Proceedings to the 7th Annual Conference of the Particle Therapy Cooperative Group North America (PTCOG-NA)

Abstract

Purpose : Cancer cells produce innate immune signals following detection of radiation-induced cytosolic DNA via signaling pathways such as cGAS-STING. High linear energy transfer (LET) radiations induce more DNA double-strand breaks (DSBs) per unit dose than low-LET radiations, potentially enhancing immunogenic effects. This work explores the in vitro dose response characteristics of pro-immunogenic interferon-beta (IFN b ) and cGAS-STING antagonist three-prime repair exonuclease 1 (TREX1) from varying-LET radiations. Methods : IFN b and TREX1 expression were measured in MCC13 cells irradiated with graded doses of x-rays or fast neutrons (comparable LET to carbon-12) via ELISA, immunofluorescence, and qPCR assays. Laboratory measurement of the RBE for IFN b production (RBE IFN b ) and TREX1 upregulation (RBE TREX1 ) was compared to the modeled RBE for DSB induction (RBE DSB ) from Monte Carlo DNA damage simulations. RBE IFN b models were applied to radiation transport simulations to quantify the potential secretion of IFN b from representative proton, helium-4, and carbon-12 beams. Results : Maximum IFN b secretions occurred at 5.7 Gy and 14.0 Gy for neutrons and x-rays, respectively (RBE IFN b of 2.5). TREX1 signal increased linearly, with a four-fold higher upregulation per unit dose for fast neutrons (RBE TREX1 of 4.0). Monte Carlo modeling suggests an enhanced Bragg peak-to-entrance ratio for IFNb production in charged particle beams. Conclusion : High-LET radiation initiates larger IFNb and TREX1 responses per unit dose than low-LET radiations. RBE IFN b is comparable to published values for RBE DSB , whereas RBE TREX1 is roughly twofold higher. Therapeutic advantages of high-LET versus low-LET radiation remain unclear. Potential TREX1-targeted interventions may enable IFNb-mediated immunogenic responses at lower doses of high-LET radiations. Aim : To implement lattice radiotherapy using proton pencil beam scanning, and demonstrate treatments that are spatially fractionated in physical dose (PD), with significant escalation of biologic dose (BD) and dose-averaged linear energy transfer (LET d ) in the vicinity of the high PD regions. Method : For 5 patients with bulky tumors, spatial proton dose fractionation inside the GTV was achieved using proton lattice radiotherapy (pLRT). This involves a 3D lattice of 1.5-cm diameter spherical dose regions separated by 3 cm on average. pLRT plans were created with Eclipse (Varian Medical Systems). Two fields with an opening angle of at least 40 degrees were used to reduce skin dose at entrance. Dose valleys between spheres were kept below 40% of the peak PD. The resulting LET d distributions were calculated with an in-house GPU-based Monte Carlo simulation. BD was estimated from LET d and PD by using published formulae that are based on the linear-quadratic model, as well as a simpler model that assumes a linear relationship between BD and the product of LET d (in keV/ l m) and PD: BD ¼ 1.1PD(0.08LET d þ 0.88). Results : Within the high dose spheres, peak BD values in excess of 140% of the prescription dose were observed (see figures). LET d values in the spheres reached values greater than 4 keV/ l m. This was achieved without using any explicit LET d optimization technique, and is a direct consequence of end-of-range energy deposition within the spheres. Conclusion : Besides spatial fractionation, a feature of pLRT is BD escalation. This can be advantageous for debulking radioresistant or hypoxic tumors. Background : This study investigates the radiosensitizing effect of Ganetespib for proton irradiation at a proximal and distal position in a SOBP in comparison to photon irradiation. Rad51, a key protein of homologous recombination repair (HRR), is downregulated by HSP90-inhibiting Ganetespib which provides a promising rational for a specifically proton-sensitizing approach. Methods and Materials : A549 and FaDu cells were treated with low-dose Ganetespib and irradiated with 200kV photons respectively protons at a proximal, low linear energy transfer (LET, 2.1keV/ l m) and a distal, higher LET (4.5keV/ l m) position within a SOBP. Cellular survival was determined by clonogenic assay, cell cycle distribution by flow cytometry, Rad51 protein levels by western blotting and c H2AX foci by immunofluorescence microscopy. Results : Ganetespib reduced clonogenicity in both cancer cell lines exclusively in response to proton irradiation of both investigated LETs. Upon proton irradiation, a more pronounced accumulation of cells in S/G2/M phase became evident with Ganetespib reducing this population. Rad51 protein levels were more extensively and more persistently elevated in proton-than in photon-irradiated cells and suppressed by Ganetespib at each investigated time point. Immunofluorescence staining demonstrated a similar induction and removal of c H2AX foci independent of Ganetespib which suggests compensation by more error-prone Rad51-independent repair pathways. Conclusion : Low-dosed Ganetespib significantly cancer Hence, this study supports pursuing research on the combination of Ganetespib with proton radiotherapy for a prospective clinical exploitation. Purpose: The normal tissue sparing effects of ultra-high dose rate radiation (FLASH) remain poorly understood. We present preliminary results of mouse FLASH proton radiation from a low-energy proton system (50 MeV) optimized for small animal radiobiological research. Methods: We radiated 6-7 week old female C57BL/6 mice with whole lung radiation using the plateau region of a cyclotron-generated 50 MeV preclinical proton beam, transmitting through the whole mouse lung, with beam-shaping via customized vertical and horizontal collimators. Mice were stratified into 3 groups: 1) control/sham radiation; 2) conventional dose rate (17Gy at ~ 0.5Gy/sec); and 3) FLASH (16-18Gy at 42-70Gy/sec). Mice were observed for dermatitis. Lung tissue was harvested post-radiation (1-hour, 5-days, 1-month, 3-months, 6-months). H&E and immunohistochemistry was performed for: yH2aX, cleaved caspase-3, and trichrome. Results: Radiation dermatitis was different between FLASH and conventional groups: FLASH (grade 0-1 ¼ ~ 90%, grade 2 ¼ ~ 10%); conventional (grade 0-1 ¼ ~ 40%; grade 2-3 ¼ ~ 60%) [Figure 1]. One-hour post radiation, lower cleaved caspase-3 IHC staining was seen in the FLASH group versus conventional group, while yH2aX staining was similar in both groups [Figure 2]. More lung airspace disease (fluid and inflammatory cells) was seen in the conventional group at 6-months. Conclusion: Preliminary results of mouse FLASH proton radiation from a 50 MeV beam suggest FLASH proton radiation leads to less normal tissue toxicity than conventional dose rate radiation. More studies are ongoing. Experimental setup: The HollandPTC R&D room is equipped with a fixed horizontal beam line providing beam from 70 up to 240 MeV, and intensities from 1 to 800 nA. The room can provide single pencil beam and large fields with 98% beam uniformity and Spread-Out-Bragg Peak (SOBP) produced with 2D passive modulators. Recently, the maximum energy of 250MeV has been released in the R&D room for FLASH applications. The full beam characterisation has been performed together with absolute dose measurements. Results: A 43% transmission efficiency of the ProBeam cyclotron is achieved at a 250 MeV energy. This resulted in a current of around 300 nA at target position. The beam spot size has a standard deviation of 3.6 mm. The fluence rate was found to be 8e6 protons/cm 2 s, more than a factor of 100 with respect to conventional beams. To further characterise the 250 MeV proton beam at maximum beam current a specific integral monitor chamber is currently under commissioning in collaboration with the company DE.TEC.TOR. Different cutting-edge solutions are adopted for the ionisation chambers to cope with FLASH intensities and minimise the recombination effects.The device is also equipped with X-Y strip ionisation chambers to measure beam size and position. compare out-of-field dosimetry in proton, neutron, and photon radiotherapy with a 3D printed anthropomorphic phantom created using a non-ionizing surface scan. Methods: We used a 3D printed phantom and tissue-equivalent chamber to measure absorbed dose in a phantom constructed from surface imaging of a female volunteer. Absorbed dose was measured in locations approximating the isocenter, thyroid, pacemaker, esophagus, and fetus positions. Square intracranial fields ranging from 2.8cm 2 to 12.8cm 2 were delivered using 6 MV flattened and flattening-filter-free (FFF) photon therapy, magnetically scanned layered proton therapy, and 50.5 MeV proton generated fast neutron therapy. out-of-field dose. For field was small but measurable with of esophagus and fetus proton therapy which measured dose not distinguishable from to proton out-of-field dose 6 MV FFF photon 60% 30% pacemaker. Out-of-field dose FFF out-of-field dose than conventional fields. Out-of-field dose in all locations. Our that out-of-field absorbed dose is reduced in magnetically scanned proton therapy more than photon and is in neutron radiotherapy. In each modality distance from the field edge the magnitude the out-of-field dose. Purpose : The purpose of this study was to investigate the impact of range uncertainty in conjunction with setup errors on dose-averaged linear energy transfer (LET d ) distribution in robustly optimized pencil beam scanning (PBS) proton lung plans. Additionally, the variability of LET d distribution in different breathing phases of 4DCT data set was evaluated. Methods : In this study, we utilized the 4DCT data set of an anonymized lung patient. The tumor motion was approximately 6 mm. A PBS lung plan was generated in RayStation using a robust optimization technique (range uncertainty: 6 3.5% and setup errors: 6 5 mm) on the CTV for a total dose of 7000 cGy(RBE) in 35 fractions. The average RBE was 1.1. The LET d distributions