Diagnostic performance of preoperative ultrasound for traumatic brachial plexus root injury: A comparison study with an electrophysiology study

Abstract

Purpose: Accurate preoperative assessment for traumatic brachial plexus injury (BPI) is critical for clinicians to establish a treatment plan. The objective of this study was to investigate the diagnostic performance of preoperative ultrasound (US) through comparison with an electrophysiology study (EPS) in the assessment of traumatic brachial plexus (BP) root injury.

Materials and methods: We performed a retrospective study in patients with traumatic BPI who had preoperative US and EPS, excluding obstetric palsy and other nontraumatic neuropathies. US examination was performed on an EPIQ 5 color Doppler equipment. EPS was performed on a Keypoint 9033A07 Electromyograph/Evoked Potentials Equipment, testing electromyography (EMG), nerve conduction studies (NCS), and somatosensory evoked potentials (SEP). Each BP root of all patients was assessed by US and EPS as completely injured or incompletely injured, respectively. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy were calculated based on the correlation with intraoperative findings. The accuracy of US and EPS were compared using the McNemar test. The added benefit of US was evaluated by comparing the sensitivity and specificity between the combined tests with EPS using the McNemar test.

Results: This study included 49 patients with traumatic BPI who underwent BP surgeries from October 2018 to September 2022. Surgical exploration confirmed 89 completely injured BP roots in 28 patients. US correctly detected 80 completely injured BP roots (sensitivity, 0.899; specificity, 0.981; PPV, 0.964; NPV, 0.944; accuracy, 0.951). EPS correctly detected 75 completely injured BP roots (sensitivity, 0.843; specificity, 0.929; PPV, 0.872; NPV, 0.912; accuracy, 0.898). US showed significantly higher accuracy than EPS (p = 0.03). When combining US and EPS for completely injured BP root detection, the sensitivity of the inclusive combination (0.966) was significantly higher than EPS alone (p = 0.000977), and the specificity of the exclusive combination (1.000) was significantly higher than EPS alone (p = 0.000977).

Conclusion: Preoperative US is an effective diagnostic tool in the assessment of traumatic BP root injury. US had higher accuracy than EPS in this study. Sensitivity and specificity were significantly higher than EPS when US was combined with EPS.

Keywords: brachial plexus; comparison; electrophysiology; root injury; trauma; ultrasound.

Figure 1

Longitudinal sonogram shows a hypoechoic irregular neural stump (arrows) with a blind-ending of the hypoechoic brachial plexus root. An empty neural foramen (arrowhead) is visualized, indicating preganglionic avulsion of the nerve root. Transverse processes (P) of cervical vertebrae are shown as hyperechoic bone prominences with posterior acoustic shadowing.

Figure 2

Longitudinal sonogram shows a completely transected brachial plexus root with a 1.92-cm-long neural gap between the proximal end (arrowheads) and the distal end (arrows). The proximal root end is adjacent to the transverse process (P). The caliber of the distal root end is enlarged in comparison with the proximal root end.

Figure 3

Transverse sonogram shows the accumulated cerebrospinal fluid, i.e., pseudomeningocele, adjacent to the intervertebral foramen. The cerebrospinal fluid (arrows) appears anechoic. The absence of brachial plexus root (arrowhead) between anterior (AT) and posterior tubercles (PT) represents preganglionic root avulsion.

Figure 4

Longitudinal sonogram shows a neuroma-like enlargement of a brachial plexus root. The proximal part (arrowheads) of the nerve root adjacent to the transverse process (P) appears as a thin stripe. The distal part (arrows) of the root is hypoechoic and significantly enlarged as a neuroma losing smooth contour and regularly arranged fascicles due to retraction.