Program Information

Obtaining Elemental Tissue Composition of Proton Therapy Patients Using Positron Emission Tomography: A Pilot Study

J Cho¹*, G Ibbott¹, M Gillin¹, C Gonzalez-Lepera², O Mawlawi³, (1) Department of Radiation Physics, MD Anderson Cancer Center, HOUSTON, TX, (2) Department of Experimental Diagnostic Imaging, MD Anderson Cancer Center, Houston, TX, (3) Department of Imaging Physics, MD Anderson Cancer Center, Houston, Tx

TU-A-BRA-3 Tuesday 8:00:00 AM - 9:55:00 AM Room: Ballroom A

Purpose: Determination of patient tissue elemental composition is vital to improve proton dose, range calculation and verification; however, currently no method of obtaining elemental composition currently exists. Our aim was to investigate the feasibility of obtaining tissue elemental composition after proton therapy by measuring several positron-emitting radioisotopes.

Methods: Six 2-mm thick samples each containing carbon[C], oxygen[O], carbon/oxygen composite[C+O], carbon/nitrogen composite[C+N], O+N, or C+O+N were irradiated parallel using a monoenergetic proton beam with a median energy of 50-MeV delivering 2.5 Gy. Irradiated samples were moved to a gamma counter and 511-keV annihilation photons were counted for 60 min. Next, three 12-cm long phantoms each composed of C, O, or C+O were irradiated simultaneously first using a monoenergetic proton beam and then using a 6-cm SOBP proton beam of a range of 10-cm with ~2 Gy. After each irradiation, phantoms were PET scanned for 30-min in dynamic mode. Distinct radioisotope fractions from irradiated C-only and O-only phantoms/samples were used to calculate the relative fraction of C and O in composite phantom/samples. In addition, two patients were PET scanned after proton treatments of the head region. Three ROIs - bone&fat, eye, soft-tissue were analyzed. Time-activity signals from all measurements were separated into the constituting radioisotopes' decay curves (O15, N13, C11, etc). The methods above were compared with Monte Carlo and analytical methods using published proton nuclear cross-section data.

Results: The fractions of each element in the C+O and C+O+N samples were estimated with errors of 8 and 15% respectively. The composition of O+C phantom was estimated within 5% error. Elemental compositions of three ROIs were compared with ICRU elemental compositions and showed consistent C/O ratios with uncertainties of 20 to 30%.

Conclusions: Elemental compositions of phantoms, samples and patients were estimated with relatively small errors. PET imaging may potentially be used to improve proton treatment planning and verification.

Contact Email