LAUSR

BACKGROUND Microvascular decompression (MVD) can be an effective intervention for trigeminal neuralgia (TN); however, an optimal system for patient selection and surgical outcome prediction has not been defined. OBJECTIVE To develop and validate a preoperative TN grading system for the prediction of long-term pain relief after MVD. METHODS This retrospective cohort study included consecutive patients suffering unilateral TN who underwent MVD with >18-mo follow-up. A grading system was formulated using 3 previously validated preoperative characteristics. The primary end-point was long-term, pain-free status without use of medication. Ability to predict pain-free status was analyzed by multiple regression and assessed by area under the receiver operating characteristic curve (AUC). Clinical utility to predict MVD success and reduce unnecessary surgeries was assessed by decision-curve analysis. RESULTS Of 208 patients analyzed, 73% were pain-free without medication at >18-mo follow-up. Pain-free status was predicted by classical TN type, positive response to carbamazepine and/or oxcarbazepine, and presence and nature of neurovascular compression demonstrated on MRI (all P < .01). The TN grading system demonstrated good discriminatory ability for prediction of pain-free status (AUC 0.85, 95% CI 0.80-0.91). Decision-curve analysis demonstrated a net reduction of 20 cases likely to be unsuccessful per 100 patients evaluated with this grading system above a decision threshold of 80%. CONCLUSION This TN grading system reliably predicts long-term pain-free status without medications following MVD. The use of the TN grading system as part of a comprehensive work-up may reduce the number of unsuccessful operations.