Correlation of Ultrasound and Electrodiagnostic Evaluation in Ballistic Peripheral Nerve Injuries

Abstract

Background:

Evaluation and management of ballistic peripheral nerve injuries remain controversial, and recent series have suggested higher rates of nerve discontinuity than previously appreciated. Ultrasound (US) may aid clinicians in the management of ballistic injuries. The goal of this study was to compare US findings to electrodiagnostic and intraoperative findings to assess its accuracy in ballistic injuries.

Methods:

We conducted a retrospective review of patients with the following criteria: (1) ballistic injury to the upper or lower extremity with suspected mixed or motor peripheral nerve injury; (2) underwent electrodiagnostic studies (EDX) and peripheral nerve ultrasound. US findings were categorized as normal, enlarged, neuroma-in-continuity, partial transection, or complete transection. EDX were reviewed for abnormalities in compound motor action potential amplitudes.

Results:

Sixteen patients met our inclusion criteria, of whom 14 had US abnormalities: 8 neuromas-in-continuity, 2 complete transections/discontinuity, 1 partial transection, 2 enlargements, and 1 hypoechoic/fascicular irregularity. US detected 14 of 16 neurapraxic, axonotmetic, or neurotmetic peripheral nerve injuries after ballistic trauma. US had 88% sensitivity, with 0 false positives and 2 false negatives (negative on ultrasound, positive on electrodiagnostic testing) compared with electrodiagnostic testing.

Conclusions:

Our findings suggest that US is an accurate way to evaluate peripheral nerve injuries after ballistic trauma. US may play a role in early diagnostics, especially when EDX are of little value. Future work should focus on the accuracy of early US in ballistic injuries and determining the effects of US and EDX at varying time intervals.

Takeaways

Question: Is ultrasound (US) an accurate way to assess peripheral nerves following ballistic injury?

Findings: This was a retrospective cohort study of patients who sustained an isolated peripheral nerve injury in the setting of a ballistic wound and underwent both US and electrodiagnostic testing. Patients’ US findings were compared with electrodiagnostic and intraoperative findings. US detected 14 of 16 neurapraxic, axonotmetic, or neurotmetic peripheral nerve injuries.

Meaning: US is an accurate way to evaluate peripheral nerves in the setting of ballistic injuries, and its use has a potential role in early diagnostics.

INTRODUCTION

The incidence of ballistic musculoskeletal trauma continues to rise in the United States.^1–3 Prior series have estimated that the frequency of nerve injury after ballistic trauma ranges from 14% to 30%.^4,5 Gunshot wounds can involve a combination of cavitation and shearing forces, leading to a wide spectrum of severity of nerve injuries.⁶ There is continued variability in treatment recommendations for ballistic nerve injuries. Although initial case series and traditional teaching have indicated that sufficient functional recovery will occur in the majority of ballistic nerve injuries,⁷ more recent series have suggested that the frequency of partial or complete neurotmetic or severe axonotmetic injury may be higher than previously appreciated. Recent series by Straszewski et al,⁵ Pannell et al,⁸ and Avery et al⁹ have demonstrated an incidence of partial or complete discontinuity in nerve injuries ranging from 17% to 22% after ballistic trauma. Because these more severe nerve injuries carry a poorer prognosis for recovery, some peripheral nerve surgeons have advocated for earlier surgical intervention.^10,11

Many peripheral nerve surgeons use electrodiagnostic studies (EDX) at 6–12 weeks after nerve injury as an adjunct to clinical examination to estimate the prognosis for spontaneous recovery.^6,12,13 Obtaining EDX before 3–4 weeks has limited utility, as Wallerian degeneration may not have occurred to an extent enough to confidently detect denervation that can be reflected on EDX.^13,14 This window of time has been referred to as the “electrodiagnostic gap.” Given the increased suspicion for partial or complete discontinuity injuries (with a poorer prognosis) and the presence of the electrodiagnostic gap, there has been interest in early imaging of nerves to identify those patients with partial or complete discontinuity injuries that may be better suited for earlier surgical intervention.¹⁴

Ultrasound (US) has become increasingly used in the evaluation of compressive neuropathy given that it (1) is less invasive than electrodiagnostic testing; (2) is less expensive than EDX¹⁵ and magnetic resonance imaging; and (3) has high sensitivity and specificity compared with magnetic resonance imaging in diagnosing peripheral nerve injury.¹⁶ Given that US also has the potential to provide prognostic information during the electrodiagnostic gap,^17,18 we sought to assess the accuracy of US compared with electrodiagnostic and intraoperative findings. We hypothesize that morphological abnormalities on US would be closely associated with motor conduction changes on EDX and intraoperative findings.

MATERIAL AND METHODS

Following approval from our institutional review board, we assembled a cohort of open ballistic nerve injury patients from our peripheral nerve clinic who were older than 18 years of age at the time of injury. Over a 50-month period (June 2018 to August 2022), 689 patients were evaluated for clinically suspected peripheral nerve injury, 69 of whom had ballistic injuries of mixed or motor peripheral nerves. Patients were indicated to receive solely US if seen within 3 weeks of their initial injury. Patients received both US and EDX if seen after 3 weeks, or received the other imaging modality if one had already been performed. We excluded patients who did not receive both US and EDX; patients with brachial plexus, lumbosacral plexus, or sciatic nerve injuries; patients with injuries to multiple peripheral nerves; patients with isolated sensory or digital nerve injuries; and patients with compressive neuropathy. Sixteen patients with ballistic nerve injury were included in our final cohort after applying these inclusion and exclusion criteria (Fig. 1).

Fig. 1.

Flowchart for the study cohort. MTF, male to female transgender.

We obtained results from US (Sonosite X-Porte Scanner, Bothell, WA) and EDX, which were performed concomitantly after the ballistic injury. (See Video 1 [online], which displays a US of the ulnar nerve at the medial epicondyle.) All studies were performed by 1 of 2 physiatrists, both with additional certification in musculoskeletal US and conducted according to American Association of Neuromuscular & Electrodiagnostic Medicine standards.¹⁹ US reports and images were reviewed to evaluate for nerve injury based on the following criteria from Symanski et al²⁰: enlargement of cross-sectional area and side-to-side ratios; partial or complete structural discontinuity; fascicular irregularity; and evidence of neuromas-in-continuity. Neuromas-in-continuity were characterized by the loss of typical fascicular pattern with mixed or hypoechogenicity with well-circumscribed enlargement of the cross-section area of the nerve, typically with a fusiform or ovoid appearance. In discontinuity injuries, bulbous enlargement is present at the end of the parent nerve. Normal nerve architecture should be present adjacent to the neuroma. EDX were reviewed to identify reduced compound motor action potential (CMAP) amplitude and slowing of nerve conduction velocity to denote nerve injury.

Video 1: Video demonstration of an ultrasound of the ulnar nerve at the medial epicondyle.

Download video file ^{(1.8MB, mp4)}

RESULTS

We identified 16 patients who met our inclusion criteria. There were 13 men and 3 women in the cohort, with a mean age of 28.5 years (range: 19.3–44.7 y). The median time from injury to US + EDX evaluation was 4.1 months (range: 0.2–14.2 mo). Within our cohort of 16 patients, 9 (56.3%) patients had concurrent adjacent fractures and 1 (6.3%) had concurrent adjacent vascular injuries that required surgical intervention. Eleven patients underwent surgical intervention for their nerve injury, 3 patients had clinical improvement without surgery noted in subsequent follow-up visits, and 2 patients were lost to follow-up after initially being treated conservatively. (See table, Supplemental Digital Content 1, which displays the comparison of US findings to intraoperative findings, https://links.lww.com/PRSGO/E92.)

Of those 16 patients with nerve injury detected on EDX, there were 2 median, 4 radial, 6 ulnar, 3 peroneal, and 1 tibial nerve injuries. The mean ± SD CMAP amplitude was 14% ± 36% of normal (range not recorded/0% to 124% of normal). Among the 16 patients, 14 had US abnormalities, including 2 complete transections/discontinuity and 1 partial transection (1 example of a discontinuity injury is seen in Figs. 2, 3). Of the 14 patients with US abnormalities, there were 13 with abnormal CMAP motor amplitudes (either not recordable or decreased compared with normal values; reflecting an axonotmetic or neurotmetic injury) and 1 had normal CMAP motor amplitudes but had delayed conduction velocity (reflecting a neurapraxic injury). There were no false positives on US (Table 1). Of the 2 patients (12.5%) with no US abnormalities, both had completely absent CMAP motor amplitudes (false negative on ultrasound). Among these, 1 had a radial nerve injury but no noted enlargement of the radial nerve on ultrasound. They were diagnosed with subacute radial neuropathy proximal to the 3 heads of the triceps brachii, but the patient did not return to the clinic for follow-up. The second patient had a peroneal nerve injury with borderline enlargement of the peroneal nerve at the fibular neck, but the enlargement was not large enough to meet diagnostic criteria based on the criteria in Symanski et al.²⁰ This patient underwent surgical decompression of the common peroneal nerve at the fibular neck for their neuropathy.

Fig. 2.

Example of a patient with discontinuity injury of the deep peroneal nerve after a ballistic injury. Radiograph showing adjacent fibula fracture (A), preoperative US demonstrating short-axis views of the common peroneal nerve proximal to the nerve injury (B), at the level of the nerve injury (C), and distal to the nerve injury (D). Operative exploration confirmed discontinuity injury of the deep peroneal nerve.

Fig. 3.

Depiction of the following nerves: common peroneal nerve, deep peroneal nerve (DPN), and superficial peroneal nerve (SPN).

Table 1.

Specificity and Sensitivity of US Detection Against Electrodiagnostics

	EDX Positive	EDX Negative
US positive	14	0
US negative	2	0

DISCUSSION

In our series, US detected 14 of 16 neurapraxic, axonotmetic, or neurotmetic peripheral nerve injuries after ballistic trauma, with 88% sensitivity compared with electrodiagnostic testing. This suggests that US is an accurate diagnostic tool that may aid physicians in the management of ballistic injuries.

Among our patients, US findings demonstrated 8 neuromas-in-continuity (axonotmetic injuries), 2 complete nerve transections (neurotmetic injuries), and 1 partial nerve transection. Complete or partial nerve transections were, therefore, present in 19% (3 of 16) of the patients with ballistic nerve injuries. This incidence is consistent with recent reports by Pannell et al⁸ (17% neurotmesis rate if nerve exploration performed), Straszewski et al⁵ (21% neurotmesis rate if nerve exploration performed), and Avery et al⁹ (22% neurotmesis rate seen with nerve exploration or ultrasound), and slightly higher than what has been historically reported by Omer⁷ (14% nerve laceration rate). Based on the more recent published literature, we believe that there should be a heightened suspicion for partial or complete nerve discontinuity in patients with a neurological deficit after ballistic trauma. Earlier identification of injuries with a poorer prognosis can aid surgeons in determining how to guide subsequent treatment,^10,17,18 whether that is through early exploration and reconstruction or alternative strategies such as distal nerve transfers or tendon transfers (or any combination of these) based on distance from the nerve injury to the target muscle.

Given that US can be performed within the first weeks following injury, US has the potential to have an initial role in the evaluation of patients with suspected peripheral nerve injury after ballistic trauma, particularly immediately after injury when EDX are of limited utility (the aforementioned electrodiagnostic gap). In our series of 16 patients, 4 (25%) patients may have potentially benefitted from early surgical intervention given the disruption (complete/partial transection or fascicular irregularities) found on US and later confirmed in intraoperative findings. As such, US may play a role in expediting decisions for early intervention by rapidly and accurately identifying the presence or absence of discontinuity injuries given its high sensitivity, as demonstrated in our series. A positive finding on US may enable surgeons to bypass electrodiagnostics and proceed directly to operative management if indicated. US may also help in counseling patients given that many patients with ballistic in-continuity nerve injuries are likely to recover without surgical intervention, as suggested previously by Omer²¹ and reinforced recently by Straszewski et al.⁵ An important area of investigation as US resolution improves will be the ability to discern among the different grades of axonotmetic injury (specifically Sunderland III and IV), ideally producing imaging that can guide those trying to prognosticate whether injuries will improve with observation or to determine the timing of surgical intervention.

Furthermore, there may be a role in using US as a manner of tracking nerve regeneration as an adjunct or potential replacement to serial EDX, as US may have the potential to detect fascicular irregularities early in recovery.¹⁴ Future work should focus on the accuracy of US during the electrodiagnostic gap, determining the respective role of US and EDX at varying time intervals when assessing patients with peripheral nerve injury, and the potential for serial assessment using US to predict spontaneous recovery.

However, there are still limitations with the usage of US in the diagnosis of peripheral nerve injury. First, accuracy, reliability, and interpretation of US images are user dependent, which may influence how US is used. Similarly, as seen with the false negatives in our group, US may not be able to detect all nerve injuries, particularly those without alterations in the macrostructure of the nerve (such as demyelinating or low-grade axonotmetic injuries). As such, we suggest using US in conjunction with clinical findings and EDX. It has been our practice to use the following algorithmic approach to the diagnosis of peripheral nerve injuries, in which ballistic injuries are evaluated sonographically as soon as possible after injury (unless direct inspection has been performed as part of concomitant fracture or open wound care) (Fig. 4). If there is a discontinuity injury seen on the early US, we consider early operative intervention. If there is no discontinuity injury seen on the early US, we continue with clinical observation and consider EDX at 6–12 weeks after injury.

Fig. 4.

Diagnostic algorithm for peripheral nerve injuries.

Our study is not without limitations. One limitation of the current study was the retrospective nature and small sample size, which may limit the power of our results. Another limitation was potential selection bias in solely including patients with both US and EDX, which may skew toward patients with more severe injuries and bias results toward a higher sensitivity. However, obtaining both US and EDX is the standard of care in our practice. Similarly, the heterogeneity of nerve injuries, whether in the mechanism of injury, the nerve injured, or the injury presentation, may limit the generalizability of the results. This study also does not include clinical outcomes, which would be of great value.

Our experience reflects an increasing enthusiasm to use US in the evaluation of ballistic nerve injuries. Given its high sensitivity and close association with EDX findings, we believe that US may have a specific role in the early evaluation of ballistic nerve injury to aid in decision-making and patient counseling. There is also an opportunity to investigate the role of US in the serial evaluation of nerve injuries to gain a better understanding of which nerve injuries may improve with observation and which injuries may be better treated surgically.

DISCLOSURE

The authors have no financial interest to declare in relation to the content of this article.