Radiotherapy is an important modality for treating cancer, either alone or in combination with other modalities such as surgery, chemotherapy and immunotherapy. It works by damaging the DNA of cancer cells directly or indirectly, causing the cancer cells to die or stop them from dividing. Radiotherapy can also influence our immune system, but the mechanisms behind this interaction are largely unknown.
We are developing improved strategies for combating cancer with radiotherapy and particle therapy
Ionizing radiation inevitably also kills healthy tissue surrounding the tumor. Healthy tissue has an increased repair capacity compared to tumorous tissue that we exploit in radiotherapy by treatment delivery in fractions over several weeks. Precision is crucial in external beam radiotherapy. This is because we focus radiation into the tumor by sending beams through the body from several directions. Image-guidance ensures that the radiation actually hits the tumor. Increased precision spares the surrounding healthy tissue, and enables treatment with higher doses and fewer fractions without an unacceptable risk of adverse effects.
The biological response to radiation can vary a lot from tumor to tumor and within a single tumor. For instance, parts of the tumor with reduced access to oxygen can be resistant to radiation (oxygen is a mediator for indirect damage on DNA). Tissue and blood samples as well as functional imaging (e.g. MR and PET) can provide valuable information about the tumor biology, enabling us to deliver a more personalized radiation treatment.
Particle therapy differs from conventional radiotherapy with photons by interacting in a more delimited area in the patient, thus damaging DNA more efficiently. Particle therapy gives less radiation to the surrounding healthy tissue, thereby reducing the risk of adverse effects such as organ damage or secondary cancers. However, we need more knowledge on the biological effects of particle therapy and on how to better clinically exploit the improved precision and biological effect.
About the initiative
The project Establishing a framework for interdisciplinary clinical particle therapy research at Haukeland University Hospital was established in 2018 with generous support from Trond Mohn Foundation, Norwegian Cancer Society, Helse Vest RHF, Haukeland University Hospital (HUH) and the University of Bergen (UoB).
The motivation behind this project was to join the forces and competence of clinicians and researchers to improve cancer care, in the context of the introduction of particle therapy in Bergen from 2024. It was our belief that high quality clinical research in radiotherapy and particle therapy would demand a framework with well-functioning interdisciplinary teams of researchers and clinical staff, scientific support to aid implementation of advanced methods and clinical translation, high quality data as well as international collaboration. These values have therefore been prioritized in both budget and activities in addition to carrying out the outlined research. Currently more than ten researchers in addition to clinical staff are daily involved in the project. Furthermore, a key role of this project has been supporting and collaborating with the other ongoing research projects in particle therapy in Bergen (i.e. at HUH, the UoB and Western Norway University of Applied Science (HVL)).
The long-term aim of the common research activities is to develop and implement technology, models and strategies to offer improved treatment with modern photon and particle therapy. This will allow for increased tumor control and/or reduced probability of normal tissue complications.
Specifically, we will:
- Perform comparative planning studies of photon and particle therapy along with predictive risk models for selected tumor sites to be used for defining clinical protocols.
- Develop and improve biological dose-response models for pelvic irradiation and establish constraints for normal tissue complications for photon, proton and carbon ion therapy.
- Expand the spectrum of suitable indications for particle therapy delivery by developing technology and strategies for image-guidance and adaptive therapy.
- Develop models to predict for patient-specific internal organ motion in the pelvis and head and neck for use in optimization of particle therapy delivery.
- Improve planning of particle therapy by quantitative use of functional imaging to re-distribute dose and LET within the tumor to maximize tumour control.
WP1: Clinical data collection and comparative planning
In WP1 we have ongoing studies collecting clinical and image data, respectively, for head and neck and lung cancer patients. We are also performing comparative planning of photon and proton therapy for children, head and neck, prostate and lung cancer patients.
WP2: Integration of functional imaging in particle therapy planning
In WP2, we have explored two different methods for quantifying hypoxia from functional images and using this information in shaping the target dose distribution with proton and photon radiotherapy.
WP3: Organ motion modelling
In WP3, we are developing a statistical motion model for pelvic organs and furthermore we are investigating the impact of interplay on dose delivery in treatment of lung cancer with scanning beam particles.
WP4: Image guidance and adaptation in particle therapy
In WP4 we have an ongoing study with an adaptive protocol for lung cancer. In addition, we are investigating methods and technology to reduce uncertainties related to estimation of proton stopping power for proton dose calculation and range uncertainties in collaboration with the University of Bergen and Western Norway University of Applied Science.
WP5: Dose-response modelling
In WP5 we have explored the impact of the variable relative biological effectiveness (RBE) of proton therapy in children and head and neck cancer patients.
WP6: Clinical integration and research infrastructure
In WP6 we have developed various tools to aid the ongoing research activities.
Department of Oncology and medical physics
Haukeland University Hospital
Liv Bolstad Hysing
Phone: +47 55 97 77 74
Liv Bolstad Hysing (project leader), Sara Pilskog, Camilla Stokkevåg, Marianne Brydøy
Olav Mella (project responsible), Anfinn Mehus, Ása Karlsdottir, Helga Gripsgaard
Scientific support group (interal):
Olav Mella, Odd Harald Odland, Olav Dahl
Scientific advisory board (external):
Cai Grau, Ivan Richter Vogelius, Antje Knopf
Publications & Outreach
Bergen Particle Therapy on GitHub
Bergen Particle Therapy on Twitter
Helge Egil Seime Pettersen, post. doc. / research support (2019.01.01 – 2021.12.21)
Helge assists with the various research activities by developing methods and software for their completion. A main project is a user friendly application for calculation of dose parameters from Treatment Planning Systems: this to compare between patients, patients cohorts and projects. A module for the calculation of normal tissue toxicity is also present with uncertainty modelling through bootstrapping.
Helge is also an active participant in the Bergen proton CT project: this is a modality to measure how the protons interact with the patient, a large step towards personalized proton therapy with reduced toxicity. His work focuses on Monte Carlo simulations of the detector design and development of methods and software platforms for particle tracking and image reconstruction.
Grete May Engeseth, PhD student (2018.01.01 – 2021.06.30)
Brain necrosis is a serious, but rare side effect after radiation therapy. Radiological changes on MR images are considered early signs of brain necrosis and the patients are diagnosed based on such image changes. In the first of three sub-studies, the outcome and patterns of the radiological changes were explored: in how many patients do these image changes occur, in what part of the brain are they located and what kind of radiation dose increases the risk of developing these image changes? In the second study, proton treatment plans and follow-up images of a small group of patients were used to explore the relationship between dose, the specific energy transfer from the proton beam (LET, Linear Energy Transfer) and the radiological changes. In the third study, models were developed to predict radiological changes in the temporal lobe of the brain using different dose models for radiobiological dose effect (NTCP, Normal Tissue Complication Probability).
The project was partly carried out at the University of Texas MD Anderson Cancer Center (Houston, TX), where Grete had a 15-month research stay, and partly at HUS and the Faculty of Medicine, UiB, in collaboration with i.a. supervisors Liv Bolstad Hysing, Camilla Stokkevåg, Marianne Brydøy and Olav Dahl at HUS and Gary B. Gunn at MD Anderson Cancer Center.
Øyvind Lunde Rørtveit, PhD student (2019.05.01 – 2022.05.01)
The aim of Øyvind’s research is to develop personalized models for motion uncertainties during a treatment course in radiotherapy. By using personalized models we expect a reduction in the irradiated volume, thus reducing the risk of normal tissue toxicities. In the first study, we have developed a population model for organ deformation for use in robust radiotherapy planning. This was done by identifying the most common form changes in rectum in 37 prostate cancer patients through Principal Component Analysis (PCA). Then, such a population model can be applied together with the patient’s CT images to create a personalized treatment volume around the rectrum, taking into account the expected changes.
The project is performed at HUH and Department of Physics and Technology (UoB) together with the supervisors Sara Pilskog, Liv Bolstad Hysing and Andreas Stordal at the Norwegian Research Center NORCE.
Kristine Fjellanger, PhD student (2020.07.01 – 2023.06.30)
The current standard of care in inoperable locally advanced non-small-cell lung cancer (LA-NSCLC) (stage III) is chemotherapy and radiotherapy, with a 5-year overall survival around 30%. Attempts to escalate radiation dose from 60—66 Gy have failed to improve survival due to toxicity of major critical organs such as the lung and heart. The aim of this project is to develop improved treatment strategies with photon and proton radiotherapy for LA-NSCLC that reduce the risk of treatment-induced normal tissue toxicity to facilitate safe dose escalation, thus increasing the 5-year overall survival.
In the first study the treatment quality of the state-of-the-art radiotherapy will be improved by the introduction of machine learning into the treatment planning workflow. In the second study, Kristine will develop a model for personalized selection of patients for proton/photon therapy, and for the usage of breathing-controlled treatment (free breathing vs breath hold). In the last study Kristine will develop an adaptive protocol for proton therapy of LA-NSCLC, where the treatment is adjust during the treatment course by means of updated imaging.
The project is performed at HUH, the Department of Physics (UoB) and Erasmus MC (Rotterdam, NL) together with the main supervisor Liv B. Hysing, co-supervisor Helge E. S. Pettersen and co-supervisor Ben Heijmen at Eramus MC.
Camilla Grindeland Boer, MSc student
Patients with locally advanced non-small-cell lung cancer (LA-NSCLC) may gain from proton therapy in order to reduce the toxicity to the spinal cord, heart, lungs and esophagus. Due to e.g. the breathing motion it’s challenging to treat this patient group using proton therapy, and usually clinical usage of protons are reserved for patients with limited body deformation due to breathing.
Camilla is researching with different strategies for breathing motion management in proton therapy by using four-dimensional CT (dynamic images) acquired in the beginning of the six weeks long treatment course. Furthermore, Camilla is comparing proton therapy treatment plans with the currenly applied (and clinically delivered) intensity modulated radiation therapy (IMRT).
The project is performed at HUH and the Department of Global Public Health and Primary Care UoB, together with supervisors Liv B. Hysing and Grete May Engeseth at HUH, and Una Ørvim Sølvik at the Department of Global Public Health and Primary Care, UoB.
Annette Høisæter, MSc student (01.01.2020 – 01.12.2020)
Knowing where the protons’ stop is vital when planning for proton therapy. The commonly applied method for identifying this is a calibration between the attenuation of photons from CT to the Relative Stopping Power (RSP) of protons. However, more and more centers apply Dual Energy CT for this calibration, where usually two separate energy spectra (DECT) are compared to give a more accurate measure for the RSP. Then, proton therapy can be given more precisely and thus with lower risk of late effects.
Annette’s research focuses on the implementation and subsequent evaluation of several methods for the calibration of CT and DECT to yield RSP. By means of this project the medical physicists at Haukeland gains important knowledge about the available methods and their usage.
The project is performed with the supervisors Helge E. S. Pettersen and Kirsten Bolstad (HUH) and Kristian Ytre-Hauge at the Department of Physics (UoB).
Christoffer G. Moen, MSc student
The aim of the study is to investigate the differences in dose distributions of advanced photon therapy versus proton therapy for head and neck cancer patients. The influence of anatomical changes and uncertainty in the relative biological effectiveness (RBE) of protons is included.
The project is performed at HUH and the Department of Global Public Health and Primary Care UoB, together with supervisors Camilla Stokkevåg and Tordis Dahle at HUH, and Una Ørvim Sølvik at the Department of Global Public Health and Primary Care, UoB.
Andreas Havsgård Handeland, MSc student
Brain stem necrosis is a rare, but severe side effect following radiotherapy. The Master's project investigates different methods used in the analysis of proton therapy dose/volume parameters related to brainstem toxicity. Cluster analysis is used to explore the association of necrosis and dose to the brain stem substructures. Additionally, dose-response models are generated based on Lyman-Kutcher-Burman parameters.
The project is performed with the supervisors Camilla Stokkevåg and Helge E. S. Pettersen at HUH and Department of Physics (UoB).
Local and regional collaboration
Molecular mechanisms and dose limits for radio- and particle therapy
Project leader: Professor
Olav Dahl, Department of Oncology and Medical physics, Haukeland University Hospital and University of Bergen. Project period 2016 -2019.
Carbon ions are more precise in dose delivery compared to protons, and have an increased biological effect by inducing more DNA damage that is difficult for tumor cells to repair. While the effect of photons and protons is reduced in areas with limited access to oxygen (hypoxia), the effect of carbon ions is similar as in well oxygenated areas. Hypoxic regions are common in most tumors and is one of the reasons why many patients are not cured by state-of-the-art photon therapy. Cells in resting phase are more vulnerable to carbon therapy.
The project is performed in collaboration National Institute of Radiological Sciences (NIRS) and Gunma University, Tokyo, Japan as well as Centro Nazionale di Adroterapia Oncologica, CNAO, Italy and Med Austron Carbon Center, Wien, Austria.
Neutron and gamma-ray imaging for real-time range verification and image guidance in particle therapy (NOVO)
Improved technology for range verification could be important for increasing precision in dose delivery with proton therapy to e.g. non-stationary tumours, and thereby exploiting the full potential of proton therapy for e.g. biologically guided dose escalation. Our collaborators at Western Norway University of Applied Science and the University of Bergen are developing the proof of concept of a new range verification system.
The idea behind this new range verification system is to combine imaging, timing and energy spectroscopy of fast neutrons in addition to commonly used prompt gamma radiation species. Fast neutrons are usually neglected or regarded as noise in state-of-the-art range verification systems. We presume that such a multi-particle and multi-feature approach to range monitoring will not only result in improved detection efficiencies and counting statistics, but also increase the amount of independent information obtained from the system that is needed to obtain a high degree of accuracy and reliability.
The project is being carried out together with project leader of NOVO Ilker Meric at Western Norway University of Applied Science, the University of Bergen and several European universities and research centres with funding from the
Norwegian Research council
Proton CT – Coming soon
Relative biological effectiveness (RBE)
In clinical practice, proton therapy is assumed to be 10% more efficient than conventional photon therapy. In other words the relative biological effectiveness (RBE) of protons is 1.1. The actual relationship in biological effect is more complex, and this is the topic of an ongoing collaborative project with the University of Bergen.
RBE is not constant, but dependent on both biological factors such as cell and tissue type, physical factors such as the linear energy transfer (LET) and also the endpoint for measuring biological effect. LET varies along the beam path and is highest for low energy protons, i.e. when the proton is coming to a stop. The concern that high LET values can cause unexpected complications is one of the reasons why the distal fall-off of proton beams are directed away from organs at risk. Monte Carlo simulations can be used for calculation of the LET distribution inside the patient, and together with clinical data we can use this to learn more about the biological effectiveness of proton therapy. Experimental in vitro data are usually used for developing phenomenological models which are more sophisticated than what is state-of-the-art clinically. There is a large uncertainty in estimation of the actual biological effects of proton therapy relative to conventional photon therapy.
In this project, led by Kristian Ytre-Hauge at Institute of Physics, University of Bergen, models for estimation of RBE are developed together with Monte Carlo tools for simulating LET and RBE in patients treated with proton therapy abroad, such as in Jacksonville, Florida, USA.