Medical procedures are the largest source of exposure to man-made radiation in Europe. This medical use of radiation has greatly improved diagnosis and treatment, and is therefore justified by a net benefit to the patient. According to European legislation, radiation doses received by individual patients must be optimised in order to maximise the benefit/ risk ratio to that patient. The past two decades have seen the competitive development of innovative medical technologies using ionising radiation which offer new diagnostic and therapeutic opportunities and require anticipation of potential risks. This is especially true for the diagnosis and treatment of cancer patients.
The ultimate goal is to optimise the use of ionising radiation for the diagnosis and treatment for each patient through an individualised approach based on generalised use of the most advanced risk-informed exposure protocols across Europe. However, significant differences in medical radiation exposure can be observed between different European countries, for similar protocols. Therefore, it is of great importance for the radiation protection community to conduct research, development, and innovation to further optimise and standardise the medical application of ionising radiation, and to disseminate best practice protocols throughout Europe, especially with respect to protecting patients from the adverse health effects of ionising radiation.
The MEDIRAD project has addressed specific areas of importance identified from clinical needs, with the aim of optimising radiation protection of patients and medical staff. While MEDIRAD Recommendations 1, 2, and 3 include further research needs specific to the technical fields covered by these recommendations, Recommendation 4 focuses on five key research issues which policymakers and relevant research communities are invited to consider.
The overarching perspective of the MEDIRAD project is to demonstrate the added value of close cooperation between medical and radiation sciences in addressing complex research questions. This set of key research issues does not seek to establish a new strategic research agenda, but rather highlight the strategic significance of addressing these five key research issues with adequate resources, and through the close cooperation of medical and radiation research communities.
The growing number of cancer survivors in Europe, over half of whom are treated with radiation therapy, emphasises the priority to prevent or mitigate radiation-adverse treatment effects, which can have a significant impact on treatment outcome and on the patient's health and quality of life (QoL) . Radiation effects in normal, healthy, tissue surrounding the tumour depends on many parameters, including the tissue type, dose/volume of the irradiation, functional status of organs at risk, age, smoking habits, comorbidities (diabetes, collagen vascular disease), and the patient’s genotype . Current understanding of the biological mechanisms underlying radiation adverse effects (cancer, fibrosis, non-cancer disease such as cardiovascular disease), combined with advanced methodological frameworks such as the adverse outcome pathway (AOP) approach, will facilitate research to identify and validate sets of molecular (genetic and epigenetic) biomarkers  of healthy tissue response, sensitivity, and adverse effects and of patient radiation susceptibility. These potential biomarkers could then be used to personalise treatments and support patient follow-up including treatment of adverse effects on healthy tissue when necessary.
Historically, diagnostic and therapeutic applications of ionising radiation have relied on consensual clinical protocols with standardised clinical dose and dose/volume recommendations. MEDIRAD stakeholders identified the development of more personalised diagnostic and therapeutic protocols as a need with significant impact on clinical practices. Given the large amounts of multisource patient data produced during routine clinical activities, AI and machine learning will play a major role in the use of big data for improving diagnostic and therapeutic applications of ionising radiation by helping to translate multisource data into clinical decision aids . AI will help to develop diagnostic and treatment approaches that are better tailored to the specific characteristics of the patient and to improve therapeutic outcomes and minimise short and long-term adverse effects of radiation.
While the role of ionising radiation exposure in inducing DNA mutations is undisputed, research in the last decades has uncovered a multitude of other biological effects that contribute to the induction of cancer or other pathologies such as cardiovascular disease [5,6]. Key biological events in response to radiation exposure have been identified at different biological levels: from genes, RNA or proteins, to cells, tissues and organs, and they are interrelated. The challenge for future research is to develop models which correctly aggregate these biological effects to provide clinicians with advanced predictive tools.
The adverse outcome pathways (AOP) approach was designed by the OECD [7,8] to understand the complex health effects of toxic chemical compounds. European radiation and clinical research groups would benefit from the AOP investigation methodology by providing a common framework for scientific and clinical data integration. It would also open the way to future models that could take into account the effects of combined oncological treatments associating chemotherapy with radiation therapy.
The use of ionising radiation in medicine represents a tremendous benefit for the diagnosis and treatment of diseases. While the benefits to the patient largely outweigh the risks, assessing the long-term adverse effects of radiation exposure is particularly important in cohorts of patients who may live for decades after exposure, particularly children . Large scale clinical epidemiology studies, that follow exposed patients for decades, are essential to identify and quantify the late health effects of medical low, moderate, and high radiation doses and provide the basis for the implementation of high standards for quality and safety of medical radiation applications [10-12]. The success of such studies relies on careful patient follow-up and collection of detailed patient demographic, clinical, and dosimetric data, images and biological samples (through linkage with clinical, radiological, and therapeutic records).
The use of ionising radiation in medicine may present some risk especially in situations that require repeated imaging for diagnostic, planning, or staging purposes and that result in non-negligible exposures, at least in certain body regions. Thus, there is a constant need for optimisation towards improved benefit/risk ratios and individualised procedures. Such optimisation will include the use of new methods and technologies, for better and more reliable diagnosis, the use of new evaluation techniques, such as those based on AI applications as described above, as well as the optimisation of existing procedures in terms of improved benefit/risk ratios on a population basis and for each individual patient. To achieve the latter, it is necessary to investigate the patient exposure in a more accurate and meaningful basis and correlate this information to the required image quality, which needs to be analysed and evaluated for each imaging procedure of each patient.