101 A PENCIL BEAM ALGORITHM FOR HELIUM ION BEAM DOSE CALCULATION

S39 into account when calculating the exposure times. The control cells were kept at room temperature and were not irradiated. The irradiated and control cells were immediately processed for RNA extraction. Five hundred nanogram of freshly extracted RNA were reverse- transcribed to cDNA for quantitative polymerase chain reaction (qPCR) using primers for the cytochrome C oxidase (COXIV) subunit 2, a mitochondrially encoded gene; COXIV subunit 4 and manganese superoxide dismutase (MnSOD), both nuclear encoded. β-actin as well as glyceraldehyde 3phosphate dehydrogenase were used as internal standards. The relative expression of target genes was quantified using comparative Ct analysis incorporated in the CFX ManagerTM Software (version 1.5) (BioRad, California, USA). Results: In our study, no time was allowed for the cells to recover before the assay. Thus, all the responses shown are acute effects. The results of the qPCR demonstrate that COXIV subunit 2 mRNA was the only gene that showed an approximate linear decrease with the increase in radiation dose (R2 = 0.8251). Gene expression from another subunit of the same enzyme complex i.e. COXIV subunit 4, did not show any change in response to various doses of radiation. Similar results were found for the MnSOD gene. The ratio of the mRNA expression of mitochondrially encoded COXIV subunit 2 and the nuclear encoded subunit 4 (i.e. mitochondrial vs nuclear gene expression of the same enzyme complex) illustrated the linear relationship between the loss in mitochondrially encoded gene expression and the increase in radiation dose (R2 = 0.9834). Conclusions: MnSOD is a mitochondrial enzyme that catalyzes the dismutation of superoxide anion into water and hydrogen peroxide, thus plays a key role in protecting the cell from oxidative damage. Our data indicate that acute effects are not measurable. It is likely that radiation dose-dependent MnSOD gene upregulation can however be detected after a longer cell survival time. COXIV is the terminal enzyme of the mitochondrial respiratory chain; in particular, its subunit 2 plays a major role in converting oxygen to water. Our data indicate a differential susceptibility of the mitochondrially encoded subunit 2 of COXIV to γradiation. The loss of mRNA expression is linear to dose and is not observed in the nuclear encoded subunit 4 of COXIV. Mature mRNAs are essential for protein synthesis while only 3% of the total DNA is known to be directly involved in mRNA transcription. It has been suggested that, therefore, RNA damage may be more detrimental than DNA damage to the overall health of the cell (Curr Biol 2003; 13: R482-R484). In this study, we have first data showing that mitochondrially encoded mRNA appears to be more susceptible to radiation damage. Given the important role of mitochondria and mRNA, even relative minor radiation effects on mitochondrially encoded mRNA could be significant determinants of acute cell responses and survival.

ICTR-PHE 2012

Figure: A to C: qPCR amplification curves for (A) COXIV subunit 2, (B) COXIV subunit 4 and (C) MnSOD. D to F: Normalized mRNA expression of (D) COXIV subunit 2, (F) COXIV subunit 4, and (F) MnSOD, at different doses of γ radiation. G: Ratio of the COXIV subunit 2 vs COXIV subunit 4 mRNA expression at different doses of γ radiation. 101 A PENCIL BEAM ALGORITHM FOR HELIUM ION BEAM DOSE CALCULATION H. Fuchs1, J. Ströbele1, T. Schreiner2, D. Georg1 1 Department of Radiotherapy, Medical University Vienna, Vienna, Austria 2 EBG MedAustron, Wiener Neustadt, Austria Purpose: To develop a flexible pencil beam algorithm for helium ion beam therapy. The algorithm was based on an existing one for scanned proton beam therapy but adapted with respect to nuclear interaction and fragmentation. Dose distributions were calculated using the newly developed pencil beam algorithm and validated using Monte Carlo methods in various phantom geometries. Material and Methods: The algorithm is based on an established theory of fluence weighted elemental pencil beam kernels for proton beams. A new splitting approach using real-time optimization has been realized. Depending on the number of sub-beams, a minimization routine selects the optimal shape for each sub-beam, disposing the need of manual optimizations. Dose depositions along the beam path are determined using a look-up table approach. The respective reference-data for look-up table generation were derived from Monte Carlo simulations in water using GATE v6.1 based on GEANT4.9.4 patch 01, for 21 energies between 50 and 250 MeV per nucleon, with increments of 10 MeV per nucleon. For materials other than water, dose depositions are calculated in the pencil beam model using a water-equivalent depth. Lateral beam spreading caused by multiple scattering has been accounted for by implementing various userselectable scattering formulas such as Lynch formula, Highland-Lynch or the Rossi formula. The pencil beam algorithm has been implemented in C++ using the ROOT framework. Validation simulations have been performed using a 40 x 40 x 100 cm3 phantom filled with either homo-

S40 geneous materials or heterogeneous slabs. The monoenergetic, mono-directional beams hit the phantom in the surface centre, initial particle energies ranged from 50 to 250 MeV per nucleon with a flux of 107 ions per beam. The position and shape of the Bragg-Peak was defined at 80 % of the maximum energy of the distal end and by the width at 60 % of the energy deposition anterior and posterior of the maximum. For comparison, a dedicated gamma index evaluation software was developed calculating the gamma indices for one, two and three dimensional dose distributions. Results: The validity of the Monte Carlo simulations was established by comparison with tabulated penetration depths. Pencil beam results in homogeneous phantoms filled with water, bone, adipose and muscle tissue were compared with reference Monte Carlo simulations. The position of the beam and shape of the Bragg-Peak showed very good agreement, with deviations of less than 0.1 mm for the maximum energy. Although differences increased slightly with increasing initial energies, lateral beam spreading showed also good conformity, with FWHM differences of less than 0.5 mm at the position of the Bragg-Peak for 250 MeV per nucleon. In addition, heterogeneous phantom configurations with slaps of varying material, thickness and position as well as alternating slabs were evaluated. Maximum deviations in the position of the Bragg-Peak of less than 1.2 % were observed. The shape of the Bragg-Peak showed differences of less than 0.2 mm. For lateral beam spreading the differences of the FWHM were similar to that ones of homogeneous phantoms. Although the pencil beam model has not been optimized in terms of speed up to now, the calculation times were comparable to existing literature describing proton pencil beams. Preliminary results of the gamma index evaluation showed good agreement in most of the regions. However, as expected sharp borders of materials with highly different properties, such as water and bone, alongside the beam path, pose larger differences between the pencil beam model and Monte Carlo simulations. Although only data for the helium beams were presented above, the performance of the pencil beam model for proton beams is comparable to that of helium ions. Conclusion: The pencil beam algorithm developed for helium ions presents a suitable tool for dose calculations. The flexible design allows easy customization to measured depth-dose distributions and accommodation of varying beam shapes, therefore making it a promising candidate for integration into treatment planning systems. 102 GEFITINIB ENHANCED RADIOSENSITIVITY OF STEM-LIKE GLIOMA CELLS BY INHIBITION OF EGFR-AKT-DNA-PK SIGNALING, ACCOMPANIED WITH INHIBITION OF DNA DOUBLE-STRAND BREAK REPAIR. K.B. Kang, C.J. Zhu, Y.L. Wong, Q.H. Gao, A. Ty, M.C. Wong Brain Tumor Research Laboratory, Humphrey Oei Institute of Cancer Research, National Cancer Centre Singapore, Singapore. Purpose: Brain tumor stem cells (BTSCs) have been identified as responsible for maintenance and growth

ICTR-PHE 2012 of brain tumors, which subsequently contributes to chemo- and radioresistance. In this study, we compared the radiosensitivity of brain tumor stem cells (BTSCs) with matched non-stem glioma cells. We determined whether the tyrosine kinase inhibitor gefitinib enhanced BTSC radiosensitivity by inhibiting epidermal growth factor receptor (EGFR)-Akt-DNAdependent protein kinase (DNA-PK) signaling, followed by enhancement of DNA double-stand breaks (DSBs)/ inhibition of DSB repair. Material and Methods: Stem-like gliomaspheres (which were cultured from patient neurosurgical resection with informed consent) were evaluated for characteristics of BTSCs to self-renew, differentiate into neuronal lineages and initiate glioma growth in vivo. Radiosensitivity of stem-like gliomaspheres and matched non-stem glioma cells were evaluated by clonogenic assays, γ-H2AX immunostaining (a marker of DNA DSBs) and cell cycle distribution. Survival of irradiated and non-irradiated NOD-SCID mice intracranially-implanted with stem-like gliomaspheres, were monitored. Glioma cells treated with gefitinib, irradiation and both, were assayed for clonogenic immunostaining, DNA-PKcs survival, γ-H2AX expression and phosphorylation of EGFR and Akt. Results: Stem-like gliomaspheres displayed BTSC characteristics of self-renewal; differentiation into lineages of neurons, oligodendrocytes and astrocytes; and initiation of glioma growth in NOD-SCID mice. Irradiation dose-dependently reduced clonogenic survival, induced G2/M arrest and increased γ-H2AX immunostaining of non-stem glioma cells, but not stem-like gliomaspheres. There was no difference in survival of irradiated and non-irradiated mice implanted with stem-like gliomaspheres. Addition of gefitinib significantly inhibited clonogenic survival, increased γ-H2AX immunostaining and reduced DNAPKcs expression of irradiated stem-like gliomaspheres, without affecting irradiated non-stem glioma cells. Gefitinib alone, and when combined with irradiation, inhibited phosphorylation of EGFR (Y1068 and Y1045) and Akt (S473) in stem-like gliomaspheres. In non-stem glioma cells, gefitinib alone inhibited EGFR Y1068 phosphorylation, with further inhibition by combined gefitinib and irradiation. Conclusions: Stem-like gliomaspheres are resistant to irradiation-induced cytotoxicity, G2/M arrest and DNA DSBs, compared with non-stem glioma cells. Gefitinib differentially enhances radiosensitivity of stem-like gliomaspheres by reducing EGFR-Akt activation and DNA-PKcs expression, accompanied with enhanced irradiation-induced DNA DSBs/ inhibition of DSB repair. 103 TECHNIQUES FOR IMAGE BASED IN-VIVO DOSIMETRY: FROM PARTICLE THERAPY PET TO IN-BEAM PROMPT GAMMA IMAGING F. Fiedler1*, U. Dersch2, C. Golnik2, S. Helmbrecht2, T. Kormoll2, D. Kunath1,2, K. Laube2, A. Müller2, M. Priegnitz1, H. Rohling2, S. Schöne1, W. Enghardt1,2 1 Helmholtz-Zentrum Dresden-Rossendorf, P.O. Box 510119, 01314 Dresden, Germany 2 OncoRay, TU Dresden, Fetscherstraße 74, 01307 Dresden, Germany