RESUMO
AIM: To evaluate the accuracy of brachial plexus magnetic resonance imaging (MRI) utilising the three-dimensional (3D)-T2-turbo spin echo (TSE) with 90° flip-back pulse ("DRIVE") myelography in detecting nerve root avulsions in patients with traumatic brachial plexus injuries. MATERIALS AND METHODS: A prospective study of 24 patients planned for brachial plexus reconstructive surgery following trauma from April 2019 to October 2021. Preoperative 1.5 T MRI of the brachial plexus was performed utilising axial T2-DRIVE, looking for signs of avulsions (absent dural rootlets, pauci-rootlet appearance and thickened rootlets; the presence of pseudomeningoceles was noted only as an ancillary sign). Comparison against the reference standard of extra-dural brachial plexus exploration was performed. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy were calculated. RESULTS: Eighteen patients had at least one root avulsion at surgery, five showed isolated post-ganglionic injuries, and one had normal brachial plexus exploration. Thirty-nine avulsed roots were found at surgery (out of 108 explored in 24 patients). Preoperative MRI identified the specific avulsed roots accurately in each patient. Two false-positive diagnoses of C5 and C6 avulsions were made in one patient. On MRI, absence of the rootlets was seen in 73.2% (n=30) of avulsions, pauci-rootlet appearance in 24.4% (n=10) and thickening of the rootlets in 2.4% (n=1). Pseudomeningoceles were found only in 68.3% (n=28) of avulsions. The overall sensitivity, specificity, PPV, NPV, and accuracy of MRI were 100%, 97.1%, 95.1%, 100% and 98.1%, respectively. CONCLUSION: 3D-T2-DRIVE is highly accurate in evaluating pre-ganglionic traumatic brachial plexus injuries. Pseudomeningoceles can be considered an ancillary feature of avulsion given the clarity of rootlet visualisation by this sequence.