Skip to main content
Radiotherapy

Radiotherapy

Could hydrogen peroxide generation underlie mini- and micro-beam radiotherapy efficacy?

28 Oct 2020 Tami Freeman
Joao Seco
Joao Seco and colleagues are investigating how radiation mini- or micro-beams can destroy tumours while sparing nearby normal tissue. (Courtesy: Joao Seco)

Micro- and mini-beam radiation therapies (MBRTs) have been shown in animal experiments to effectively destroy tumours while minimizing damage to normal tissue nearby. MBRT is delivered by irradiating a tumour with a series of high-dose beamlets of protons or photons alternated with low-dose valleys. The reduced side-effects to organs-at-risk are thought to be due to the differential response of normal and tumour tissue to such spatially fractionated radiation. The mechanisms underlying this differential response, however, are still unknown.

Aiming to fill this gap in knowledge, researchers in Germany have proposed and investigated a chemical mechanism to describe the efficacy of mini-beams and micro-beams. They performed a simulation study to find a surrogate to the tumour control seen in MBRT. Building on a proposed correlation between tissue damage and the level of reactive oxygen species (ROS) in tissue, the team examined whether the distribution of a radiation-induced molecule or radical could provide such a surrogate.

“From all published animal studies with micro- and mini-beams, it was clear that dose coverage of the tumour in the valley region was too low to allow any form of tumour control. So the dose delivered to the tumour was not a good surrogate for establishing the biological effect,” explains senior author Joao Seco from DKFZ. “Several years ago, my group started working in radiochemistry, with the focus of evaluating which radical or molecule could be a good surrogate of biological effect.”

In a study described in Frontiers in Physics, Seco and colleagues investigated 12 potential generic radicals and molecules (gRMs) produced by a radiation beam. As physical dose coverage is not achieved in MBRT, they hypothesized that biological damage is instead provided by the distribution of such a gRM reaching a uniform coverage of the tumour target.

As such, they proposed that a candidate gRM must meet four conditions: it should be stable enough to diffuse during beam-on to cover the dose-valley regions; it should reach a steady state in production versus removal, within a few microseconds; it should be a product of water radiolysis; and it should have oxidizing capacity to create cellular damage.

The researchers, also from Heidelberg University and GSI, used the Monte Carlo code TRAX-CHEM to model the production, removal and diffusion of the 12 gRMs. The simulations revealed that the steady state was only reached by three of the gRMs, with only one of these – hydrogen peroxide (H2O2) – an oxidizing species. Thus they restricted all further analysis to H2O2.

Temporal diffusion

To assess the potential of H2O2 as a candidate for the biological efficacy of MBRT, the researchers evaluated their simulations against previous animal experiments. One limitation of TRAX-CHEM is that it only runs up to 10-6 s. So to extend their predictions to times suitable for experimental comparison (up to 103 s), they modelled the time evolution of H2O2 with a convolution model that uses a Gaussian Kernel (calculated from TRAX-CHEM) to convert delivered dose into H2O2 spatial distribution at a specified time.

They calculated the H2O2 spatial distribution for four proton mini-beam and photon micro-beam studies, using the published values of peak spacing, peak FWHM and peak-to-valley dose ratio (PVDR). Based on these parameters, they calculated the beam-on time at which the H2O2 distribution reached a uniform tumour coverage.

Comparisons with the actual irradiation times revealed that the calculated minimum irradiation times were reached in three of the four experiments. In these cases, the predicted H2O2 distribution had a coverage of at least 95%. The team notes that these three experiments were all associated with high probabilities of tumour ablation or growth delay.

Minimum beam-on time

In the experiment that did not reach the minimum beam-on time, the model predicted that H2O2 did not uniformly cover the target. In this experiment (which used synchrotron-generated X-ray microbeams), only two tumours were ablated in 32 irradiated rats. In contrast, an experiment using the same synchrotron beams but with smaller spacing, in which H2O2 diffused uniformly over the target, produced five tumour ablations out of 11 rats.

The researchers surmise therefore that homogeneous H2O2 distribution is a highly relevant parameter for tumour ablation and should be further investigated. The fact that this uniform coverage may be generated in the tumour but not in normal tissue, combined with the higher tolerance of normal cells to ROS relative to cancer cells, may underlie the differential effect between tumour and normal tissue in MBRT.

Seco notes that the developed model can be used to assess whether MBRT will provide uniform H2O2 coverage of the tumour volume. There is still a need, however, to investigate the correlation between this coverage and treatment response. The team is looking to discuss with other groups a possible animal study to understand how H2O2 could be used to quantify treatment response.

“In our study, we demonstrated that we should use H2O2 – not dose – as a marker of biological effect for micro- and mini-beam radiotherapy,” Seco concludes.

Copyright © 2024 by IOP Publishing Ltd and individual contributors