Background and Aims

There has been increasing evidence of the protective effect of ultra-high dose rate (FLASH) irradiation on different normal tissues recently. Here we intend to verify the FLASH effect of proton irradiation of the intestine, brain and skin.

Methods

For the partial abdominal irradiation, 6-8-week-old C57BL/6j mice and Rag1^-/-/C57 mice were exposed to FLASH (120-130 Gy/s) or conventional dose rate (CDR, ~0.4 Gy/s) proton irradiation of 16 Gy or 16.2 Gy.

For the brain irradiation, 10-12 weeks C57BL/6j mice received a single dose of 10 Gy whole brain irradiation. BrdU was injected to label the proliferating neural stem/progenitor cells in the hippocampus.

To study radiation-induced skin injury, Indian ink was injected intracutaneously into the skin of FVB/N mice the day before irradiation. The distance between the two ink dots was measured with a vernier caliper.

Results

Long-term observation (278 days) of the C57BL/6j mice after 16.2 Gy abdominal irradiation showed no significant survival difference between the FLASH and CDR groups (Figure A). The survival of Rag1^-/-/C57 mice in the 16 Gy FLASH group was lower than that of the 16 Gy CDR group (Figure B). No significant difference was observed in the number of the BrdU labeled cells within the subgranular zone of the hippocampus. Skin contraction after 25 and 27 Gy was significantly greater in mice receiving conventional irradiation compared to FLASH groups (Figure C).

Conclusions

Proton FLASH irradiation protection was observed in skin tissue. However, no significant FLASH sparing effect was observed for intestine and brain.

Background and Aims

A common methodology to report dose rates for pulsed and scanned FLASH beams is missing. We propose the dose-rate dose-fraction (DR-DF) histogram as a common method across all FLASH modalities.

Methods

The DR-DF histogram specifies the fraction of the dose in a point that is delivered with a given minimum mean dose rate (Figure 1). Two parameters can directly be extracted from the DR-DF histogram: (1) the mean dose rate that X% of the dose is at minimum delivered with (DR_X%) and (2) the fraction of the dose that is delivered with a minimum mean dose rate of Y Gy/s (DF_YGy/s,). As an example, DF_40Gy/s is the fraction of the dose that is delivered under FLASH conditions if the FLASH effect is triggered at 40Gy/s mean dose rate. The DR-DF histogram concept was applied to characterize the proton PBS field used in pre-clinical FLASH experiments at our institution.

Results

Figure 2 shows the distribution of dose and dose rate DR_95% for the pre-clinical FLASH beam and the FLASH fraction and FLASH weighted dose in illustrative simulations with the dose rate reduced to 10% of the actual value. In the FLASH weighted dose, the dose delivered with FLASH (DF_40Gy/s) was given a weight of 80% while the remaining dose (100% - DF_40Gy/s) was given a weight of 100%.

Conclusions

The concept of DR-DF histograms was proposed as a common framework to characterize the time structure of scanned and pulsed FLASH beams through mapping of minimum mean dose rate, FLASH fraction and FLASH weighted dose.

Background and Aims

Recent proton FLASH research focuses on transmission planning to exploit the higher beam transport efficiency for high energies, with some compromises to plan quality. This study explores the plan quality and achievable dose rates for transmission and spread-out Bragg peak using a ridge-filter. The results utilise a FLASH effectiveness factor that incorporates dose and dose rate to facilitate an overall plan quality comparison between plans.

Methods

Three spot-reduced, single-field plans were optimised for three patients: IMPT using down-stream range-shifters with consecutive delivery of energy layers; IMPT using down-stream range-shifters with simultaneous delivery of all energies for each lateral spot position, simulating a personalised variable ridge-filter; transmission using the highest available energy of 229MeV. The fraction dose was 1x22.3Gy (equivalent to 30x2Gy and tumour alpha/beta=10Gy). The potential FLASH effects were estimated by multiplying any dose contribution delivered in FLASH state by a FLASH effectiveness factor FEF=0.67. The FLASH state was triggered in a voxel if any dose contribution was delivered above 40Gy/s and 5/10/15Gy and was assumed to persist for 200ms after the trigger has ended.

Results

Figure 1 shows very similar and pronounced FLASH effects for ridge-filter deliveries compared with transmission plans. Figure 2 suggests a potential clinical benefit of the FEF-weighted ridge-filter plans compared with a clinical reference plan’s physical dose, with positive values indicating a reduced integral dose.

Conclusions

With encouraging FLASH characteristics and excellent dose conformality, variable ridge-filters might be a promising approach to bring FLASH proton therapy to clinics.

Background and Aims

The sharp energy deposition of a pulsed ion beam (ideally using short pulses of a few microseconds) results in thermoacoustic waves (ionoacoustics). Signals acquired at multiple positions could allow to infer the in-vivo dose, to localize the Bragg peak or be used for dosimetry. This study assesses dose reconstruction for a small-animal proton irradiator under FLASH conditions (94Gy/s instantaneous dose rate, at the Bragg peak, delivery duration below 200ms). Particular attention is paid to the reconstruction from transient waves emerging from millisecond-long proton pulses at clinical cyclotron facilities.

Methods

The ionoacoustic signals recorded by a realistic 32-element linear array were simulated in water, accounting for the sensor response and acquisition noise (Fig.1). 10ms square proton pulses (30ns rising/falling time and 50% duty cycle) were considered, giving rise to transient ionoacoustic emissions from the pulse edges. The array geometry and piezoelectric material thickness were optimized to improve the accuracy of 2D-dose reconstruction obtained from time-reversal method.

Results

Using the detection of transient ionoacoustic waves from millisecond proton pulses, the dose can be reconstructed (Fig.2.) under FLASH conditions. Optimizing the sensor response and number of pulses to average the signal, the error in the range can be reduced to as low as 0.5% (30µm).

Conclusions

This feasibility study shows that transient ionoacoustics-based dose reconstruction allows for accurate range verification at cyclotron accelerators, assuming dedicated pulsing structure. Thanks to the direct relation between dose and pressure, this method could be applied for in-vivo dosimetry.

Acknowledgements: ERC-725539, H2020-EU INSPIRE-730983, CALA

Background and Aims

The recent rediscovery of the “Flash-effect” revived the interest in high dose-rate radiation effects throughout the radiobiology community, promising protection of normal tissue, while simultaneously not altering tumour control. Systematic preclinical studies resulted in a “recipe” of necessary beam parameters for inducing an electron Flash effect (https://doi.org/10.3389/fonc.2019.01563). For protons, the Flash effect was confirmed in a few animal experiments using the beam parameters available at clinical cyclotrons. Extending the clinical parameter range, the “Dresden platform for high-dose rate radiobiology” enables proton experiments with dose-rates of up to 10⁹ Gy/s.

Methods

The general applicability of the different proton beams for radiobiological studies was proven using biological models of increasing complexity, from cellular models to zebrafish embryo to mouse, at the Draco laser accelerator and, for comparison, at the University Proton Therapy Dresden (UPTD).

Results

A proof-of-principle irradiation campaign was performed using a mouse ear tumour model (https://doi.org/10.1371/journal.pone.0177428) to study the effects of the continuous beam delivery at UPTD and the pulsed beam delivery at Draco with peak dose-rates of 10⁸ Gy/s. Moreover, to investigate the interplay of oxygen consumption and proton dose-rate up to 300 Gy/s and 10⁹ Gy/s, respectively, were applied at UPTD and Draco to study the radiation response of zebrafish embryos.

Conclusions

The successful performance of comparison experiments at Draco laser accelerator and UPTD cyclotron paves the way for upcoming in vivo experiments at both machines. At the conference, we will provide an overview of our radiobiological experiments and the obtained results.

Background and Aims

The ultra-high dose rates thought to be required for the FLASH effect are two orders of magnitude higher than envisaged when current proton therapy shielding was designed. Existing proton therapy equipment offers the exciting prospect of such ultra-high dose rate beams, but consideration of personnel safety is essential. This work aims to assess the suitability of existing shielding for ultra-high dose rate proton beams in the Christie Proton Centre’s research facility.

Methods

A Monte-Carlo model of the shielding, beam delivery equipment and beam stop was created in MCNPX, and calculations carried out to determine neutron and photon flux in the surrounding areas. The model was used to identify areas that would require access restrictions to be in place for an ultra-high dose rate survey. Radiation surveys were carried out using a WENDI-2 detector and Thermo-Scientific FH-40-G Geiger counter to establish neutron and photon dose rates around the room, and to determine the extent of on-going access restriction requirements.

Results

An accelerator current of 800nA at 244MeV was achieved, providing an estimated in-room current of 79nA. Dose-rates at the maze entrance and a plant room above were found to be highest with dose rates exceeding 100µSv/h. The shielding was found to be suitable under these conditions, but only with a significant increase of restricted areas surrounding the research room.

Conclusions

The safe use of ultra-high dose rates for significant beam-on times in our research room using existing shielding is possible, but only by restricting access to some surrounding areas.