Validation of a Monte Carlo Framework for Out-of-Field Dose Calculations in Proton Therapy

Proton therapy enables to deliver highly conformed dose distributions owing to the
characteristic Bragg peak and the finite range of protons. However, during proton therapy,
secondary neutrons are created, which can travel long distances and deposit dose in outof-field volumes. This out-of-field absorbed dose needs to be considered for radiationinduced secondary cancers, which are particularly relevant in the case of pediatric
treatments. Unfortunately, no method exists in clinics for the computation of the out-offield dose distributions in proton therapy. To help overcome this limitation, a
computational tool has been developed based on the Monte Carlo code TOPAS. The
purpose of this work is to evaluate the accuracy of this tool in comparison to experimental
data obtained from an anthropomorphic phantom irradiation. An anthropomorphic
phantom of a 5-year-old child (ATOM, CIRS) was irradiated for a brain tumor treatment
in an IBA Proteus Plus facility using a pencil beam dedicated nozzle. The treatment
consisted of three pencil beam scanning fields employing a lucite range shifter. Proton
energies ranged from 100 to 165 MeV. A median dose of 50.4 Gy(RBE) with 1.8 Gy(RBE)
per fraction was prescribed to the initial planning target volume (PTV), which was located
in the cerebellum. Thermoluminescent detectors (TLDs), namely, Li-7-enriched LiF : Mg, Ti
(MTS-7) type, were used to detect gamma radiation, which is produced by nuclear
reactions, and secondary as well as recoil protons created out-of-field by secondary
neutrons. Li-6-enriched LiF : Mg,Cu,P (MCP-6) was combined with Li-7-enriched MCP-7
to measure thermal neutrons. TLDs were calibrated in Co-60 and reported on absorbed
dose in water per target dose (mGy/Gy) as well as thermal neutron dose equivalent per target dose (mSv/Gy). Additionally, bubble detectors for personal neutron dosimetry (BDPND) were used for measuring neutrons (>50 keV), which were calibrated in a Cf-252
neutron beam to report on neutron dose equivalent dose data. The Monte Carlo code
TOPAS (version 3.6) was run using a phase-space file containing 1010 histories reaching
an average standard statistical uncertainty of less than 0.2% (coverage factor k = 1) on all
voxels scoring more than 50% of the maximum dose. The primary beam was modeled
following a Fermi–Eyges description of the spot envelope fitted to measurements. For the
Monte Carlo simulation, the chemical composition of the tissues represented in ATOM
was employed. The dose was tallied as dose-to-water, and data were normalized to the
target dose (physical dose) to report on absorbed doses per target dose (mSv/Gy) or
neutron dose equivalent per target dose (mSv/Gy), while also an estimate of the total organ
dose was provided for a target dose of 50.4 Gy(RBE). Out-of-field doses showed
absorbed doses that were 5 to 6 orders of magnitude lower than the target dose. The
discrepancy between TLD data and the corresponding scored values in the Monte Carlo
calculations involving proton and gamma contributions was on average 18%. The
comparison between the neutron equivalent doses between the Monte Carlo simulation
and the measured neutron doses was on average 8%. Organ dose calculations revealed
the highest dose for the thyroid, which was 120 mSv, while other organ doses ranged
from 18 mSv in the lungs to 0.6 mSv in the testes. The proposed computational method
for routine calculation of the out-of-the-field dose in proton therapy produces results that
are compatible with the experimental data and allow to calculate out-of-field organ doses
during proton therapy.