Multileaf collimator end leaf leakage: implications for wide-field IMRT

Phys Med Biol. 2007 Nov 7;52(21):N493-504. doi: 10.1088/0031-9155/52/21/N01. Epub 2007 Oct 11.

Abstract

The multi-leaf collimator (MLC) of a particular linear accelerator vendor (Millennium MLC, Varian Medical Systems, Palo Alto, CA, USA) has a maximum leaf extension of 14.5 cm. To achieve intensity modulated radiotherapy (IMRT) for fields wider than 14.5 cm all closed leaf pairs are restricted to placement inside the field. Due to the rounded leaf end design of the MLC end leaf leakage will occur in the treatment field. The implementation of direct aperture optimization in the IMRT module of a radiotherapy treatment planning system (Pinnacle, Philips Radiation Oncology Systems, Milpitas, CA) has facilitated the delivery of IMRT fields wider than 14.5 cm. The end leaf leakage of the Millennium MLC has been characterized for 6 MV photons using gafchromic and radiographic film, and the accuracy of the planning system verified. The maximum leakage measured for a single field was 0.39 cGy MU(-1) for a 0 mm leaf gap and 0.51 cGy MU(-1) for a 0.6 mm leaf gap. For a clinical IMRT field leaf end leakage contributed an additional 2-3 Gy over the course of treatment. The planning system underestimated the magnitude of end leaf leakage by 20-40%. The ability to deliver IMRT fields wider than 14.5 cm with the Millennium MLC has improved the efficiency and flexibility of IMRT treatments; however, significant extra dose can be introduced due to end leaf leakage. Caution should be exercised when delivering wide field IMRT as it is not a complete panacea. Any significant occurrences of end leaf leakage predicted by the planning system should be independently verified prior to delivery.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Equipment Design
  • Film Dosimetry / methods
  • Humans
  • Particle Accelerators*
  • Photons
  • Radiometry
  • Radiotherapy Planning, Computer-Assisted
  • Radiotherapy, Intensity-Modulated / methods*
  • Reproducibility of Results