Radiologic and Neurosurgical Diagnosis of Arterial Neurovascular Conflict on Magnetic Resonance Imaging for Trigeminal Neuralgia in Routine Clinical Practice

Objective: The most common cause of trigeminal neuralgia is neurovascular conflict, particularly arterial compression of the trigeminal nerve (ACTN). It is possible to show this condition preoperatively on fine-cut constructive interference in steady state magnetic resonance imaging (MRI), supplemented by time-of-flight magnetic resonance angiography. We have noticed significant variability in the interpretation of these studies between radiologists and the treating neurosurgeon. We have assessed the sensitivity and specificity of these 2 styles of interpretation compared with the intraoperative observations. Methods: We studied 68 patients who underwent de novo microvascular decompression from 2011 to 2018 under the care of a single neurosurgeon in Melbourne, Australia. Data was recorded prospectively in the radiology reports, operation reports, hospital admission records and neurosurgeon correspondence from the perioperative clinic reviews. In particular, the surgical interpretation of the MRI was clearly described prospectively and preoperatively in the correspondence. The presence or absence of ACTN was recorded prospectively in the operation report. These data were collated retrospectively by the first author. Results: Of patients, 83.8% (57/68) had clear ACTN confirmed at surgery. Radiologists detected this abnormality in 50.9% (29/57) of these patients, with a sensitivity of 50.9% and specificity of 81.8%. The operating neurosurgeon detected ACTN in 87.7% (50/57) of the positive cases with a sensitivity of 87.7% and a specificity of 72.7%. Statistical analysis showed a significant disagreement between both styles of interpretation, with a radiologic accuracy of 55.9% compared with 85.3% neurosurgically (P < 0.0001). Follow-up review of the patient's response to surgery further supported the neurosurgical diagnosis of ACTN. Conclusions: There was a strong tendency for radiologists to underreport ACTN on the preoperative MRI compared with the intraoperative findings. In this series, the neurosurgeon was able to accurately detect ACTN in 88% of patients preoperatively. However, the diagnosis of absent ACTN was still associated with a false-negative rate of 46.7% when the neurosurgeon performed the interpretation. At the current standard of fine-cut constructive interference in steady state MRI in Melbourne, underdetection of ACTN remains common. Clinicians using this test need to be acutely aware of its limitations in deciding whether to proceed to microvascular decompression.