How Big Data Is Redefining Diseases

It’s the end of the world as we know it – and I feel fine: If one is familiar with it, this catchy old R.E.M. tune inevitably comes to mind when reading this year’s ETIM program. “The end of medicine as we know it” was the title of the lecture held there by Prof. Dr. Harald Schmidt, Head of Department of Pharmacology & Personalised Medicine, Maastricht University. However, whether those present were faring well with this prognosis is debatable. Because one thing was made clear during the lecture: Harald Schmidt is not prone to exaggeration.

Away from the Organ

Right at the start of the lecture, he shook the cornerstones of today's medicine: the organ-based classification of diseases which predefines the structure of healthcare institutions nowadays as well as the administration of medicinal products. In cardiology, patients are treated with dedicated drugs for arrhythmias, for example, and in pulmonary medicine, people with asthma receive medicines which have been specifically approved for this disease. “We know that, on average, drugs work only in one out of five to 25 patients. However, the pharmaceutical industry has still not found a way to improve this ratio. This is not surprising, because the entire field of pharmaceutical research is based on the definitions of disease from the 19th and 20th centuries – organ-based and descriptive. It is therefore about time for diseases to be redefined on the basis of 21st century knowledge,” says Prof. Dr. Harald Schmidt.

Medical Data Shows New Patterns of Diseases

And this is where big data is helping. Based on large medical databases, for example from Great Britain or Scandinavia, Harald Schmidt and his team developed networks of all human diseases with a focus on risk genes. “In doing so, we identify patterns of diseases which have nothing to do with the conventional classification by organs. Instead, we see disease clusters of a variety of organs with the same underlying mechanism. This means that, as long as we practice medicine on the basis of organs and individual diseases, we will never understand the mechanisms of diseases well enough to effectively treat and actually cure them.”

Disease Clusters Define the Patient Treatment of Tomorrow

What does this mean specifically for patient care? According to Prof. Dr. Harald Schmidt, this implies a softening of the traditional specialist physician model. In the future, the patient will still seek out a physician based on symptoms. However, the physician’s task will not be to treat the symptoms based on the organ(s), but rather to identify the disease mechanism and the correlation to other diseases with the aid of artificial intelligence – and to treat the disease in an interdisciplinary manner. However, this also means the following: A drug which is used to treat disease A of a certain cluster would also have to be effective in all other diseases of the cluster. “We refer to this approach as drug repurposing. We take approved drugs which are already available and use them in the case of diseases other than those described. This is not a dream of the future; many institutions are already taking this approach. We are, for example, currently preparing a clinical study for the use of drugs for the treatment of diabetes and heart failure in stroke patients,” explains the expert.

Welcome to Network Pharmacology

This approach does not only sound logical but is also easy to achieve. Scientists finally have an entire series of tried and tested active substances and medicines available to them which they only need to release from the confines of organ-specific therapy. Experts such as Harald Schmidt speak of a paradigm change to systems medicine and network pharmacology: Instead of continually developing new drugs according to the same principle, existing medicines are used for diseases of the same network and administered in a combined manner – because the more disrupted components of a network are synergistically corrected, the higher the likelihood of a successful therapy. Harald Schmidt: “New studies show that, for example, all tumor diseases can ultimately be reduced to about 33 signaling pathways – independent of where a tumor occurs. We thus try to correct each of these signaling pathways in a highly effective manner.” For the traditional pharmaceutical company, this is admittedly not good news. Many insiders assume that the previous business model of the research and development company will be largely obsolete in about ten years. Money can then be made only with the production and sales of medicines – and with the medical data which provides new knowledge about the network mechanisms.