Ultrasound assessment of extensor pollicis longus tendon rupture following distal radius fracture: a sonographic and surgical correlation

Extensor pollicis longus (EPL) tendon ruptures occur more commonly after undisplaced or minimally displaced distal radius fractures^(1–3). Rupture of the EPL tendon can develop without antecedent trauma, being reported in patients with rheumatic disease, following systemic or local steroid injection^(4,5) and in patients with carpometacarpal osteoarthritis( ⁶ ). Cortical irregularities and erosions result in roughening of the bone surface at the level of Lister’s tubercle, which contributes to progressive attrition and consequent rupture of the EPL tendon( ⁷ ). Excessive and repetitive wrist motion in certain occupational groups can result in tendon friction against Lister’s tubercle, predisposing it to rupture^(8–10). Iatrogenic tendon irritation can be caused by prominent screws after volar plate fixation( ¹¹ ), against a dorsal plate or the pins of an external fixator( ⁵ ). Theories on the pathogenesis of EPL ruptures have been postulated in the literature to account for these clinical settings, which describe the interplay of anatomical, mechanical and vascular factors, making the tendon inherently vulnerable to rupture^(1,3).

Ultrasonography (US) plays an important role in the assessment for the presence of EPL tendon ruptures. US depicts the tendon in real time and allows rapid assessment of its integrity( ¹² ), thereby providing a prompt diagnosis for the surgeon to plan their most appropriate course of action. The location of the incision is dependent on the level of the retracted ends, and this can be assessed on US to enable surgical planning. US can also influence the operative technique chosen by the surgeon, such that a direct primary repair of the tendon may not be possible if the tendon is found to be severely retracted or if the ends are atrophied or degenerated. A free tendon graft after harvesting palmaris longus if present, may then be used to augment a direct repair or bridge retracted ends. Alternatively, the tendon transfer technique which involves using extensor indicis to replace a non-salvageable EPL tendon usually confers a predictable outcome^(13,14).

Ultrasound evaluation of the EPL tendon can be technically challenging for the sonographer due to the complex anatomy of the EPL within the extensor compartment and the importance of getting the diagnosis right for the patient and surgeon. In this paper, we describe the sonographic features of EPL rupture that can be identified, through a series of eleven cases which were correlated with surgical findings.

Written consent was obtained from the patients prior to submission for publication. A retrospective case series review was performed in 11 patients over a 2-year period at the John Radcliffe Hospital in Oxford, United Kingdom. Patients included were those referred from both orthopedic and plastic surgery specialties with suspected EPL tendon ruptures. For each patient, their age, gender and presumed cause of EPL rupture (trauma, non-traumatic and spontaneous causes) were recorded.

The ultrasound examinations were performed by a consultant musculoskeletal radiologist, with 20-year experience in musculoskeletal ultrasound. The procedure was performed using a 12 MHz high-frequency linear-array transducer from a GE Logiq E9 ultrasound machine.

The examination commenced with both wrists palmar side down on the table, with the normal side acting as a control. The transducer was applied in the axial plane over the dorsal surface of the distal radius and the bony landmark of Lister’s tubercle was identified. The EPL tendon within the 3^rd compartment was found on the ulnar side of Lister’s tubercle. The tendon was followed proximally to the myotendinous junction and then distally along its full length as its curves round Lister’s tubercle and crosses the tendons of the second compartment, terminating at its insertion onto the base of the distal phalanx of the thumb.

The findings identified on ultrasound were then correlated surgically using the operation note found in the patient’s electronic record.

Eleven patients were referred to radiology from the orthopedic and plastic surgery specialties. The average age of our patients was 51 years (age range 23–71 years). In this cohort of patients, 8 were female and 3 were male. The average interval between the US examination and the initial injury was 4.5 weeks. A total of 8 patients out of the 11 proceeded to operative intervention. The rest either refused surgical treatment or their function had improved following conservative management and did not require surgery (Tab. 1).

No visible tendon was seen in extensor compartment III at the level of Lister’s tubercle in all 11 cases (Tab. 2). The average gap length was measured as 2.4 cm (gap range 1.4–3.6 cm). The most common (10 patients) sonographic finding was effusion seen in the compartment III sheath (Fig. 1). There were 9 patients who had fluid also in the sheath of the extensor compartment II. An empty tendon sheath was observed in 1 patient (Fig. 2). Evidence of associated tenosynovitis of the EPL sheath was demonstrated in 2 cases with increased vascularity seen on Doppler interrogation (Fig. 3). In 9 patients, sonography revealed the retracted tendon ends as enlarged and hypoechoic with loss of their normal fibrillar appearance (Fig. 4). None of the cases demonstrated atrophic or degenerated changes of the tendon ends. In the 8 cases that proceeded to operative intervention, all of these were confirmed as EPL rupture. In 1 patient, the retraction gap was measured as 8 cm intra-operatively but was underestimated by US which measured the gap length as 3.6 cm.

Fig. 1.

Fig. 2.

Fig. 3.

Fig. 4.

Of the 8 patients who sustained a fracture of the distal radius, radiography demonstrated the fracture line extending into Lister’s tubercle in only half of these patients. In contrast, US depicted fracture involvement of Lister’s tubercle (Fig. 5) in all of 8 cases and even identified a fracture of Lister’s tubercle in 1 patient who had no visible fracture on the initial radiograph (Tab. 3).

Fig. 5.

In 1 patient, Lister’s tubercle was seen as a variant in which both the radial and ulnar peaks were of similar heights (Fig. 6).

Fig. 6.

Sonography identified complete rupture of the EPL tendon proximally at the level of Lister’s tubercle in all patients, which was later supported by operative findings. De Maeseneer et al. (2008)⁽¹⁵⁾ performed a sonographic correlation with anatomic features of the EPL tendon ruptures in 5 cadavers. They found a characteristic hypoechoic tubular area at the expected position of the EPL tendon, which corresponded with a dissection to the synovial sheath being distended by fluid, hemorrhage or scarring. In our study, this was not a consistent feature and is in keeping with Santiago et al. (2008)⁽¹⁶⁾, who showed that tendon rupture tended to be illustrated by the absence of the tendon or tendon hypoattenuation.

Following confirmation of EPL rupture, sonography was able to identify the extent of proximal and distal tendon end retraction in all the cases. In evaluating the state of the retracted tendon ends, none of the cases revealed an atrophic appearance, which was not an expected finding as most cases were scanned less than 4 weeks following presentation. However, in many of our patients, the tendon ends appeared as thickened and hypoechoic stumps, which is a feature supported in previous literature⁽¹⁶⁾.

In our study, the ruptured EPL tendon sheath frequently contained a significant amount of fluid, and in several cases there was evidence of sheath thickening and neovascularity consistent with post-traumatic tenosynovitis. These changes are likely accounted for by a secondary inflammatory response following rupture. The current study also frequently identified fluid within the 2^nd extensor compartment sheath, a finding supported by previous studies which have described a foraminal intercommunication between the sheaths of the EPL and extensor carpi radialis tendons^(1,17,18).

In most of the cases, EPL tendon rupture was associated with an undisplaced or minimally displaced distal radius fracture. Several previous studies^(1–3) have also found this association. It has been postulated that rupture ensues from increased pressure within the sheath that impedes blood flow in an already tenuous vascular part of the tendon adjacent to Lister’s tubercle⁽¹⁾. Helal et al. (1982) propose a similar theory, suggesting that an intact extensor retinaculum holding the EPL tendon closely against the fracture callus can lead to attritional rupture⁽²⁾.

In our study, sonography consistently identified an irregular cortical surface of Lister’s tubercle. Fracture lines extending into Lister’s tubercle were clearly demonstrated in all patients with distal radius fractures, and in one case US identified an occult fracture not depicted radiographically. The finding of an irregular Lister’s tubercle supports the mechanical attrition theory for EPL rupture. Furthermore, this is an essential finding that should be communicated to the surgeon due to the implication of potential future risk to the tendon graft or tendon transfer, which may prompt the surgeon to smooth down the tubercle as a preventative measure.

Sonography incidentally identified an anatomical variant of Lister’s tubercle in one patient, where both the ulnar and radial peaks were of similar height and size. Chan and Chong (2017) proposed a new system for anatomically classifying variants of Lister’s tubercle based on size and morphology of its radial and ulnar peaks, which in our study would conform to the ‘Type 2 variant’. They suggest the ‘Type 2 variant’ predisposes the EPL to inherent microinstability and may result in increased risk of tendon rupture. Variant anatomy can also have surgical implications for pre-procedural planning of volar plate fixation, highlighting the importance of making a qualitative assessment of the morphology of Lister’s tubercle⁽¹⁹⁾.

Several limitations were associated with the current study. As this was a retrospective study, the ultrasound examinations were performed only by a single operator. We found that US underestimated the extent of the retraction gap in one patient (3.6 cm) compared with surgical findings (8 cm); however, there was a delay of 8 weeks between the scan and the operation. Ideally, the retraction gap determined on US would have been correlated with the extent of retraction found operatively in all patients. However, surgeons do not routinely report the retraction gap in their operation notes. In addition, three patients did not proceed to operative repair, rendering it difficult to correlate with the sonographic determination of EPL rupture. Furthermore, the number of patients included in the study was small, which is a limitation also shared by other previous studies in the literature and reflects the fact that EPL rupture is not a common complication (0.3%)⁽²⁰⁾.

Ultrasound is an excellent modality in detecting EPL tendon ruptures, enabling the operator to assess the extent of tendon end retraction, state of the tendon ends, concomitant tendon changes of the EPL tendon and the neighboring extensor carpi radialis tendons. Fractures extending into Lister’s tubercle resulting in a sharp, jagged cortex were frequently identified on ultrasound in our series of patients and have significant surgical implications.