Introduction

The ulnar collateral ligament (UCL), the primary restraint against valgus stress of the elbow [1], is a superficial structure amenable to sonographic evaluation. It is composed of an anterior band, a posterior band, and a transverse band, but it is the anterior band that is the clinically relevant component. Injury to this ligament usually occurs as a result of overuse, typically in the throwing athlete, such as baseball pitchers, due to the valgus stress placed on the elbow during the late cocking and early acceleration phases of throwing [2]. The purpose of the study is to describe the sonographic appearances of tears of the ulnar collateral ligament.

Materials and methods

Patients

Eight males, aged 13–35 years (average 18 years) were referred for imaging of suspected ulnar collateral ligament injury. All were non-professional baseball pitchers, and all had been referred by their orthopedic surgeons because of medial elbow pain and tenderness along the course of the UCL; on physical examination, six patients had instability with valgus stress. One patient had had prior surgery to remove osteophytes and calcification from the proximal aspect of the UCL 2 years before, and was having new symptoms.

Five patients were initially referred for magnetic resonance (MR) imaging, one for both MR imaging and sonography, and two for sonography only. The five patients who were referred for MR imaging also agreed to undergo sonographic evaluation, for which they were not charged and which was performed during the same visit as the MR examination.

The time from onset of symptoms to imaging ranged from 1.5 weeks to 6 months.

Imaging

Sonography was performed using either an HDI 3000 unit or HDI 5000 unit (Advanced Technology Laboratories, Bothel, Wash., USA), with 13–7 MHz linear, 12–5 MHz linear, or 10–5 MHz compact linear transducers, or a GE 700 unit (General Electric Medical Systems, Milwaukee, Wisc., USA), with a 13-7 MHz linear transducer. The scans were performed by any one of the authors, all of whom knew the clinical histories and who were aware of the MR imaging findings in four patients. The patients were seated facing the operator, who held their supinated and slightly flexed elbows with one hand and scanned with the other hand. The asymptomatic contralateral elbows were also evaluated. Neither valgus stress nor power Doppler imaging was applied.

MR imaging was performed on a 1.5-T unit (Signa; GE Medical Systems), and consisted of T1-weighted and fat-suppressed fast spin-echo T2-weighted sequences in the coronal, axial, and sagittal planes. Parameters for the T1-weighted images were 500–700/13–15 (TR/TE), and for the T2-weighted images were 2600–5000/48–57 with 8–19 echo train length. Other parameters included 2–3 mm slice thickness with no interslice gap, 256–384×192 matrix, 12–18 FOV, and 1–2 excitations. The patient with the prior UCL repair also underwent MR arthrography, at the surgeon’s request, after the conventional study; in the MR suite, using sterile technique and local anesthesia but without imaging guidance, a 22 gauge hypodermic needle was placed into the radio-capitellar joint, and 8 ml of dilute gadolinium chelate (0.2 ml Omniscan in 50 ml normal saline; Nycomed, Milwaukee, Wisc., USA) were injected. Fat-suppressed T1-weighted sequences 400–500/15 (TR/TE) with other parameters similar to above were then performed in the coronal and axial planes.

Imaging criteria

The sonographic criteria for rupture of the UCL were non-visualization, discontinuity, or detachment of the ligament, with surrounding hypoechoic edema, while sprain consisted of surrounding hypoechoic edema with/without thickening or thinning of the ligament compared to the contralateral normal UCL.

Similarly, MR imaging criteria of rupture included detachment, focal discontinuity, or poor or non-visualization of the ligament, with internal and/or surrounding high signal intensity edema on T2-weighted images, while sprain consisted of thickening or thinning with surrounding high signal intensity edema on T2-weighted images.

The images were reviewed retrospectively by the authors in consensus. Three patients had surgical confirmation of their injuries, with time from imaging to surgery of 2 days to 9 months.

Results

In four patients, the UCL was ruptured from its attachment to the undersurface of the medial epicondyle, manifested sonographically in three cases as discontinuity of the normally hyperechoic ligament with anechoic fluid in the gap (Fig. 1), and in one case as non-visualization of the ligament with heterogeneous echogenic swelling in the expected location of the ligament (Fig. 2). Corresponding coronal fat-suppressed T2-weighted MR imaging sequences demonstrated focal discontinuity of the ligament with high signal intensity fluid in the gap, and complete disruption of the normal ligamentous architecture, respectively. Two of these cases were confirmed surgically.

Fig. 1
figure 1

Twenty-year-old man with surgically confirmed rupture of ulnar collateral ligament, 6 weeks after injury. A Longitudinal sonogram of symptomatic elbow shows discontinuity of the ligament (large arrow) and hypoechoic edema in the gap and around the ligament. Round-tail arrow points to the ulno-humeral articulation. Small arrow points to the ulna. B Longitudinal sonogram of contralateral asymptomatic elbow shows a continuous ulnar collateral ligament (arrows). Ulno-humeral articulation (round-tail arrow). C Coronal fat-suppressed fast spin echo T2-weighted MR image (3700/51, echo train length 8) shows the discontinuous ruptured ligament (arrow) with surrounding high intensity edema

Full size image
Fig. 2
figure 2

Seventeen-year-old male with ruptured ulnar collateral ligament, 2 weeks after injury. Injury was not surgically confirmed, but the patient was unable to return to pitching. A Longitudinal sonogram of symptomatic elbow shows heterogeneous swelling of the common flexor mass (small arrows) and lack of a discernible ligament. Ulno-humeral articulation (round-tail arrow). Ulna (thin long arrow). B Longitudinal sonogram of contralateral asymptomatic elbow shows the ulnar collateral ligament (arrows). Ulno-humeral articulation (round-tail arrow). Ulna (thin long arrow). C Coronal fat-suppressed fast spin echo T2-weighted MR image (4000/48, echo train length 8) shows the markedly swollen and disrupted ligament (black arrow), with internal and surrounding high signal intensity edema/hemorrhage and swelling of the common flexor mass (white arrows)

Full size image

Two patients, 13-year-old and 15-year-old boys, had avulsions of the UCL from the medial epicondyle, with sonographic demonstration of the avulsed echogenic bony fragment in both cases and thickening of the ligament due to interstitial tear (Fig. 3); the 15-year-old underwent surgical repair and both the avulsion and thickening were confirmed.

Fig. 3
figure 3

Thirteen-year-old male with avulsion of ulnar collateral ligament from the medial epicondyle, 10 days after injury. Not surgically confirmed. A Anteroposterior radiograph of symptomatic elbow shows a small avulsed fragment (arrow) from the medial epicondylar apophysis. B Anteroposterior radiograph of contralateral asymptomatic elbow shows a normal apophysis. C Longitudinal sonogram of symptomatic elbow shows the avulsed echogenic bony fragment (large solid white arrow) and the ulnar collateral ligament (black arrows), which is swollen proximally near the avulsed fragment. Open white arrow points to the anechoic cartilage of the humeral trochlear epiphysis. Ulna (thin white arrow). D Longitudinal sonogram of contralateral asymptomatic elbow shows the normal ulnar collateral ligament (black arrows) and the single echogenic epicondylar apophysis (large solid white arrow). Humeral trochlear epiphysis (open white arrow). Ulna (thin white arrow). E and F Sequential coronal fat-suppressed fast spin echo T2-weighted MR images (3000/49, echo train length 19) show the swollen ulnar collateral ligament (black arrows) and avulsed bony fragment (white arrow)

Full size image

One patient had a mild sprain, manifest as mild thickening and decreased echogenicity of the ligament sonographically compared with the contralateral normal ligament, and with mild surrounding hypoechoic edema. Corresponding high signal intensity edema was also present on the coronal T2-weighted MR images (Fig. 4).

Fig. 4
figure 4

Twenty-year-old pitcher with 1.5 weeks of medial pain due to a sprain, not surgically confirmed. A Longitudinal sonogram of symptomatic elbow shows a mildly hypoechoic ligament (solid white arrows) and adjacent hypoechoic soft tissue edema (open arrow). Ulno-humeral articulation (round-tail arrow). B Longitudinal sonogram of contralateral asymptomatic elbow shows the normal echogenic ulnar collateral ligament (arrows) and lack of adjacent edema. Ulno-humeral articulation (round-tail arrow). C Coronal fat-suppressed fast spin echo T2-weighted MR image (4000/57, echo train length 8) shows high signal intensity edema (arrow) adjacent to ligament

Full size image

The eighth patient, who had undergone previous surgical repair of the proximal attachment of the ligament, was examined sonographically prior to MR imaging and had a small partial tear of the deep surface of the distal aspect of the ligament, visualized as anechoic fluid between the deep surface of the ligament and its ulnar attachment. This appearance was confirmed on the MR arthrographic examination as high signal intensity fluid tracking under the ulnar attachment, forming the “T-sign” described by Timmerman et al. [3] (Fig. 5). This patient did not undergo operative treatment.

Fig. 5.
figure 5

Twenty-two-year-old male with 6 months history of pain, 2 years after surgical removal of osteophytes and calcification from ulnar collateral ligament. Imaging shows a partial tear of the deep surface of the ligament, not surgically confirmed. A Longitudinal sonogram of symptomatic elbow shows anechoic fluid (black arrow) between the distal aspect of the ulnar collateral ligament and the ulna. Ulno-humeral articulation (round-tail arrow). B Longitudinal sonogram of contralateral asymptomatic elbow shows the normal ulnar collateral ligament (arrows) inserting directly on the ulna. Ulno-humeral articulation (round-tail arrow). C Coronal fat-suppressed T1-weighted arthrographic image (400/15) shows the high signal intensity contrast (arrow) insinuating itself between the ulnar collateral ligament and the ulna, forming the “T” sign

Full size image

Discussion

The anterior band of the ulnar collateral ligament extends from the undersurface of the medial epicondyle to the sublime tubercle of the ulna. When injured, the patient complains of medial elbow pain and tenderness to palpation along the course of the ligament, but these symptoms can mimic those of medial epicondylitis and ulnar neuritis, and the clinical assessment of valgus instability can be difficult [2]. While not interfering with routine daily activities, injury to the UCL may hinder athletic performance, and a distinction between sprain and rupture is important since a low grade injury may respond to conservative management, while a high grade injury may require surgical reconstruction if the athlete wishes to return to the pre-injury level of performance.

Attempts to assess the status of the ligament based on the amount of medial joint space widening radiographically during valgus stress have had conflicting results. Rijke et al. [4] used stress radiography to evaluate 42 throwing athletes with medial elbow pain and clinical suspicion of UCL injury, and found that an increased joint space width of 0.5 mm or more in the symptomatic side compared with the contralateral side correlated with rupture or large partial tear at surgery. However, 25% of 136 asymptomatic athletes studied by Singh et al. [5] also exhibited 0.5 mm or more of laxity. Similarly, the average difference in radiographic joint space widening between the dominant and non-dominant elbows in 40 asymptomatic elite handball players was 0.41 mm [6]. Thus, UCL laxity per se may not be abnormal in the dominant arm of the throwing athlete.

While a study by Mulligan et al. [7] suggests that the radiographic presence of heterotopic calcification within the UCL is highly associated with complete or partial tear of this ligament, only 6% of the patients with elbow pain in their series had such calcification. Because the majority of UCL injuries are radiographically occult, magnetic resonance is currently the imaging gold standard. The appearances of UCL injury, originally described by Murphy [8] and by Mirowitz and London [9] include thickening of the ligament with increased signal intensity within and around the ligament, poor visualization or poor margin definition of the ligament, detachment or focal discontinuity, and waviness of the ligament.

Sonographically, the normal UCL has a hyperechoic, fibrillar appearance and a fan-like or cord-like shape in the longitudinal plane [10,11]. In Jacobson et al.’s study of the sonographic evaluation of cadaveric elbows, two specimens had hypoechoic disruption of the normal fibrillar architecture, confirmed as tear both by MR arthrography and gross dissection [11].

In our series, complete tears were manifest as either focal discontinuity of the ligament with surrounding hypoechoic edema or complete disruption with non-visualization of the ligament, this latter appearance having also been described by Jacobson and van Holsbeeck [12]. Avulsed bony fragments from the medial apophysis were also visualized in skeletally immature patients, representing the sonographic appearance of “little leaguer’s elbow”, but sonography also demonstrated partial tear of the ligament in these cases. Partial tears in the adults were seen as either thickening of the mildly sprained ligament with decreased internal echogenicity and surrounding hypoechoic edema, best appreciated by comparison with the contralateral asymptomatic elbow, or as a partial detachment from the insertion on the sublime tubercle of the ulna, representing a sonographic “T” sign. The sonographic appearances of these complete and partial tears corresponded to the appearances on MR imaging.

Our series did not have enough sprained thickened ligaments to establish a distinction from normal thickness, but decreased internal echogenicity and surrounding hypoechoic edema should suggest the acutely sprained ligament rather than a ligament thickened by chronic overuse. Sonographic assessment of chronic change of this ligament in asymptomatic collegiate [13] and professional [14] baseball pitchers has documented increased thickness of the UCL in the throwing arm compared to the non-dominant arm. Other reported findings of chronic stress are ligamentous calcification and focal areas of hypoechogenicity within the ligament [14].

Valgus stress was not applied during the sonographic evaluation of any of our patients because a structural abnormality was visualized which corresponded to the patient’s symptoms. Moreover, given normal laxity in asymptomatic athletes as demonstrated radiographically [5,6] and sonographically in both college pitchers [13] and professional pitchers [14], the significance of widening with valgus stress has yet to be determined. Similarly, the ability of power Doppler sonography to distinguish the injured from non-injured UCL has not yet been determined. Power Doppler sonography showed no hyperemia in or around the UCL of two baseball pitchers with ligamentous laxity in De Smet et al.’s [15] series, even though MR imaging demonstrated surrounding edema in one.

Weaknesses of our study include the non-blinded performance of the sonographic examinations, the small number of cases and few surgical confirmations, and its retrospective nature, all of which preclude determination of sensitivity and specificity of sonography, as well as precluding an estimation of interobserver variability for evaluating UCL injury. Nonetheless, we believe that sonography should be the first advanced imaging modality used to evaluate clinically suspected UCL injury, reserving MR imaging for sonographically normal or equivocal cases. Prospective studies comparing the performance of these two modalities will be necessary to determine their accuracy.