insertion points). In the third part of the study, 4 previously dissected cadaveric elbows from the Laboratory of Descriptive and Topographic Anatomy was examined to evaluate anatomic characteristics of the anterior bundle of the UCL (shape and insertion points). In 1 cadaveric elbow, the anterior bundle thickness was measured in the proximal third, near the humeral insertion, as well as the medial and distal thirds, near its ulnar insertion at the sublime tubercle. For statistical analysis, initially all variables were analyzed descriptively. An inference analysis and analysis of variance were used to estimate the intraclass correlation coefficient (ICC) with a 95% confidence interval. A coefficient of less than 0.4 represents poor reproducibility; a coefficient of 0.75 or greater represents excellent reproducibility; and all other readings represent satisfactory reproducibility.12 To describe the intensity of agreement between 2 or more observers, the κ coefficient was used. P < .05 was considered significant in all analyses.13

Figure 2. Longitudinal sonogram showing the anterior bundle of the UCL (between calipers) in the right elbow of a 48-year-old volunteer. The hyperechogenicity of the anterior bundle in relation to the adjacent flexor muscles (asterisks), its triangular shape, with the apex at the insertion site running along the sublime tubercle of the ulna (arrowheads), and a homogeneous echo texture are shown. D, indicates distal aspect of the elbow; P, proximal aspect; H, cortical surface of the humerus; SC, subcutaneous tissue; thick arrow, humeroulnar joint; and U, cortical surface of the ulna.

379

3403jum363-518 copy_Layout 1 2/20/15 3:14 PM Page 380

Ferreira et al—Sonography of the Anterior Bundle of the Ulnar Collateral Ligament

Results The characteristics of the volunteers are summarized in Table 1. The anterior bundles of the UCLs were identified by all observers in 100% of the cases, resulting in a 96.2% confidence level regarding characterization of the ligament and precise identification of its limits and triangular appearance. The characteristics of the anterior bundle of the UCL regarding echogenicity, echo texture, and additional findings are summarized in Table 2. The samples with iso/ hypoechoic anterior bundles were grouped for κ analysis. Table 1. Characteristics of the 24 Volunteers Included in the Study Characteristic

Value

Age, y Mean ± SD Range Sex, n (%) Male Female Race, n (%) White Nonwhite Height , m Mean ± SD Range Weight, kg Mean ± SD Range Body mass index, kg/m2 Mean ± SD Range Right-hand dominant, n (%)

43.7 ± 14.7 25–82

Additional findings were observed in 13.2% of the elbows, and no additional findings were observed in 86.8%. The mean anterior bundle measurements ± SDs from the right elbow in the proximal, middle, and distal thirds were 4.13 ± 0.81, 2.70 ± 0.58, and 1.42 ± 0.35 mm, respectively. The means, standard deviations, minimums, and maximums of the measurements that were obtained from the volunteers by the 3 investigators are summarized in Table 3. The agreement among the examiners for the anterior bundle measurements in the right elbow was considered excellent in the proximal third (ICC = 0.853) and satisfactory in the middle and distal thirds (ICC = 0.738 and 0.698, respectively). By contrast, in the left elbow, the agreement was considered excellent in the proximal and middle thirds (ICC = 0.890 and 0.773) and satisfactory in the distal third (ICC = 0.717). Overall, the agreement was

16 (66.7) 8 (33.3)

Table 2. Distribution of Qualitative Sonographic Characteristics

19 (79.1) 5 (20.9)

Echogenicity Hyperechoic Iso/hypoechoic Echo texture Homogeneous Heterogeneous Additional findingsa Observed Absent

1.69 ± 0.10 1.50–1.86 76.9 ± 13.4 54–102 27.0 ± 4.4 18.3–37.8 24 (100)

Characteristic

n (%) 139 (96.5) 5 (3.5) 128 (88.9) 16 (11.1) 19 (13.2) 125 (86.8)

aInsertional ligament calcifications, thin layer of intra-articular anechoic liquid, and the presence of a hypoechoic strap superficial to the anterior bundle of the UCL.

Table 3. Anterior Bundle Measurements by Examiner Right Elbow Measurement Examiner 1 Mean, mm SD, mm Minimum, mm Maximum, mm Examiner 2 Mean, mm SD, mm Minimum, mm Maximum, mm Examiner 3 Mean, mm SD, mm Minimum, mm Maximum, mm

380

Left Elbow

Proximal

Middle

Distal

Proximal

Middle

Distal

4.35 0.98 2.40 6.60

2.75 0.68 1.60 4.60

1.40 0.34 0.90 2.20

4.14 0.87 2.80 5.70

2.62 0.72 1.50 4.20

1.44 0.35 0.90 2.30

4.08 0.71 2.30 5.10

2.55 0.52 1.30 3.30

1.23 0.30 0.80 1.90

4.05 0.77 2.80 5.50

2.72 0.46 2.00 3.40

1.35 0.28 1.00 1.90

4.15 0.81 2.60 6.10

2.71 0.51 1.60 3.70

1.51 0.34 0.90 2.20

4.03 0.72 2.60 5.30

2.85 0.59 1.60 3.80

1.59 0.40 0.90 2.30

J Ultrasound Med 2015; 34:377–384

3403jum363-518 copy_Layout 1 2/20/15 3:14 PM Page 381

Ferreira et al—Sonography of the Anterior Bundle of the Ulnar Collateral Ligament

considered excellent for in proximal third (ICC = 0.868) and satisfactory in the middle and distal thirds (ICC = 0.752 and 0.708), with P < .0001 (Table 4). There was no statistically significant variation in the mean anterior bundle measurements in the volunteers at any of the 3 points measured. On sonography, the anterior bundle had a triangular shape (in the coronal plane), with the triangular distal apex located at its insertion site in the sublime tubercle of the ulna (Figure 3). The qualitative agreement among the observers was considered good for echogenicity (κ = 0.617), excellent for echo texture (κ = 0.931), and moderate for additional findings (κ = 0.466), with P < .0001. In the second part of the study, a sonogram of the anterior bundle of the UCL from the left cadaveric elbow was obtained, which was found to be hyperechoic compared to the adjacent muscle. It was also found to have a homogeneous echo texture, a fibrillar pattern, and a triangular shape, with the apex located at the insertion site. Table 4. Agreement Among Examiners for the Anterior Bundle Measurements Elbow Right Proximal Middle Distal Left Proximal Middle Distal Mean Proximal Middle Distal

ICC

95% CI

Agreement

P

0.853 0.738 0.698

0.711–0.931 0.487–0.878 0.409–0.859

Excellent Satisfactory Satisfactory