Molecular imaging biomarkers hold the promise to provideunparalleled opportunities for elucidating disease phenotypes, understanding pathophysiology, facilitating development of targeted therapies and driving personalised medicine. To realise the full potential of imaging biomarkers, our scientific community is compelled to translate them into clinical research and practice. This talk will present the current and emerging imaging biomarkers in clinical practice, including MRI and hybrid-PET imaging biomarkers, and describe the applications, strategies, opportunities and challenges in clinical translation.
Molecular imaging biomarkers are tools that can be used to visualize abnormal processes at a molecular or cellular level of function. Its capabilities represent a paradigm shift from conventional anatomical and functional imaging and when translated to the clinics, is powerful in driving personalized medicine and health care.
The role of imaging biomarkers in the clinics is manifold; it includes identifying disease phenotype, patient selection and risk stratification, tailoring dose, response assessment to therapy, and drug development. In clinical trials, biomarkers have been found to provide complementary information, and outperform conventional size-based criteria as surrogate end-points which could reduce the size, complexity and duration of clinical trials (1). Molecular imaging that may be used in the clinics include magnetic resonance spectroscopy (MRS), chemical exchange saturation transfer (CEST) and hyperpolarized MRI using endogenous information, and a host of exogenous direct molecular imaging probes of physiology, metabolism and specific biological targets using MRI contrast agents, positron emission tomography (PET) and single photon emission computed tomography (SPECT) imaging. Whilst the opportunities for molecular imaging biomarkers are unparalleled, it is widely recognized that the process of clinical translation is challenging with only a few having crossed the translational gaps into clinical practice.
Biomarkers should be biologically specific, accurate, reproducible, standardized, easy to implement in the clinical workflow and of impact in clinical practice. The imaging biomarker ‘roadmap’ has been described as follows (2): the process of discovery, followed by technical validation which serves to test the accuracy, precision, repeatability and reproducibility of the biomarker measurement, and biological and clinical validation that establishes the biomarker as a surrogate for a specific biological feature and clinical outcome variable respectively. Technical, biological and clinical validation are parallel processes occurring concurrently and extends from single center to multi-center studies. Moreover, the process may involve pre-clinical imaging in animal models for validation with biology/histopathology and intervention. The biomarker should be qualified for a particular clinical use and evaluated prospectively in multi-center trials. Finally, the biomarker becomes a clinical tool used in patient management by demonstrating its clinical utility on health outcomes together with evidence of cost-effectiveness.
Fundamentally, the biomarker must be of clinical relevance and validated against clinically meaningful endpoints, preferably addressing a high priority question/knowledge gap. This requires an in-depth understanding of the complexities of biology, diseases, and its management. The role of the imaging biomarker must be considered vis-a vis other non-imaging molecular biomarkers, including laboratory tests, genomics/proteomics and liquid biopsies for example. Compared to pathology and laboratory tests, imaging is unique in that it can capture information of spatial and temporal heterogeneity (of cancer, including intratumoral and inter-tumoral heterogeneity with the use of whole-body imaging which can also identify evolving clones on serial imaging), assess disease extent and distribution and direct tissue sampling. It is expected to have a complementary role to, and integrated with clinical parameters and other molecular diagnostics for patient management (3).
Practicality and simplicity are key in the clinics, be it in aspects of standardization, data acquisition, analysis or interpretation (4). Many quantitative imaging biomarkers are challenging to implement because of the complexity of the techniques. Practical considerations of scan time, motion sensitivity, need for exogenous contrast administration, ease of scanner optimization for good quality data acquisition are factors to be considered in clinical practice. The level of expertise required for of data analysis, including the need for use of off-line programs must be considered. Imaging parameters should be simplified from continuous variables to categorical data using putative cut-off points so that it can be easily applied to clinical decision making. Many of the above processes have not been standardized, and hence causing added variability and limiting its application in multi-center studies (5).
The focus of the development process should not end with technological development. Collaboration with the clinical team of technologists and radiologists is essential for successful clinical translation, and buy-in by the referring clinician counterparts is critical. It will require all stakeholders to embrace this paradigm shift in clinical imaging in order to bring these tools into clinical practice.