A dosimetric comparison of linear accelerator (LA)-based (BrainLAB) and robotic radiosurgery (RS) (CyberKnife) systems for acoustic schwannoma (Acoustic neuroma, AN) was carried out. Seven patients with radiologically confirmed unilateral AN were planned with both an LA-based (BrainLAB) and robotic RS (CyberKnife) system using the same computed tomography (CT) dataset and contours. Gross tumour volume (GTV) was contoured on post-contrast magnetic resonance imaging (MRI) scan [planning target volume (PTV) margin 2 mm]. Planning and calculation were done with appropriate calculation algorithms. The prescribed isodose in both systems was considered adequate to cover at least 95% of the contoured target. Plan evaluations were done by examining the target coverage by the prescribed isodose line, and high- and low-dose volumes. Isodose plans and dose volume histograms generated by the two systems were compared. There was no statistically significant difference between the contoured volumes between the systems. Tumour volumes ranged from 380 to 3,100 mm(3). Dose prescription was 13-15 Gy in single fraction (median prescribed isodose 85%). There were no significant differences in conformity index (CI) (0.53 versus 0.58; P = 0.225), maximum brainstem dose (4.9 versus 4.7 Gy; P = 0.935), 2.5-Gy volume (39.9 versus 52.3 cc; P = 0.238) or 5-Gy volume (11.8 versus 16.8 cc; P = 0.129) between BrainLAB and CyberKnife system plans. There were statistically significant differences in organs at risk (OAR) doses, such as mean cochlear dose (6.9 versus 5.4 Gy; P = 0.001), mean mesial temporal dose (2.6 versus 1.7 Gy; P = 0.07) and high-dose (10 Gy) volume (3.2 versus 5.2 cc; P = 0.017). AN patients planned with the CyberKnife system had superior OAR (cochlea and mesial temporal lobe) sparing compared with those planned with the Linac-based system. Further evaluation of these findings in prospective studies with clinical correlation will provide actual clinical benefit from the dosimetric superiority of CyberKnife.

No abstract