Predictability of vascular conflict by MRI in trigeminal neuralgia.

OBJECTIVE: Magnetic resonance imaging (MRI) has been investigated extensively in its success or failure to identify preoperative vascular compression in patients with trigeminal neuralgia (TN). To this end, we reviewed our case load to evaluate the concordance or discordance between preoperative MRI and intraoperative findings. PATIENTS AND METHODS: Sixty-nine patients with Type 1 TN and retrievable MRI images, operative reports, and intraoperative photographs were retrospectively reviewed. RESULTS: Our review shows that MRI predicted conflict (arterial or venous) in 58 cases that was confirmed at surgery in 55 cases. MRI predicted no conflict in 11 cases, whereas surgery revealed no conflict in a total of 6 cases. Thus, in predicting conflict at surgery, MRI had a sensitivity of 87%, and specificity of 50%, respectively. Conversely, MRI accurately predicted intraoperative conflict (positive predictive value) in 95% of cases, and the absence of conflict (negative predictive value) in 27%. These results reveal that MRI is more accurate in predicting conflict than the absence of conflict at surgery. CONCLUSION: Our results support the reliance on the clinical diagnosis of Type 1 TN to recommend microvascular decompression (MVD). The presence of vascular compression by MRI should encourage the surgeon to persevere in search of the offending vessel when it proves elusive. MRI positive and negative predictive values for conflict are expected to increase with better resolution imaging. The absence of neurovascular conflict on high-resolution MRI should not negate MVD in the treatment of a patient with classic TN.