40. Evaluation of the Incise2™ multileaf collimator for stereotactic radiotherapy on Cyberknife®

Abstract

Introduction The CyberKnife® (Accuray) is a stereotactic robotic system using historically circular collimators to define the beam apertures. A new multileaf collimator (MLC) is now available on the Cyberknife® M6 version. The purpose of this study was to evaluate the quality of treatment plans using MLC compared to treatment plans using a variable circular aperture collimator IRIS™ for radiosurgery (SRS) and stereotactic body radiotherapy (SBRT). Methods The study was performed on a Cyberknife® M6™ v10.6. A total of 17 cases including cranial, liver, spinal and prostatic cases were investigated. For each case, two treatment plans were generated with Multiplan® v5.3 treatment planning system: one plan using the Incise2™ MLC associated with the Finite Size Pencil Beam (FSPB) dose calculation algorithm and one using the IRIS™ collimator associated with Ray Tracing or Monte Carlo dose calculation algorithm. The dose was prescribed near the 80% isodose level relative to max dose. Plans were normalized to have the same target coverage at ±0.5%. The quality of the MLC and IRIS™ treatment plans was assessed using two criteria: - Dose distribution by reporting the conformal index (COIN), Homogeneity Index (HI) and dose gradient index (GI). - Efficiency by reporting the number of delivered monitor units (MUs) and the treatment time. To evaluate the accuracy of dose calculation algorithm within heterogeneity, a typical lung treatment plan for each collimator was created on a QUASAR™ Respiratory phantom (Modus Medical Inc) with a lung target insert. Calculated planar dose was compared with measurements by radiochromic EBT3 films (Ashland). The analysis was done using gamma analysis 3% (local) / 2 mm (threshold: 30% of the max dose) Results Results are summarized in Table 1. Dose conformality and homogeneity were equivalent for both collimator systems (mean differences less than 3%) except for targets with volumes smaller than 2 cm3 for which dose conformality was better of 14% on average with the IRIS collimator. The dose gradient index was improved by 10.5% on average with the MLC and up to 26% on average for volume larger than 100 cm3. The number of MUs decreased by 25% on average (up to 70%) and the treatment times were systematically reduced with a mean reduction of 28% (max 54%) with the MLC. The comparison between calculated and measured dose in lung phantom showed a gamma passing rate of 51.6 %, 45.5 % and 98.7% for FSPB, Ray Tracing and Monte Carlo respectively. Conclusion The MLC Incise2™ on Cyberknife® allows a significant reduction of MUs and treatment time compared to circular collimators while maintaining a high degree of conformality and a steep dose gradient. For treatment of target volumes smaller than 2 cm3, circular collimators should be still used. The use of MLC for SBRT of lung targets should not be recommended currently due to the absence of type B dose calculation algorithm.