Big Data medicine identifies appropriate chemotherapy for cancer

Computer programs combine data to determine the best possible therapy for patients
In the future, medicine could benefit immensely from computer programs that link information about patients with active ingredients. In this way the best possible therapy should be determined. Because the success of a treatment depends on many individual factors. For example, a particular chemotherapy may work well in one patient but not in another. The solution should deliver "big data".

Find the optimal active ingredient with Big Data
The buzzword "big data" causes many people fear. Because the ability to collect, store and link huge amounts of data entails many risks from a data protection point of view. In the field of medicine, special computer programs that link patient data to current research, therapies, and related case histories can add value by identifying the optimal treatment for the patient. But privacy advocates warn that such data could be used, for example, by employers and insurance to the detriment of those affected. In view of the rapid progress made in the area of ​​"Big Data", the Ethics Council convened an annual meeting under the motto "Big Data in Medicine" at the Hasso Plattner Institute (HPI) for Software Systems Engineering in Potsdam. Over the past two days, medical and IT experts discussed the progress and drawbacks of the new opportunities.

Big Data Against Cancer. Image: vege - fotolia

The "Drug Response Analyzer", which is already in use at the Charité Medical Clinic for Hematology, Oncology and Tumor Immunology, was also introduced. The program aims to help physicians select the most appropriate chemotherapy for tumors of the mouth, pharynx, larynx, nose and throat within a few minutes. The system was developed by the mathematician, computer scientist and HPI director Christoph Meinel and the head of the HPI program "E-Health", Matthieu Schapranow. The program considers substances, combinations of cytotoxins and also antibody therapies that specifically intervene in cell signaling pathways.

"Big Data" can help the doctor significantly to find the optimal therapy
So far, oncologists must, among other things on the basis of the patient's history and genome characteristics of the tumor as well as taking into account international study results and treatment plans from guidelines partially Sisyphus work to find the optimal treatment. With the "Drug Response Analyzer", the HPI researchers want to significantly accelerate this process. Among other things, the "high-speed database" links data on genes and proteins of the tumor, on signaling pathways of the cells as well as on its reaction to the various active substances. In addition, all available scientific publications on the subject are continuously added to the database.

"Cancer researchers will also be able to identify relationships between variants in the genetic makeup of patients and the effect of drugs on them," Meinel explained. "We combine data from historical cases to help select the right treatment for acute patients," adds Shapranov. "A mathematical model evaluates the data from different patients and thus allows a prognosis of the individual response of a single patient to a specific drug in real time."

Although only data from about 40 patients with head and neck cancer are stored in the system, they are comprehensive. Only the population-based studies provide so far large data collections. "Epidemiology always delivers big data," explains the physician and public health expert Reinhard Busse from the Technical University of Berlin to the "Tagesspiegel". He sees new technologies as an opportunity to network and use them even more effectively. (Ag)