Pathology and Radiology Beyond Looking at Pictures

Abstract

I have seen a lot of innovations during this meeting that we can almost copy and use in radiology. The last speaker mentioned that there is going to be a potential merger between radiology and pathology, not just in the clinical world but also in working out systems biology and personalized medicine. For us to merge, we have to first look at how radiologists see the world of genomic medicine and personalized medicine. That is what I am going to do. I am also going to examine how we can put radiology and pathology together so that we fundamentally change the way that we practice medicine in the future.First of all, I have to say that without pathology, there is no radiology. The way we practice radiology today is totally dependent on 100 years of experience of correlating pictures with pictures. What we do today is look at the images we generate and, in our mind, go back to the correlative pictures and say, if I see a picture like this, in a computed tomography scan, what are the likelihoods that this will be a lymphoma or other disease. How do we get this experience? It is from radiologic/pathologic correlation.Ninety-nine percent of radiology residents go through the Armed Forces Institute of Pathology to learn radiologic/pathologic correlation. Without pathology, there is no radiology. It is as simple as that. But this paradigm is likely going to change because currently what we are looking at is doing diagnosis and therapy; in the future, what we are thinking about, in the era of genomic medicine, is trying to predict which patient will have high susceptibility for certain diseases, do focus screening, prevention, early detection, and personalized treatment and then monitor appropriately. That is really the view of the world in the next decade and beyond. So how are we going to make that happen? With radiology and pathology working together things can happen a lot quicker.Let us now look at the development of genomic medicine and molecular biology. It really is a rapid progression from the discovery of the double helical structure of DNA to mapping out the entire human genome. What people are thinking about now is to try to find a genomic explanation for all the common diseases, like Alzheimer, cardiovascular disease, and so on. But it is turning out to be terribly difficult. It is not as simple as people think. If you look at the number of Mendelian-based diseases that people have discovered, it has gone up tremendously because of the acceleration of the molecular biology techniques. But the number of polygenic diseases for which one can actually find all the responsible genes has not increased significantly, because there are complex interactions between the human genome and also the environment. Using only this type of bottom up approach to understand systems biology is very difficult.People find other ways to use the genome. They start looking at pharmacogenomics, trying to predict the response of a patient to different drugs, both in terms of side effects and also disease response. There are some early successes like the CYP45 chip. Another success story is that you can now use Oncotype to look for whether the patient should be put on therapy or not after breast cancer surgery.Pathologists have actually led the way in terms of adopting genomic medicine into their practices to do personalized medicine. Where is medical imaging in this process? Are we moving ahead and how fast are we moving ahead? What is the paradigm that we are thinking about?If you look at the beginning of medical imaging, we were much slower than pathology. Pathologists started with eyeball observations. In gross pathology you touch the specimen and then you use your eyeballs to observe. You are hundreds of years ahead of us. Then came the microscope. That was a couple of hundred years before we actually had the first in vivo imaging, which is x-ray, that was discovered slightly more than 100 years ago. At that time, people saw the potential already.