How can we better identify and prevent the side effects of radiotherapy, one of the main techniques used to treat cancer? To try and answer this question, the IRSN launched an ambitious research program, ROSIRIS*, in 2009. The program's objective is to acquire new fundamental knowledge of the main physical and biological mechanisms involved, which will ultimately help us better predict these effects and risks and reduce the possible sequelae of radiotherapy. More generally, the program's results will also help us better understand the effects of radiation on living subjects. The aim is to connect, step by step, the initial events of radiation energy transfer within molecules to the biological effects that occur much later.
ROSIRIS is a multi-disciplinary program that brings together a wide range of skill sets (radiobiology, radiopathology, physical dosimetry, bioinformatics, particle physics, modeling and programing, biomathematics, imaging, and more).
Pulmonary lesions in a mouse revealed by high-resolution micro-CT imaging following stereotactic millimetric irradiation using the SARRP. © IRSN
Background and Objectives
Cancer remains the leading cause of death in Europe, with 4.2 million new cases in 2018
(source: IARC) and a substantial economic impact. The cancer plan developed by the French National Cancer Institute (INCa) includes a national research strategy to drive progress in cancer treatment. Seventeen operational objectives were set for the period 2014–2019, most relating to radiotherapy. Ensuring comprehensive, personalized care (objective 7) and reducing the risks of sequelae and second cancer (objective 8) should help "extend the lives and preserve the quality of life" of patients.
Radiotherapy consists of delivering large doses of ionizing radiation to the tumor, while minimizing the dose delivered to healthy tissue found within the radiation field. The reactions that occur within this healthy tissue can lead to complications. While some of these effects disappear spontaneously, other can have a significant and long-lasting effect on patient quality of life. Radiation-induced toxicity within healthy tissue is therefore a limiting factor when determining the dose that can be delivered to the tumor. In addition, the technical side of radiotherapy has become much more complex in recent years, with the appearance of new technologies (stereotaxis, use of ions, high dose rates, etc.) that can deliver sophisticated forms of irradiation. Technological developments bring new patient benefits, but also introduce new risks that must be better understood if the existing arsenal is to be used in the best possible way.
In recent years, the IRSN has decided to focus in particular on the side effects of abdominal, pelvic and thoracic radiotherapy. These "side effects" are now sometimes described as diseases in their own right: "pelvic radiation disease" and "radiation-induced pulmonary fibrosis", for example. The IRSN is working closely with the Gustave Roussy Institute on this shared objective.
Tissue response to ionizing radiation exposure is complex, and the mechanisms of tissue damage progression involve numerous molecules that eventually cause chronic injury. The main aim of the ROSIRIS experimental research program is to understand the physical and biological mechanisms that are triggered following exposure to radiation, from radiation energy transfer to molecules through to the later biological effects. Ultimately, the program should improve our ability to predict whether complications involving healthy tissue will occur, to optimize and potentially personalize radiotherapy protocols, and to discover new treatment approaches, both prophylactic and curative. The effects studied by the program are severe deterministic effects.
Left image: Representation of a cell nucleus used to simulate energy deposits in DNA. The three images show the three different scales: the top right shows the DNA molecules in the voxels (a notional intermediate structure with a volume of 50 nm cubed, containing a chromatin fiber fragment made up of 23 nucleosomes); the top left shows the voxel cubes; and below is the cell nucleus full of chromosomes created from voxels.
The bases and each sugar and phosphate are represented by spherical volumes that come together to form the nucleotides, followed by the DNA double helix wrapped around a histone to model a nucleosome and, finally, the chain of nucleosomes that form chromatin, the raw material of DNA. © IRSN/ www.nature.com/Scientific Reports
Five voxel configurations were created to enable chromatin fiber to be assembled flexibly within the nucleus (to form chromatin loops, for example). © IRSN/ www.nature.com/Scientific Reports
Challenges and Approach
Based on the current understanding of the pathogenesis of radiation-induced damage in healthy tissue, the Institute's teams identified the vascular endothelium as a key aspect of tissue response to radiation. Endothelial cells are in direct contact with the blood and play an essential role in maintaining vascular homeostasis and in tissue response to stresses such as radiation. They play a key role in the progression of radiation-induced damage, since they lose their ability to regulate the movement of inflammatory and immune cells between the blood and the irradiated tissue, which contributes to the failure to heal properly and the progressive development of fibrosis. The ROSIRIS program therefore elected to focus on this cell compartment.
The program is split into three phases:
- The first phase looks at the correlations between the spatial distribution of radiation energy deposits at a nanometric level and the initial biological events triggered by this radiation.
- The second phase seeks to understand the dynamics of the physiopathological mechanisms—primarily molecular—involved in the cellular response to radiation. It focuses on the response of the vascular endothelium
- The third phase aims to experimentally validate in vivo the physiopathological hypotheses and the main molecular actors identified as being involved in the development of side effects. It focuses primarily on fibrosis caused by abdominal, pelvic and thoracic radiotherapy.
Figure 1. Timescale of the effects of ionizing radiation from the moment of exposure onwards. © IRSN
Phase 1: Characterization of the relationship between the physical characteristics of ionizing radiation and the early biological effects it causes
This phase of the ROSIRIS research program aims to describe the causal link between the nature of the ionizing radiation and the biological events that are produced in the cell when it is exposed to this radiation. It adopts a multi-disciplinary approach comprising physics, biology, IT, and more. An in silico model is being developed to predict whether certain early biological effects will occur, based on the physical characteristics of the ionizing radiation used and the physical and chemical effects it is known to have on DNA. The Geant4-DNA simulation code and the DNA-Fabric software developed by the IRSN are used in particular for this project. A variety of biological effects can be simulated, but the IRSN is focused on quantifying double-strand breaks, which offer the added benefit of being observable within the cell nucleus thanks to the H2Ax protein that “activates” when close to these breaks.
The aim is to understand the various stages and key parameters that influence the nature of the first radiation-induced effects detected at the cellular level, as well as their likelihood of occurring depending on the nature of the radiation, using observations made at the molecular (DNA) level. The data used to validate the various stages of the model are primarily obtained from the literature, but are also taken from various collaborations, such as the BioQuart international project and the Geant4-DNA collaboration. The MIRCOM microbeam, introduced in 2018, can be used to gather new data to validate the models obtained with accelerated ions. In parallel to this, the simulation for gamma radiation primarily uses data obtained with the IRSN’s SARRP and ALPHÉE (medical LINAC) irradiators.
Phase 2: Modeling the overall molecular response of endothelial cells to radiation
In order to understand the dynamics of the molecular events involved in the endothelium’s response to ionizing radiation, IRSN researchers have adopted a recent approach that incorporates systems biology. The approach involves integrating different types of information (models, simulations, theories, experiments, etc.) in order to identify the interactions between the different “components” (cells, genes, proteins, molecules, etc.) of a biological system and how this system functions as a whole. Systems biology is based on the “omics” (genomics, proteomics, metabolomics, transcriptomics), mathematics and bioinformatics.
The aim is to decipher the molecular mechanisms of the endothelial compartment in response to radiation and model this response over time.
The work is designed to further develop the concept of relative biological effectiveness (RBE). Part of the research therefore focuses on using the two irradiators available at the IRSN (the SARRP and the ALPHÉE medical accelerator) to take various biological measurements of endothelial cells, including clonogenic survival, cellular senescence, cell cycle progression, apoptosis, transcriptome profile, interactions with circulating cells, and initial DNA damage. These measurements can establish whether the endothelium has been damaged or is dysfunctional following exposure to ionizing radiation. They are acquired for perfectly defined radiation conditions (i.e. a reference situation). The aim is to establish dose-response relationships for each biological parameter in the reference situation.
The IRSN’s second aim is to develop an in silico model to predict the molecular and cellular consequences of all types of irradiation. The model has to include all the previously acquired biological measurements, coded mathematically. The researchers study in vitro the impact on the biological measurements of changing the exposure parameters, by comparing them to the reference situation. The ultimate objective is to establish a multiparametric RBE (see box below on RBE).
In vivo validation of the physiopathological hypotheses used to determine the risk of radiotherapy complications, using transgenic animal models
The aim is to validate in vivo the hypotheses formulated based on “Modeling the overall molecular response of endothelial cells to radiation” (phase 2), and to validate in vivo the robustness of the predictive mathematical model that the teams aim to establish during phase 2 of the program. To achieve this, preclinical models are used in complex irradiation situations involving different doses, volumes or targeted organs. These might include localized irradiation of the digestive tract (“serial” tissue organization) or the lungs (“parallel” tissue organization). Transgenic mice models are used to study the role of the vascular compartment in the progression of tissue damage at the molecular level, in particular the specific deletion of certain key components in the endothelium. Some of these will be proteins identified in phase 2 of the program as being involved in the endothelium immune response that leads to radiation-induced fibrosis. These experimental models are designed to reveal the main actors and their associated functions involved in the progression of tissue damage, from early to late effects.
A demonstration of changes in glycosylation following irradiation. Healthy human endothelial cells were cultivated
in vitro following irradiation (20 Gy) or no irradiation (0 Gy). The cells were then incubated at different times in the presence of a fluorescent protein (here, a lectin, concanavalin A [ConA]) (in green). This lectin binds specifically to glycosylated radicals (highly mannosylated N-glycans) bound to endothelial cell membrane proteins and found on the external surface of the cells, which play a role in transporting immune cells during inflammation. The cell nuclei appear in blue, following specific staining with DAPI. The photos show the increased fluorescence caused by radiation compared to the cells that were not irradiated. ©
The main next step for the ROSIRIS program is to continue to optimize the nanometric-level energy deposit simulation in order to link the deposits to the observed biological variables (DNA damage, senescence). However, the results obtained thus far also indicate that the concept of RBE needs to be developed further, while new risk prediction models that are more robust than the existing ones need to be developed. On the biology side, ROSIRIS is currently working to characterize in detail the dynamic cellular and molecular processes of post-radiation vascular dysfunction. It is also studying the vascular endothelium/inflammatory response interface in relation to the progression of radiation damage in healthy tissue.
* Radiobiology of integrated systems for the optimization of treatments using ionizing radiation and evaluation of the associated risks
Further developing the relative biological effectiveness (RBE) measurement
At present, radiation protection risk assessments are based on relative biological effectiveness (RBE) measurements. RBE is defined as the ratio between doses of two types of ionizing radiation—or two different radiation techniques—that produce the same biological effect. This measurement is used in particular in the medical field to predict the biological effects of radiation and radiation protocols, on both cancerous tissue and healthy tissue.
RBE measurements are based almost entirely on cell mortality or the cell becoming permanently incapable of division. The measurements are obtained via the "clonogenic survival assay", which calculates the survival of cell lines irradiated
in vitro. This is the standard radiobiology assay and serves as the basis for the reference mathematical model (the "linear quadratic model") used to determine RBE. This model is still applied directly to clinical situations in order to predict risk, yet is not suited to the latest radiotherapy techniques, in particular the increasingly high dose rates and the use of fewer and fewer fractions. Furthermore, only the absorbed dose is used in these measurements to qualify the radiation used.
One of the final aims of the ROSIRIS project is to develop multiparametric RBE measurements to provide more realistic predictions of the risk associated with using ionizing radiation as part of new radiotherapy techniques and practices (high-dose fractions, high dose rate, stereotactic radiotherapy etc.), and to accompany these with a micro- and nanodosimetric characterization of the radiation in question.