Development of radiochromic film dosimetry at micrometric scale for microbeam radiation therapy

Motivation: Microbeam radiation therapy (MRT) is a novel radiotherapy approach for cancer treatment and neurological diseases based on spatially fractionated fields at micrometric scale. MRT is performed with orthovoltage X-rays and fields are typically generated as array of 50 mm wide planar beams spaced by a 400 mm pitch. MRT provides better healthy tissue sparing and tumor control compared to conventional radiotherapy based on homogeneous field irradiations.

Reliable and robust dosimetry protocols are the basis of any radiation therapy because a precise and accurate dose definition is mandatory for a positive treatment outcome. Dosimetry in MRT is challenging due to the extreme conditions under which it is performed. Differences between simulated and measured dose values in MRT are typically well above the 3% limit used in conventional radiotherapy. Improvements in MRT dosimetry are required for both experimental and computational methods to make this innovative radiotherapy technique available for clinical application.

Methods: Radiochromic films have several useful properties for MRT dosimetry and a read-out protocol for their analysis at micrometric scale was developed using an optical microscope. The definition of some critical factors influencing film dosimetry and, more in general, dosimetry in MRT was possible. The energy dependence of Gafchromic® HD-V2 films was characterized for X-rays in the range 30 – 100 keV and a procedure was defined to overcome the problem of the read-out delay due to the stabilization time of the film after the irradiation.

Monte Carlo (MC) radiation transport simulations, implemented with the Geant4 toolkit, were used to improve MRT dosimetry models. When films are placed between the plastic slabs of a phantom for irradiation with an array of microbeams, the presence of air gaps between film and plastic may influence the dose distribution and this factor was quantified. The interaction of X-rays with a multislit collimator (MSC), i.e. a machined block of metal that spatially fractionates the homogeneous X-ray beam, is one of the most critical elements for the correct description of the dose profile of the microbeam array. Radiation reflection from the inner surface of the MSC was investigated as cause of dose distribution variations in the valley regions (between the microbeams). Furthermore, the radiation leakage through the metal sections of the MSC was studied to understand the differences between calculated and experimental valley dose values.

All these factors were included to benchmark MC simulation versus film dosimetry for the dose evaluation in a water-equivalent cubic phantom and a cell flask.

Results: The characterization of the components of the optical microscope and of the image processing protocol provided a total uncertainty (2σ) for the film read-out of 3.8%. Gafchromic® HD-V2 films show a negligible energy dependence for the photon energies used in MRT. Film read-out can be done after a few hours if the same time interval between irradiation and analysis is respected for all the used films.

The presence of air gaps between the slabs of the utilized phantom degrades the dose distribution when arrays of microbeams are used: 30 μm air gaps increase the valley dose by up to 12%. The air gap was reduced from 14 μm to 6 μm by machining the plastic slabs, thus obtaining a better, less rough, surface. The study of the photon-MSC interaction shows that radiation reflection can bring between 5% and 15% more dose into the valley region, with the effect being largest for small fields. The radiation leakage through the MSC is a further important factor, increasing valley doses from 5% to 30%; this contribution is more relevant for small fields and when using a harder X-ray spectrum in the clinical configuration.

Considering all these factors in MRT dosimetry, agreement between MC simulations and film dosimetry has been significantly improved from the 10-30% range obtained so far. Differences between calculated and measured doses are now less than 5% for dosimetry in a water-equivalent phantom and less than 10% for the cell flask study.

Conclusion: This work has improved MRT dosimetry, establishing a robust and reliable protocol for radiochromic film dosimetry at micrometric scale using an optical microscope. Factors such as the presence of air gaps between film and phantom slabs during irradiation, and the photon-MSC interactions have proven to be critical for a precise and accurate dose evaluation in MRT. Several problems of MRT dosimetry have been solved and agreement between simulated and measured dose is now close to the radiotherapy requirements in medicine, bringing MRT closer to clinical applications.