Characterization of LiF TLD-100 in carbon ion beams for remote audits

Background

A passive dosimeter framework for the measurement of dose in carbon ion beams has yet to be characterized or implemented for regular use.

Purpose

This work determined the dose calculation correction factors for absorbed dose in thermoluminescent dosimeters (TLDs) in a therapeutic carbon ion beam. TLD could be a useful tool for remote audits, particularly in the context of clinical trials as new protocols are developed for carbon ion radiotherapy.

Methods

TLD-100 were irradiated in a carbon ion beam at the Centro Nazionale di Adroterapia Oncologica (CNAO) in Pavia, Italy. The dose correction factors for linearity, fading, and beam quality were characterized. Fading was characterized from 5 to 100 days post-irradiation. For linearity, the TLDs were irradiated to absorbed doses ranging from 1 to 15 Gy in both the entrance of a high-energy pristine carbon ion peak and the center of a 2 cm spread-out Bragg peak. For beam quality, the TLD was irradiated to the same absorbed dose (3 Gy) in several pristine carbon ion Bragg peaks, as well as in several spread-out Bragg peaks. Each correction factor was calculated and compared to photon correction factors. The correction factors were also compared between high and low dose-averaged linear energy transfer (LET_D) in the carbon ion beams. The absorbed dose was compared between ion chamber and TLD-100 in the several tissue substitute phantom materials, applying the carbon ion TLD correction factors.

Results

There was no statistically significant difference in the TLD fading correction factor between photons, low LET_D carbon ion beams, or high LET_D carbon ion beams. The TLD linearity correction factor did differ between photons, low LET_D carbon ions, and high LET_D carbon ions. The beam quality correction factor was large and changed linearly with LET_D. The overall uncertainty of the carbon ion absorbed dose calculation was 3.9% at the one-sigma level, driven largely by a 3.5% uncertainty in the beam quality correction. TLD measurements were within 1.2% of ion chamber measurements in the phantom material for polyethylene, solid water (Gammex and Sun Nuclear), acrylic, blue water, and techtron HPV. TLD measurements in balsa wood were within 3.0% and cork was 6.6% low compared to ion chamber.

Conclusion

TLD-100 can be used for passive dosimetry in a therapeutic carbon ion beam. Importantly, the linearity and beam quality correction factors are both different from photon therapy, and dependent on LET_D of the carbon ion beam. This opens the possibility of TLD use for carbon ion output audits, phantom audits, and in vivo dose measurements, but may be subject to more complicated management and slightly larger uncertainties than are achieved in photon beams.