Stereotactic Arrhythmia Radioablation (STAR) of Ventricular Tachycardia: A Treatment Planning Study

PURPOSE The first stereotactic arrhythmia radioablation (STAR) of ventricular tachycardia (VT) was delivered at Stanford on a robotic radiosurgery system (CyberKnife® G4) in 2012. The results warranted further investigation of this treatment. Here we compare dosimetrically three possible treatment delivery platforms for STAR. METHODS The anatomy and target volume of the first treated patient were used for this study. A dose of 25 Gy in one fraction was prescribed to the planning target volume (PTV). Treatment plans were created on three treatment platforms: CyberKnife® G4 system with Iris collimator (Multiplan, V. 4.6)(Plan #1), CyberKnife® M6 system with InCise 2(TM) multileaf collimator (Multiplan V. 5.3)(Plan #2) and Varian TrueBeam(TM) STx with HD 120(TM) MLC and 10MV flattening filter free (FFF) beam (Eclipse planning system, V.11) (Plan #3 coplanar and #4 noncoplanar VMAT plans). The four plans were compared by prescription isodose line, plan conformity index, dose gradient, as well as dose to the nearby critical structures. To assess the delivery efficiency, planned monitor units (MU) and estimated treatment time were evaluated. RESULTS Plans #1-4 delivered 25 Gy to the PTV to the 75.0%, 83.0%, 84.3%, and 84.9% isodose lines and with conformity indices of 1.19, 1.16, 1.05, and 1.05, respectively. The dose gradients for plans #1-4 were 3.62, 3.42, 3.93, and 3.73 with the CyberKnife® MLC plan (Plan #2) the best, and the TrueBeam(TM) STx co-planar plan (Plan #3) the worst. The dose to nearby critical structures (lung, stomach, bowel, and esophagus) were all well within tolerance. The MUs for plans #1-4 were 27671, 16522, 6275, and 6004 for an estimated total-treatment-time/beam-delivery-time of 99/69, 65/35, 37/7, and 56/6 minutes, respectively, under the assumption of 30 minutes pretreatment setup time. For VMAT gated delivery, a 40% duty cycle, 2400MU/minute dose rate, and an extra 10 minutes per extra arc were assumed. CONCLUSION Clinically acceptable plans were created with all three platforms. Plans with MLC were considerably more efficient in MU. CyberKnife® M6 with InCise 2(TM) collimator provided the most conformal plan (steepest dose drop-off) with significantly reduced MU and treatment time. VMAT plans were most efficient in MU and delivery time. Fluoroscopic image guidance removes the need for additional fiducial marker placement; however, benefits may be moderated by worse dose gradient and more operator-dependent motion management by gated delivery.