Factors associated with radiation toxicity and long-term tumor control more than 10 years after Gamma Knife surgery for non-skull base, nonperioptic benign supratentorial meningiomas.

Objective: Gamma Knife surgery (GKS) is a well-established treatment for benign intracranial meningiomas; however, the dosimetric factors associated with long-term GKS efficacy and safety remain to be elucidated. Using data obtained with at least 10 years of follow-up, the authors aimed 1) to analyze GKS efficacy and safety for the treatment of benign meningiomas confined to non-skull base, nonperioptic supratentorial locations and 2) to determine the radiation dose window that allows for long-term efficacy and safety, namely the minimum dose to achieve long-term local control (LC) and the maximum safe dose to avoid adverse radiation effects (AREs).

Methods: A retrospective analysis was performed on patients who underwent GKS for benign meningiomas in the abovementioned location at the authors' institution between 1998 and 2010 and who received follow-up for more than 10 years. The authors meticulously extracted the values of various dosimetric factors by using a dose-volume histogram. Cox proportional hazard regression analyses were performed to investigate the dosimetric factors associated with LC and ARE.

Results: Fifty-five patients (male/female ratio 1:4.2) with 68 tumors were enrolled. The median (range) gross target volume and marginal dose were 4.2 (0.2-31.7) cm3 and 14.3 (9-20) Gy, respectively. In total, 23.5% of tumors progressed at an average of 72 months, with 10- and 15-year progression-free survival rates of 80.9% and 73.5%, respectively. In univariate analysis, higher marginal dose, coverage (%), Dmin, D98%, Dmean, D2%, Dmax, and Paddick conformity index were significantly associated with LC. In multivariate analysis, D98% was the significant factor, with a cutoff value of 11 Gy (HR 0.754, p < 0.001). Symptomatic AREs occurred in 7 patients at an average of 7 months after GKS. AREs were significantly associated with the volume of normal tissue irradiated with more than 14 Gy (nV14Gy), with a cutoff value of 0.66 cm3 (HR 2.459, p = 0.002).

Conclusions: D98% was a barometer of the minimum required dose associated with long-term LC, and nV14Gy was related to symptomatic AREs. The authors recommend a marginal dose ranging from 11 to 14 Gy to achieve long-term efficacy and safety in patients with non-skull base, nonperioptic benign supratentorial meningiomas, with the assumption of thorough tumor coverage.