Objective: Adequate radiation doses for repeat gamma-knife radiosurgery (GKS) for trigeminal neuralgia is investigated in our series and meta-analysis.
Methods: Fourteen patients treated by ipsilateral repeat GKS for trigeminal neuralgia were included. Median age of patients were 65 (range 28 – 78). Median target dose of the first GKS was 85 (range 70 – 90, within 100% isodose line). Median target dose of the second GKS was 85 (range 50 – 90). Median cumulative dose was 170 (range 140 – 180). Patients were followed median 10.8 months (1 – 151) after the second gamma-knife surgery. Brainstem dose analysis and vote-counting meta-analysis of 19 studies were performed.
Results: After the second gamma-knife radiosurgeries, pain was relieved effectively in 12 patients (86 %) (Barrow Neurological Institute Pain Intensity Score (BNI) I – III). Post-gamma-knife radiosurgery trigeminal nerve deficits were mild in five patients. No serious anesthesia dolorosa was occurred. The second GKS radiation dose ≤ 60 Gy was significantly associated with worse pain control outcome (p = 0.018 in our series, permutation ANOVA, and p = 0.009 in the meta-analysis, two-tailed Fisher’s exact test). Cumulative dose ≤ 140 – 150 Gy was significantly associated with poor pain control outcome (p = 0.033 in our series and p = 0.013 in the meta-analysis, two-tailed Fisher’s exact test).) A cumulative brainstem edge dose > 12 Gy tended to be associated with trigeminal nerve deficit (p = 0.077).
Conclusion: Our study suggests that the second GKS dose is a potentially important factor.
Fig. 1 Comparison between low dose group (≤ 60 Gy) and high dose group (> 60 Gy) in the second GKS in relation to pain control outcome (a). Low dose group showed significantly worse outcome than high dose group (p = 0.01 in the Fisher’s exact test) (b). Comparison between the long inter-treatment interval group (≥ 70 months) and the short inter-treatment interval group (< 70 months) (c). Long treatment interval group tended to have worse pain control outcome (p = 0.066 in the Fisher’s exact test). Two patients with BNI IV outcome had both low radiation dose (≤ 60 Gy) and long treatment interval ≥ 70 months (c). In two-way permutation ANOVA analyzing effects of both factors together by the resampling method, the radiation dose (p = 0.018) in the second GKS was the only significant factor and the inter-treatment interval (p = 0.71) was not significant.
Fig. 2 Meta-analysis by the vote counting method. One previous study was counted as a data point in the scatter diagram (a, b). All studies with the mean or median second GKS dose > 60 Gy showed acceptable pain control percentage greater than 80 % (a). The studies with the mean or mean second GKS dose ≤ 60 Gy had worse pain control outcome (acceptable pain control percentage < 80 %) more frequently (five of seven studies, 71.4 %) (a). This difference was statistically significant (p = 0.001, two-tailed Fisher’s exact test) (a). The cumulative dose > 140 Gy was also significantly related with higher acceptable pain control percentage ≥ 80 % (p = 0.003, two-tailed Fisher’s exact test) (b). The studies with the cumulative GKS dose ≤ 140 Gy had worse pain control outcome (acceptable pain control percentage < 80 %) more frequently (five of eight studies, 71.4 %).