Assessment of single isocenter linear accelerator radiosurgery for metastases and base of skull lesions.

BACKGROUND: Newer technology for stereotactic radiosurgery (SRS) should be assessed for different multi-leaf collimators (MLC). OBJECTIVE: Assess plan quality of an automated, frameless, linear accelerator based (linac) planning and delivery system (HyperArc) for SRS using both standard MLC (SMLC) and high definition MLC (HDMLC) compared to a cobalt-60 based SRS system (Gamma Knife, GK). METHODS: We re-planned twenty GK Perfexion-treated SRS patients (27 lesions) for HyperArc using SMLC and HDMLC. We assessed plan quality using the following metrics: gradient index (GI), Paddick and RTOG conformity indices (CIPaddick, CIRTOG), volume receiving half of prescription isodose (PIVhalf) and maximum dose to 0.03 cc for brainstem, optic chiasm and optic nerves, and V12Gy for brain-GTV. RESULTS: Linac plans had better conformity with HDMLC being most conformal. GK exhibited better GI. PIVhalf demonstrated no statistically significant difference between HDMLC and GK, and SMLC was nominally worse than GK. Mean PIVhalf was generally 0.85 cc larger for SMLC than HDMLC. For TV > 1.0 cc, the relative differences in CIRTOG, GI, and PIVhalf for SMLC vs. HDMLC were less than 21%. For TV less than < 1.0 cc, there were more obvious relative differences for SMLC vs. HDMLC in CIRTOG (mean 146%, max 700%), GI (mean 49%, max 162%), and PIVhalf (mean 77%, max 522%). Organ at risk doses were met in all plans. CONCLUSIONS: New linac-based plans positively compare to GK plans overall. HDMLC should be strongly considered for treatment of lesions < 1.0 cc given the significant improvements in conformity and PIVhalf over SMLC.

Assessment of intra-fraction motion during automated linac-based SRS treatment delivery with an open face mask system.

PURPOSE/OBJECTIVE: To evaluate intra-fraction target shift during automated mono-isocentric linac-based stereotactic radiosurgery with open-face mask system and optical real-time tracking. MATERIALS/METHODS: Ninety-five patients were treated using automated linac-based stereotactic radiosurgery in 1-5 fractions with single isocenter for a total of 195 fractions. During treatment, patient positioning was tracked real-time with optical surface guidance and immobilized with a rigid open-face mask. Patients were re-positioned if optical surface guidance error exceeded 1 mm magnitude or 1°. Translational and rotational intra-fractional changes were determined by post-treatment CBCT matched to the planning CT. Target specific error was calculated by translation and rotation matrices applied to isocenter and target spatial coordinates. RESULTS: For 132 fractions with isocenter within a single target, the median shift magnitude was 0.40 mm with a maximum shift of 1.17 mm. A total of 398 targets treated for plans having multiple or single targets that lied outside isocenter, resulted in a median shift magnitude of 0.46 mm, with median translational shifts of 0.20 mm and 0.20° rotational shifts. A 1 mm PTV margin was insufficient in 18% of targets at a distance greater than 6 cm away from isocenter, but sufficient for 96% of targets within 6 cm. CONCLUSIONS: The findings of this study support 1 mm PTV expansion due to intra-fraction motion to ensure target coverage for plans with isocenter placement less than 6 cm away from the targets.