Gamma knife radiosurgery for trigeminal neuralgia: the initial experience of The Barrow Neurological Institute.

PURPOSE: To assess the efficacy and complications of Gamma Knife radiosurgery for trigeminal neuralgia. METHODS AND MATERIALS: The Barrow Neurological Institute (BNI) Gamma Knife facility has been operational since March 17, 1997. A total of 557 patients have been treated, 89 for trigeminal neuralgia (TN). This report includes the first 54 TN patients with follow-up exceeding 3 months. Patients were treated with Gamma Knife stereotactic radiosurgery (RS) in uniform fashion according to two sequential protocols. The first 41 patients received 35 Gy prescribed to the 50% isodose via a single 4-mm isocenter targeting the ipsilateral trigeminal nerve adjacent to the pons. The dose was increased to 40 Gy for the remaining 13 patients; however, the other parameters were unvaried. Outcome was evaluated by each patient using a standardized questionnaire. Pain before and after RS was scored as level I-IV per our newly-developed BNI pain intensity scoring criteria (I: no pain; II: occasional pain, not requiring medication; III: some pain, controlled with medication; IV: some pain, not controlled with medication; V: severe pain/no pain relief). Complications, limited to mild facial numbness, were similarly graded by a BNI scoring system.
RESULTS: Among our 54 TN patients, 52 experienced pain relief, BNI score I in 19 (35%), II in 3 (6%), III in 26 (48%), and IV in 4 (7%). Two patients (4%) reported no relief (BNI score V). Median follow-up was 12 months (range 3-28). Median time to onset of pain relief was 15 days (range 0-192), and to maximal relief 63 days (range 0-253). Seventeen (31%) noted immediate improvement (</= 24 h). Prior to RS, all patients were on pharmacologic therapy felt to be optimal or maximal. Twenty-two (41%) were able to stop medications entirely (BNI score I or II). Another 16 (30%), with BNI Score III relief, decreased medication intake by at least 50%. Patients with classical TN pain symptoms were more likely to stop medications than those with atypical features, 49% (21 of 43) versus 9% (1 of 11). This difference was significant at p = 0.040. Statistically, the finding most predictive for pain relief was new facial numbness following RS. Each of the 5 patients with new numbness after RS developed BNI score I relief, contrasting with 35% for the 49 patients with no new numbness (p = 0.019). Complications have been limited to delayed, mild facial sensory loss. Before RS, 17 patients had numbness from prior invasive procedures, none of whom reported a worse numbness score after treatment. Thirty-seven patients had no facial numbness at the time of RS, of whom 5 developed facial hypesthesia. Each rated this as "mild, not bothersome." There have been no other sequellae.
CONCLUSION: RS is an effective treatment, and is the least invasive nonpharmacologic therapy for TN. It carries a small risk of mild facial hypesthesia, a side effect which, somewhat ironically, may be desirable, because it appears to correlate favorably with an excellent pain response. We currently include radiosurgery among the appropriate options for TN patients who have failed optimal medical management, with or without prior invasive neurosurgical procedures. We present here BNI scoring systems for pain intensity and facial numbness. These have proved simple and reliable, have facilitated data collection, rendered analysis more objective, and improved our ability to discuss results with patients and colleagues.