Background and Aims

Currently, many research groups are more interested in the FLASH radiotherapy characterized by irradiation with ultra-high dose rate. A first usual step is to validate the beam line for FLASH studies by reproducing published FLASH effect in animals. However, it classically requires time consuming animal studies with dedicated skills, authorizations and infrastructures. Thus, to provide alternative methods and facilitate the implementation and validation of new FLASH beams, we aimed at developing in vitro and ex vivo models that will allow rapid and pertinent evaluation of the FLASH effect.

Methods

For our studies, we are using the ElectronFLASH LINAC manufactured by SIT company. To achieve this goal, we first used an in vitro model of human lung basal stem cells obtained from patients. Cultured in specific air-liquid conditions, this model allows the monitoring of stem cells survival and their capacity to differentiate after irradiation. In parallel, we adapted organotypic lung slices model, recapitulating lung complexity, architecture and microenvironment interactions, for radiation toxicity studies.

Results

Our results indicate that organotypic lung slices enables a rapid evaluation of the FLASH effect.

Conclusions

These models developed in the lab allow to rapidly determine the impact of the various beam parameters on FLASH effect with a robust and reproducible assay. With the inclusion of tumoral cells within the organotypic lung slices, we hypothesize that this ex vivo model can assess concomitantly the FLASH sparing effect on healthy tissue as well as the antitumoral efficacy. Moreover, the model can apply for human patient samples as well as rodent tissues.

Background and Aims

During linac irradiation Cherenkov light is induced in tissue as part of the dose deposition process. This instantaneous light intensity can be used to excite molecular probes, and the signal is directly proportional to the instantaneous dose rate. Thus, during FLASH irradiation, it is possible to use this signal to sense the concentration of molecules and probes in tissue, as is investigated here.

Methods

In vivo measurement of light was done by time-gated intensified cameras that are synchronized to the linac pulses of a 10 MeV FLASH linac. Full emission spectrum light from tissue is imaged for a signal that is proportional to the Cherenkov light coming off the tissue surface. Time-delayed luminescence is imaged with different sequences of time are used to measure the emission kinetics of molecular probe Oxyphor 2P over the course of 50 microseconds after the pulse.

Results

The imaging of Cherenkov can be calibrated to dose (+/-0.5Gy accuracy) for a fixed geometry and fixed tissue optical properties. The imaging of scintillator patches is possible for quantitative dose, independent of the tissue optical properties(+/-0.1Gy accuracy). The imaging of oxygenation is possible for sensing local pO2 from up to 5 pulses of radiation to obtain a reliable lifetime (+/-1mmHg accuracy), with the camera settings used currently.

Conclusions

Pulse to pulse sensing of dose and tissue metabolism is critical to understanding the dose delivered and the tissue responses to FLASH. The tools of Cherenkov luminescence imaging can be deployed with a single time-gated camera to sense dose delivered or tissue oxygenation.

Background and Aims

In this study, we investigated the effects of tumor oxygen tension on the anti-tumor efficacy of ultra-high-dose-rate (FLASH) radiotherapy (RT).

Methods

U87 glioblastoma cells were xenografted in Swiss Nude mice and irradiated using a single 20-Gy dose administered at UHDR (2 pulses, 100 Hz, 1.8 µs pulse width, 0.01 s delivery) or CONV (~ 0.1 Gy/s) dose rates with the Oriatron/eRT6 (PMB, CHUV) under normoxia, hypoxia (vascular clamp), and hyperoxia (carbogen breathing). In situ oxygen tension was measured during and following irradiation using an OxyLite probe. Tumor growth was monitored using caliper measurements and tumor were sampled for RNA and protein profiling (GIF, UNIL). Metabolic analysis and ROS measurements were performed in vitro using Seahorse XF96 Analyzer and CellROX.

Results

Surprisingly, the anti-tumor efficacy of FLASH-RT was not affected by hypoxia in this U87 xenograft model, whereas hypoxia induced radioresistance with CONV-RT. Genomic profiling revealed a decrease in hypoxia signaling in the FLASH-treated compared to the CONV-treated and control tumors 24h post-RT. Oxidative metabolism was also altered in response to FLASH-RT. Real-time tumor oxygen readout, ROS levels, and metabolic testing at different oxygen tensions and timepoints post-RT are in progress.

Conclusions

FLASH-RT anti-tumor efficacy does not seem to be affected by hypoxia supporting a differential role for oxygen signaling between FLASH and CONV-RT and opening new venues for clinical application of FLASH-RT in a subset of highly radiation resistant tumors.

Acknowledgement: The study is funded by SNF Synergia grant (FNS CRS II5_186369)

Background and Aims

Little is known about the mechanism of the biologically proton versus photon irradiation in healthy tissue. Differences in DNA double strand breaks could initiate a distinct secondary response of normal tissue cells, such as the initiation of senescence. To investigate these potential differences between proton and photon irradiation, we used our murine salivary gland organoid model that closely resembles the in vivo response after irradiation.

Methods

First, we established the induction of senescence and senescence-related genes upon both proton and photon irradiation using SA-b-Gal, p16 and p21 as markers.

Results

Interestingly, no differences were observed in the initiation of senescence. However, bulk RNA sequencing showed clear distinctive transcriptome profiles at different time points and radiation doses, especially the inflammatory response. Significant differences were observed in genes related to cytoplasmatic dsDNA and dsRNA recognition, immunity activation, Interferon 1 (IFN1) response and tissue development. Next to this, proton irradiation enhanced cGAS-related signalling, the number of micronuclei containing both cGAS and dsDNA and both cytoplasmic dsRNA and genes related to cytosolic dsRNA sensing. All this coincided with a higher expression of Interferon stimulated genes.

Conclusions

Our study suggests that the DNA damage induced by proton irradiation might lead to more dsDNA leakage into the cytoplasm and consequently more cGAS activation. This might lead to a more pronounced activation of Interferon stimulated genes and dsRNA release that act as positive feedback for the activation of immune responses and inflammation. This enhanced cellular immune response could lead to dissimilar normal tissue response to proton versus photon irradiation.

Background and Aims

Radiation induced skin and soft tissue toxicity remains a complication even for targeted proton pencil beam scanning (PBS) therapy. In this study, we determined the feasibility and benefit of Flash PBS therapy on these toxicities in mice.

Methods

A uniform dose of 35 Gy (toxicity study) or 15 Gy (tumor control study) was delivered to the right hind leg of mice at 1 Gy/s (Conv), 57 Gy/s (FLASH60) and 115 Gy/s (FLASH115) using the plateau region of a 250MeV proton beam. Acute radiation effects were quantified by measurements of TGF-β1 in the plasma and skin and by skin toxicity scoring. Delayed irradiation response was defined by hind leg contracture and plasma levels of 13 cytokines (CXCL1, CXCL10, Eotaxin, IL1-beta, IL-6, MCP-1, Mip1alpha, TNF-alpha, TNF-beta, VEGF, G-CSF, GM-CSF and TGF-β1). Tumor control was quantified in vivo using MOC1 and MOC2 murine oral squamous cell carcinoma (OSCC) cells transplanted into the flank of immunocompetent mice.

Results

Plasma and skin levels of TGF-β1, skin toxicity and leg contracture were significantly decreased in FLASH compared to Conv groups. Maximal FLASH effect was already observed at 60 Gy/s. Plasma levels of CXCL1, GM-CSF, G-CSF and IL-6 were significantly different between FLASH and Con PBS treated animals. FLASH and Conv PBS had similar efficacy on MOC1 and MOC2 tumor growth in vivo.

Conclusions

FLASH PBS radiation can be delivered to mice at dose rates up to 115 Gy/s in a clinical gantry and can improve radiation induced skin and soft tissue toxicity while remaining isoefficient in delaying OSCC growth.

Background and Aims

Radiation-induced gastrointestinal toxicity is a dose-limiting factor in radiotherapy. Recent preclinical studies have shown in multiple models that irradiation at ultra-high dose rates (FLASH) reduces normal tissue toxicity compared to irradiation at conventional dose rates. Here we used a 6 MeV electron LINAC to study acute normal tissue toxicity in C3H mice, where the whole abdomen was irradiated to various radiation doses with FLASH irradiation (mean dose rate = 2-6 MGy/s) or conventional dose rate irradiation (mean = 15 Gy/min). Subsequently, the importance of the temporal pulse structure for the FLASH sparing effect was investigated by varying the number of pulses and the different time interval between two pulses.

Methods

Mice were weighed daily and culled 3.75 days after irradiation and the gastrointestinal systems collected for histological assessment. Stools were collected one day before mice culling to assess the gastrointestinal functionality. Normal tissue damage was quantified using a modified Swiss roll-based crypt assay. Whole blood and plasma were also collected for immunological assessment.

Results

Compared to conventional irradiation, FLASH irradiation showed a dose modifying factor of ≈1.1(a 10% higher dose was required for FLASH irradiation to cause the same level of acute toxicity as conventional dose rate irradiation). Mice irradiated with FLASH also showed reduced weight loss compared to mice that received conventional irradiation. Overall, the normal tissue sparing effect seen in our study seemed to be primarily governed by the mean dose rate.

Conclusions

This study further demonstrates the clinical potential of FLASH radiotherapy, with its capacity of sparing gastrointestinal normal tissue.