Ethical Considerations for MRI Research in Human Subjects in the Era of Precision Medicine.
Hui Mao,Eduardo A. Garza-Villarreal,Linda Moy,Tarique Hussain,Andrew D. Scott,Janine M. Lupo,Xiaohong Joe Zhou,Candace C. Fleischer
DOI: https://doi.org/10.1002/jmri.28969
IF: 4.4
2023-01-01
Journal of Magnetic Resonance Imaging
Abstract:In the field of magnetic resonance imaging (MRI), the involvement of human subjects is an essential part of research investigations, technology development, and patient care. High-definition anatomical, physiological, functional, and molecular information offered by MRI is increasingly sought after in scientific discovery and precision medicine. Furthermore, the latest developments in translational and clinical research, including MRI-guided interventions and artificial intelligence-based algorithms, rely on the wealth of MRI data generated from studies on patients. As we continue to advance MRI technology to empower research and clinical care, the recent shift to an era of precision medicine must be examined in the context of the relevant ethics when involving human subjects in research studies. The opening plenary session on “Moral & Ethical Issues in MRI Research” at the 2023 annual meeting of the International Society for Magnetic Resonance in Medicine (ISMRM) brought several important issues to the forefront for the community to consider. Here, we highlight key points from the session within the framework of the Belmont Report introduced 45 years ago, specifically the importance of Respect for Persons, Beneficence, and Justice. In the modern era, investigators must consider a broad range of characteristics, including biological variables, socio economic status, cultural heritage, and life experience. The early years after adoption of the Belmont Report and other guidelines for research using human subjects, such as guiding principles for ethical research by the U.S. National Institutes of Health Clinical Center,1 were focused on establishing regulations for monitoring safety, rights, and well-being of human subjects in research. More recently, increased emphasis has been placed on the validity and quality of scientific research, including the rigor of the approaches taken, reproducibility of data, and quality of unbiased study design. A key consideration for precision medicine is the scope of inclusion for research cohorts to encompass diverse health conditions and populations. For example, it is disturbing to realize women are still underrepresented in acute coronary syndrome trials with a low ratio of participation to prevalence of 0.6,2 which may be responsible, in part, for the wide differences in healthcare outcomes. As many diseases, including cancer and neurodegenerative disorders such as Alzheimer's disease, are highly heterogeneous, recruiting a diverse set of subjects to capture within and between subject variability is critical for validating novel technologies.3 Furthermore, epigenetic research has shown disparities in healthcare conditions and long-term environmental and socioeconomic effects can alter biological and genetic traits of healthy individuals and patients.4 Using existing datasets for developing algorithms for artificial intelligence that do not take into consideration a wide breadth of factors can result in biased and ungeneralizable results, even when the same experimental conditions are utilized.5 Understanding individual variability and incorporating thoughtful approaches to design human subject studies are particularly relevant in clinical trial design and assessment of drug efficacy as the scientific community moves toward more individualized, adaptive trial designs, and personalized medicine practices. Over representing vulnerable or specific groups is one option to address the apparent bias in existing cohorts and recruitment. Importantly, developments in MRI technology continue to enable more inclusive and diverse human subject cohorts. A notable example is the rapid technical advances to image infants and underserved populations, as most recently demonstrated by the development of portable and low-field MRI scanners.6 With the goal of generating beneficial and broadly translatable results, the plenary session also highlighted the importance of MRI safety and, more broadly, beneficence of our research to both current and future patients. There is a clear benefit in incorporating input from both MRI and non-MRI experts to ensure appropriate study design, safety of volunteers and patients in clinical trials, and data interpretation. As we develop cutting-edge methodologies using healthy volunteers for eventual clinical applications in patients, there is an important balance between relentlessly pursuing innovation and carefully assessing the risk to the subjects, particularly those with unique health or physical conditions introduced by recruiting a more diverse and heterogeneous study population. The potential safety concerns from powerful new technology must be carefully managed. The ISMRM community has established an excellent track record in assessing and mitigating the risks involved in new and advanced technologies developed in the field.7 Our continued responsibility is particularly crucial as inclusion criteria are expanded to include more diverse study populations and data sources become more open. Data sharing among scientific communities in different fields is essential to maximize the benefits and improve outcomes of human subject studies, as is unbiased testing and validation of new findings and technologies. The push toward more open science, however, also imposes challenges for data safety management.8 The delicate balance between open science and protecting the individual right to privacy is a key topic of consideration for our global MRI community. Federated learning is a potential approach for improved security, where machine learning is performed at local sites using single MRI datasets and updated neural network weights are then shared between sites. The number of clinical trials that incorporate and implement MRI methods continues to increase globally.9 International collaborations offer opportunities for broadening access to geographically, ethnically, and socioeconomically diverse populations that are critical for democratizing and globalizing MRI technology. While different regions and countries may, and should, have their own organizations and agencies that develop guidelines for research in human subjects, it is important to recognize and respect the cultural and regulatory differences before reaching a consensus for establishing fundamental principles and universal standards.10 Without acceptable global systems and policies in place, a basic understanding and good practice of ethical research in human subjects, along with the unwavering integrity held by each investigator, are the foundation for formulating and developing cross-border collaborations. ISMRM, as a scientific community of investigators in this field, is well-poised to play a prominent role in educating, promoting, and facilitating research in human subjects performed with the utmost ethical standards, highlighting the continued relevance of the Belmont Report nearly half a century later.