„Translational studies on microRNA biomarkers – how to integrate clinical insight, biostatistics and artificial intelligence”
Medical University of Łódź, al. Kościuszki 4, 90-419 Łódź, Poland
Translational medicine is a broad term encompassing all studies that aim at bridging the gap from the laboratory bench to the bedside. Mine and prof. Chowdhury’s teams have collaborated for the last 8 years on various biomarker studies aiming to bridge the gap between basic science and the clinical world with its distinct needs and challenges. The main area of our focus is detection of radiation exposure and the focus to harness the potential of miRNAs associated with this for diagnostic use. Initial studies in mice showed that a signature of irrevocable bone marrow destruction may be defined and proven through in depth mechanistic studies, that such microRNAs are a specific feature of bone marrow stem cells losing their potential to repopulate the irradiated individual leading to death unless allogenic bone marrow is transplanted [1]. The translational potential of such a test was obvious, but obvious ethical considerations precluded the calibration of it on humans. To bring it closer to the bedside we devised a framework for diagnostic test design and validation and supported our claim with bioinformatic analyses of evolutionary conservation predicting the test to work in humans. Given the test excellent performance we hypothesized that circulating microRNAs could be used for diagnostics of malignancies and turned our attention to ovarian cancer. Due to the inherent clinical complexity of human studies, we decided to perform experiments on non-human primates [2]. Surprising overlap between the species allowed to make informed projections onto humans and a metaanalysis of available studies on microRNAs in radiotherapy largely confirmed our estimates on evolutionary conservation of mechanisms that drive the miRNA response to radiation [3]. This project necessitated the use of high-level data mining techniques and multilevel validation procedures documenting the appropriateness of biomarker and classification method selection, the universality of the chosen miRNAs when quantified using different molecular techniques, their specificity towards ovarian cancer and finally, the accuracy of the test on a separate clinical group [4]. Ultimately, the project was wrapped up with a website for easy access of researchers and doctors alike. The projects described above outline the process of devising applicable biomarker tests and highlight the need for seamless integration of biological, clinical, statistical and bioinformatic approaches to solve modern health challenges. The challenges encountered when working with microRNA biomarkers will be highlighted during the talk to highlight the potential and limitations of this novel class of biomarkers in translational medicine.
References
1.Acharya SS, Fendler W, (…) Chowdhury D. Serum microRNAs are early indicators of survival after radiation-induced hematopoietic injury. Sci Transl Med. 2015;7(287):287ra69. doi:10.1126/scitranslmed.aaa6593. 2.Fendler W, (…) Chowdhury D. Evolutionarily conserved serum microRNAs predict radiation-induced fatality in nonhuman primates. Sci Transl Med. 2017;9(379). doi:10.1126/scitranslmed.aal2408. 3.Małachowska B(…) Chowdhury D, Fendler W. Circulating microRNAs as Biomarkers of Radiation Exposure: A Systematic Review and Meta-Analysis. Int J Radiat Oncol Biol Phys. 2020;106(2):390-402. doi:10.1016/j.ijrobp.2019.10.028. 4.Elias KM, Fendler W, (…) Chowdhury D. Diagnostic potential for a serum miRNA neural network for detection of ovarian cancer. Elife. 2017;6. doi:10.7554/eLife.28932.