Precision Medicine and Precision Diagnostics

John Tomaszewski, MD, is a SUNY Distinguished Professor and Peter A. Nickerson, PhD Chair of Pathology and Anatomical Sciences.

By John E. Tomaszewski, MD

What if every medical decision always resulted in the best outcome for every patient? This, of course, is a cherished dream of all patients and medical practitioners. Until recently this dream has seemed to be an utter impossibility. But now, maybe not? The first part of the 21st century has seen an explosion in medicine’s access to “big data” in the digital forms of human molecular biology, radiology and cell and tissue imaging. All of these approaches have expanded our understanding of the complexities of human biological systems, and these approaches have been made even more powerful and accessible by the advent of modern computing. This means that fast computers running advanced algorithms may be able to help physicians make important decisions by looking for patterns in the digital data. These digital data come from gene and protein analyses, high-resolution microscopy, and radiological examinations of human tissues and organs. The decisions made from massive digital data may be better because they take into account lots of very detailed quantitative data sources, which makes for the possibility of more precise choices. This sequence is known as the big-data-to-big-knowledge paradigm, and it is the process that can fuel choosing the best diagnosis and the best treatment for an individual patient every time. We call this process precision medicine.

Precision diagnostics, in turn, is the toolbox which pathologists and laboratory medicine physicians use to assemble precision diagnostic statements. How does precision diagnostics work? An example in breast cancer care may help. If a woman has a small breast cancer that is completely excised, she has an overall good prognosis, but she still has some risk for tumor spread in the years to come. We know that this risk can be dramatically reduced by adjuvant chemotherapy. There are however many risks with chemotherapy, and the majority of these women really do not need it. So how could we make a precise decision on which women will progress, offer them treatment and avoid the adjuvant therapy risks for the rest? Tests of multiple genes associated with breast cancer progression can be performed on the tumor tissue and/or we can use computer algorithms to precisely exam the pixels in microscopic pictures from the breast cancer. Amazingly, each method can precisely diagnose which woman needs the adjuvant chemotherapy and which woman does not. The best precision diagnostic tests predict a patient’s response to a chosen therapy such that almost 100 percent of the treated patients have a good response. Precision diagnostics makes precision medicine possible.

So, what will the practice of precision medicine and precision diagnostics look like in 25 years? Physicians will continue to start by taking a patient history. The patient’s signs and symptoms along with their full medical history will be entered into a precision diagnostics artificial intelligence (AI) system, and the physician along with the AI assistant will choose the best precision diagnostic tests. Noninvasive high-resolution cell and tissue imaging will be used in targeted scans to find disease hot spots. The data from the imaging, along with molecular microsampling of these hotspots, will be used by the physician and AI assistant to create precision medicine treatment plans that are 99 percent effective.

For the last 10 years at UB, the departments of Pathology and Anatomical Sciences, Biomedical Engineering, Biomedical Informatics, Computer Science, and Materials Design and Innovation, and such centers as the New York State Center of Excellence in Bioinformatics and Life Sciences, have been building the basic science programs on which precision medicine and precision diagnostics will be established. Because of this, our university is well positioned to make the jump to this new paradigm of medical practice.