Radiation damage to normal mammalian tissue in vivo with laser-driven protons at ultra-high instantaneous dose rate
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Physics. The day in general physics: instruments, methods and results across the field.
Kai: We're looking at the paper: "Radiation damage to normal mammalian tissue in vivo with laser-driven protons at ultra-high instantaneous dose rate".
Mira: Here's the gist of the paper. The first in vivo investigation of normal tissue response to laser-driven proton irradiation reveals a reduction in tissue swelling following laser-driven proton treatment compared with X-ray irradiations at conventional dose rates <ref:2602.20460#pg2> RNA sequencing identified differential gene expression associated with immune and epidermal programs following laser-driven proton irradiations at two different dose levels <ref:2602.20460#pg10> Laser-driven (LD) proton sources are an emerging technology enabling multi-disciplinary research with societal impact in areas like radiobiology, material science, and fusion energy These compact sources produce proton bunches of unique parameters, namely a high flux of up to 10 12 particles per bunch, short picosecond scale duration at the source, and low transverse emittance This is of interest for research into the FLASH effect that requires radiation sources capable of ultra-high dose rate and has received increasing attention in the past 10 years for its potential to revolutionize cancer radiation therapy The FLASH effect describes the differential sparing of normal tissue from adverse effects following radiation exposure, while maintaining an equivalent tumor response, when exposing tissue to radiation of ultra-high dose rate (UHDR) of >40 Gy/s in sub-second exposures Experimental Setup and Dosimetry The experiment was performed using the BELLA PW laser facility at LBNL The LD proton source was implemented at the BELLA PW’s short focal length setup, called iP256,57 to enable irradiations of murine ear samples with 8 mm diameter proton beams Online and offline diagnostics were used in different locations along the beamline to characterize the beam parameters and provide online dosimetry for every sample irradiation The average dose per shot over all sample exposures was (2.0 ± 0.4) Gy Tissue Response Observations Ear thickness was measured longitudinally using a constant-pressure micrometer (Mitutoyo) at baseline (pre-radiation) and at 18 subsequent timepoints spanning up to 85 days post-radiation When averaged across animals, maximum ear thicknesses of (286 ± 28) µm for group A, (316 ± 28) µm for group C, and (468 ± 260) µm for group D were observed Treatment with a delivered single-fraction of 36 Gy LD protons showed reduced tissue effects compared to conventional X-rays where the applied dose was adjusted conservatively with proton radiobiological effectiveness values from the literature, suggesting a possible relative sparing Transcriptomic and Late-Phase Analysis RNA sequencing at multiple time points was performed to characterize immune and epidermal transcriptional programs LDA of RNA-seq data from BELLA Group A (36.0 Gy) shows clear separation between 28-day and 84-day irradiated transcriptomes, with both distinct from their respective shams At 84 days post-exposure, genes associated with keratinocyte differentiation remained over-represented among upregulated genes, indicating ongoing epidermal turnover Discussion on Sparing Mechanisms Moderate-dose LD protons maintain the capacity to resolve inflammation and reestablish tissue homeostasis, whereas X-ray exposure is associated with prolonged immune activation and high-dose LD protons induce broad suppression, suggesting potential intrinsic sparing and recovery advantages of LD proton irradiation Emerging evidence suggests that the instantaneous (also called “peak” or “bunch”) dose rate plays an important role in triggering the FLASH effect Future studies will benefit from straightforward automation of online dosimetry coupled with direct shot-go integration into the laser control system, which will allow operation closer to the intrinsic repetition rate of the laser Future Directions Future work at the BELLA proton beamline will extend proton energies to approximately 30 MeV, enabling irradiation of samples several millimeters in depth This will allow direct assessment of normal tissue sparing alongside tumor control within a single experiment With variable pulse separations and proton energies (by varying the laser intensity on target), and energy-variable beam transport systems, this will enable broad systematic studies of the proton FLASH effect across a wide range of MDR and LET regimes within compact LD systems available in university and industry labs Table 1 summarizes the beam parameters used for sample exposures at the LD proton beamline of the BELLA PW For an 8 MeV proton, the bunch length is 11 ns, and it delivers an instantaneous dose rate of 1.3 × 10 8 Gy/s The mean dose rate is approximately 0.1 Gy/s The beam parameters used for sample exposures at the LD proton beamline of the BELLA PW are summarized in Table 1 Key Findings on Dose Effects The total dose error ΔD was between 4.3 Gy (group A) and 5.8 Gy (group C) This variability did not correlate with variations in the applied dose ranging from a minimum of 2% in group D to a maximum of 9% in group C Assuming an RBE factor of 1.1, we observed significantly reduced ear swelling following LD proton exposures compared to conventional X-rays Consistent with this observation, Limoli et al.
Kai: reported that FLASH sparing was maintained in a hypofractionated regime with 6 MeV electrons Data Availability and Funding All data are available in the main text, the supplementary materials, and in the repository doi.org/10.5281/zenodo.18263365 The work was supported by the U.S.
Mira: Department of Energy’s (DOE) Office of Science (SC) Fusion Energy Sciences (FES) program: the LaserNetUS initiative at the BELLA Center, and the Office of High Energy Physics (HEP), both under Contract No.
Kai: DE-AC02-05CH11231 The Lawrence Berkeley National Laboratory Laboratory Directed Research and Development (LDRD) funding provided by the Director Acknowledgments and Authorship The authors gratefully acknowledge the contributions from Samuel Barber, Aodhán McIlvenny, Csaba Tóth, Art Magana, Joe Riley, Zac Eisentraut, Mark Kirkpatrick, Tyler Sipla, Nathan Ybarrolaza, Derrick McGrew, and Teo Maldonado Mancuso A.
Title and authors: Kai: Now, let's talk about what this paper actually found. It's titled "Radiation damage to normal mammalian tissue in vivo with laser-driven protons at ultra-high instantaneous dose rate" >
Mira: The main finding is that they saw a reduction in tissue swelling when they used these laser-driven proton treatments compared to using conventional X-rays at standard dose rates >
Kai: That's the key observation. They also did RNA sequencing and found differences in gene expression related to immune and epidermal programs after the laser-driven proton irradiations at two different dose levels >
Lev: From an engineering standpoint, that differential gene expression analysis is important because it tells us *how* the tissue is reacting biologically, not just that it looks different on a microscope >
Kai: Right. And they showed this reduction in swelling was consistent with what Limoli et al. reported about FLASH sparing being maintained in a hypofractionated regime with six me electrons >
Mira: So they’re linking the physical swelling reduction to the biological changes seen through gene expression, which is a solid connection >
Lev: If we were building hardware for this, that RNA sequencing data would be what we need to feed into a model to predict what kind of tissue response we can actually handle on real hardware >
The paper's summary: Kai: They summarized it by saying that laser-driven proton accelerators are capable of delivering uniquely high instantaneous dose rates in ultrashort bunches, which is what makes them interesting for the FLASH effect research >
Mira: They set up experiments using eight me protons delivering total doses up to fifty gy at two gy per laser shot and they found this reduction in tissue effects compared to X-rays where the dose was adjusted conservatively based on proton radiobiological effectiveness values > <ref:2602.20460#pg2,protons delivering total doses up to 50 gy at>
Lev: The paper points out that the mechanism underlying this sparing effect is still poorly understood, which means we don't have a clear physical theory yet for why it works so well when the dose rate is ultra-high >
Kai: So they are basically saying these sources can do the FLASH effect, and they measured some tissue response changes using transcriptomics to see what was going on in the immune and epidermal systems >
Mira: The RNA sequencing revealed that at eighty four days post-exposure, genes associated with keratinocyte differentiation were still overrepresented among the upregulated genes, which suggests ongoing epidermal turnover even long after the treatment >
Lev: That’s a concrete biological detail. If we were running this on real hardware, we’d need to know if those transcriptomic markers are stable or if they change based on how many shots you deliver >
The paper's improvements: Kai: The paper suggests a few things for improvement. One big thing is that they should focus on integrating online dosimetry measurements, like the integrated current transformer, with post-hoc calibration using radiochromic film data to account for shot to shot variations >
Mira: That addresses the dose error issue, which was a point in their analysis; they found a total dose error between four point three gy and five point eight gy that didn't match the applied dose variation >
Lev: From an engineering view, integrating online dosimetry with film calibration is how you get the most precise picture of what the tissue actually received over time, which is critical for validating any FLASH claim >
Kai: And they also suggest using AI systems to model dose-dependent biological responses by incorporating that RNA sequencing and differential gene expression analysis to predict transcriptional states at different time points >
Mira: That means we can use those transcriptomic patterns to help us figure out what kind of tissue damage we are actually seeing, separating the low-dose effects from the very high-dose suppression >
Conclusion: Kai: So to wrap up, this paper on "Radiation damage to normal mammalian tissue in vivo with laser-driven protons at ultra-high instantaneous dose rate" shows that these laser-driven proton treatments cause a reduction in tissue swelling compared to conventional X-rays >
Mira: They also showed that this sparing effect is linked to changes in the immune and epidermal programs, specifically observing ongoing keratinocyte differentiation at eighty four days post-irradiation >
Lev: For real hardware, the main implication is that we need better ways to precisely measure the instantaneous dose rate because that's what triggers this sparing effect in the first place >
Kai: So they are pushing for better online dosimetry coupled with AI modeling of gene expression to fully understand and control this phenomenon >
Mira: It’s a step toward understanding why FLASH sparing happens, which is something we still don't have a clear mechanism for >
Lev: I just think the next step is figuring out how to scale this on real systems so we can move beyond the animal experiments and actually use these proton sources for therapy or material science applications >
Lawrence Berkeley National Laboratory · Lawrence Livermore National Laboratory
physics.med-ph, physics.plasm-ph
Submitted: 2026-02-24
Updated: 2026-11-09
License: http://creativecommons.org/licenses/by/4.0/
Importance score: 91/100
The gist: The first in vivo investigation of normal tissue response to laser-driven proton irradiation reveals a reduction in tissue swelling following laser-driven proton treatment compared with X-ray…
Key concepts
- Laser-Driven (LD) Proton Sources
- These are compact sources that create proton bunches with high flux and very short durations (picoseconds). They are important for studying the FLASH effect because they can deliver ultra-high dose rates necessary to observe unique radiation effects on tissue.
- FLASH Effect
- This describes the phenomenon where normal tissue is spared adverse effects from radiation when exposed to an ultra-high dose rate (above 40 Gy/s). The paper suggests that the instantaneous (peak) dose rate might be crucial for triggering this sparing effect.
- Transcriptomic Analysis
- RNA sequencing was used to study changes in gene activity after proton irradiation. This analysis helped researchers see how the immune system and skin cells (epidermis) responded differently at various time points following laser treatment, revealing distinct biological programs.
- Tissue Swelling Measurement
- Ear thickness was measured over time using a micrometer to track physical changes in the tissue after radiation exposure. The results showed that moderate LD proton doses caused less swelling than conventional X-rays, implying a degree of protection.
Terminology
Summary
The first in vivo investigation of normal tissue response to laser-driven proton irradiation reveals a reduction in tissue swelling following laser-driven proton treatment compared with X-ray irradiations at conventional dose rates <ref:2602. p. 1 Here, we report the first in vivo investigation of normal
tissue response to laser-driven proton irradiation, with exposures to 8 MeV protons,
delivering total doses up to 50 Gy at 2 Gy per laser shot. Our findings reveal a reduction in
tissue swelling following laser-driven proton treatment compared with X-ray irradiations
at conventional dose rates. Our findings reveal a reduction in
tissue swelling following laser-driven proton treatment compared with X-ray irradiations
at conventional dose rates. RNA sequencing identified differential gene expression
associated with immune and epidermal programs following laser-driven proton
irradiations at two different dose levels.Where the paper says it
Where the paper says it
Introduction to the Technology
Laser-driven (LD) proton sources are an emerging technology enabling multi-disciplinary research with societal impact in areas like radiobiology, material science, and fusion energy These compact sources produce proton bunches of unique parameters, namely a high flux of up to 10 12 particles per bunch, short picosecond scale duration at the source, and low transverse emittance This is of interest for research into the FLASH effect that requires radiation sources capable of ultra-high dose rate and has received increasing attention in the past 10 years for its potential to revolutionize cancer radiation therapy The FLASH effect describes the differential sparing of normal tissue from adverse effects following radiation exposure, while maintaining an equivalent tumor response, when exposing tissue to radiation of ultra-high dose rate (UHDR) of >40 Gy/s in sub-second exposures
Experimental Setup and Dosimetry
The experiment was performed using the BELLA PW laser facility at LBNL The LD proton source was implemented at the BELLA PW’s short focal length setup, called iP256,57 to enable irradiations of murine ear samples with 8 mm diameter proton beams Online and offline diagnostics were used in different locations along the beamline to characterize the beam parameters and provide online dosimetry for every sample irradiation The average dose per shot over all sample exposures was (2.0 ± 0.4) Gy
Tissue Response Observations
Ear thickness was measured longitudinally using a constant-pressure micrometer (Mitutoyo) at baseline (pre-radiation) and at 18 subsequent timepoints spanning up to 85 days post-radiation When averaged across animals, maximum ear thicknesses of (286 ± 28) µm for group A, (316 ± 28) µm for group C, and (468 ± 260) µm for group D were observed Treatment with a delivered single-fraction of 36 Gy LD protons showed reduced tissue effects compared to conventional X-rays where the applied dose was adjusted conservatively with proton radiobiological effectiveness values from the literature, suggesting a possible relative sparing
Transcriptomic and Late-Phase Analysis
RNA sequencing at multiple time points was performed to characterize immune and epidermal transcriptional programs LDA of RNA-seq data from BELLA Group A (36.0 Gy) shows clear separation between 28-day and 84-day irradiated transcriptomes, with both distinct from their respective shams At 84 days post-exposure, genes associated with keratinocyte differentiation remained over-represented among upregulated genes, indicating ongoing epidermal turnover
Discussion on Sparing Mechanisms
Moderate-dose LD protons maintain the capacity to resolve inflammation and reestablish tissue homeostasis, whereas X-ray exposure is associated with prolonged immune activation and high-dose LD protons induce broad suppression, suggesting potential intrinsic sparing and recovery advantages of LD proton irradiation Emerging evidence suggests that the instantaneous (also called “peak” or “bunch”) dose rate plays an important role in triggering the FLASH effect Future studies will benefit from straightforward automation of online dosimetry coupled with direct shot-go integration into the laser control system, which will allow operation closer to the intrinsic repetition rate of the laser
Future Directions
Future work at the BELLA proton beamline will extend proton energies to approximately 30 MeV, enabling irradiation of samples several millimeters in depth This will allow direct assessment of normal tissue sparing alongside tumor control within a single experiment With variable pulse separations and proton energies (by varying the laser intensity on target), and energy-variable beam transport systems, this will enable broad systematic studies of the proton FLASH effect across a wide range of MDR and LET regimes within compact LD systems available in university and industry labs
Summary of Beam Parameters
Table 1 summarizes the beam parameters used for sample exposures at the LD proton beamline of the BELLA PW For an 8 MeV proton, the bunch length is 11 ns, and it delivers an instantaneous dose rate of 1.3 × 10 8 Gy/s The mean dose rate is approximately 0. p. 2 In contrast, LD proton exposure setups have been
established that deliver UHIDR to larger irradiation fields of up to 10 mm diameter4,40–43
. Currently, the LD proton mean dose rate is limited by the drive laser’s repetition rate, which is 1-
10 Hz with today’s multi-Joule laser systems.Where the paper says it
Key Findings on Dose Effects
The total dose error ΔD was between 4.3 Gy (group A) and 5. p. 5 Samples were exposed in four dose and fractionation groups. The table in Figure 2c displays the
average total dose (D), sample to sample variations (SSV), and dose error (ΔD) that samples
received in each dose group.Where the paper says it
Conclusion on Sparing
Assuming an RBE factor of 1.1, we observed significantly reduced ear swelling following LD proton exposures compared to conventional X-rays Consistent with this observation, Limoli et al. p. 9 Assuming an RBE factor of 1. 1, we observed significantly reduced ear swelling following LD
proton exposures compared to conventional X-rays.Where the paper says it
Data Availability and Funding
All data are available in the main text, the supplementary materials, and in the repository doi.org/10.5281/zenodo. p. 0 Data and materials availability: All data are available in the main text, the supplementary
materials, and in the repository doi. 5281/zenodo. The dosimetry
methods are described in the following section. Dosimetry for normal mouse ear irradiations
Sample irradiations were supported by online dosimetry to guide application of the target dose to
each sample, with the final delivered dose to each sample subsequently determined with high org), Offices of Fusion Energy Sciences, and High Energy
Physics, of the Office of Science of the U. Department of Energy Contract No.Where the paper says it
Where the paper says it
Where the paper says it
Acknowledgments and Authorship
The authors gratefully acknowledge the contributions from Samuel Barber, Aodhán McIlvenny, Csaba Tóth, Art Magana, Joe Riley, Zac Eisentraut, Mark Kirkpatrick, Tyler Sipla, Nathan Ybarrolaza, Derrick McGrew, and Teo Maldonado Mancuso A. M. S., L. O., J. I. and K. N. were the lead scientists of the project E E. p. 4 It also allows for a flexible extension with additional magnetic elements to transport up to 30
MeV protons64. Lead shielding was implemented surrounding the proton beam axis that blocked
X- and gamma-rays along the direct line of sight between the TNSA target and the sample location. DEAC02-05CH11231. Lawrence Berkeley National Laboratory Laboratory Directed Research and Development
(LDRD) funding provided by the Director.Where the paper says it
Where the paper says it
References
The references listed are 1 through 85 The reference list includes works such as H. Daido et al., A Macchi et al., and M-C Vozenin et al. The references also include papers on FLASH radiobiology, laser-driven ion sources, and proton therapy applications
Specific Dosimetry Details
The dose applied on every shot was measured with the ICT placed in front of the sample The final dose reported on the samples is a combination of both ICT and film measurements The dose error ΔD was calculated using Gaussian error propagation and included several contributions, such as sample to sample variation (2%-9%, standard deviation) The uncertainty introduced by shot to shot variations during ICT calibrations was 1.
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Precise dose prediction for normal tissue exposure can be achieved by integrating online dosimetry measurements like those from an Integrating Current Transformer (ICT) with post-hoc calibration using Radiochromic Film (RCF) data to account for shot to shot variations, as described in the
Proton dosimetry
section. -
AI systems can be improved to model dose-dependent biological responses by incorporating RNA sequencing and differential gene expression analysis, allowing them to predict transcriptional states at different time points following laser-driven proton irradiation, such as distinguishing between 28-day and 84-day responses.
-
Machine learning models can be trained to correlate radiation parameters, specifically instantaneous dose rate (IDR) versus mean dose rate (MDR), with the occurrence of the FLASH effect by analyzing data from LD proton sources operating outside nominal limits, as noted in the discussion on
mean dose rate and dose per pulse.
-
The system can incorporate RBE factors derived from comparison studies, such as those ranging between 1.2 and 1.430 for LD protons versus conventional X-rays, to provide more accurate biological effectiveness estimations for proton doses in ultra-high dose rate regimes.
-
AI can be developed to analyze complex transcriptomic data to identify hallmark pathway enrichment modules that distinguish between the effects of low-dose BELLA irradiation (retaining epithelial turnover) and very high-dose BELLA irradiation (inducing broad suppression).
Abstract
The differential sparing of normal tissues relative to tumor control observed at ultra-high dose rates, referred to as the FLASH effect, has recently gained considerable attention. The therapeutic advantages of FLASH radiotherapy are expected to be further amplified through the use of protons and ions, which enable precise dose deposition at tumor depth while minimizing irradiation of healthy tissues proximal and distal to the target. Nevertheless, the mechanism underlying this sparing effect remains poorly understood. Laser-driven proton accelerators are capable of delivering uniquely high instantaneous dose rates in ultrashort bunches. Here, we report the first in vivo investigation of normal tissue response to laser-driven proton irradiation, with exposures to 8 MeV protons, delivering total doses up to 51 Gy at 2 Gy per laser shot. Our findings reveal a reduction in tissue swelling following laser-driven proton treatment compared with X-ray irradiations at conventional dose rates. RNA sequencing identified differential gene expression associated with immune and epidermal programs following laser-driven proton irradiations at two different dose levels.
Related papers
- BART Online Open-Source Sequence Toolbox for Computational MRI
- Constrained estimation of rotational invariants of the cumulant expansion (RICE) for rapid tensor-valued diffusion MRI
- Improved Myelin Water Fraction Estimation and Uncertainty Mapping in Multicomponent T2 Relaxometry via NNLS Bootstrap Aggregation
- BEAM3R: Beam's-eye-view architecture with Mamba-3 for implicit dose reconstruction
- MORSE-PI -- Flexible and artefact-free image reconstruction for structural and functional QSM and other phase-critical imaging applications
- RadMaps: A Geospatial Framework for Simultaneously Modelling Capacity and Geographic Constraints on Radiotherapy Access