A virtual patient will “take” the medicine

A virtual patient will “take” the medicine

By: Marit Slavin

An Israeli company is developing a computerized model to test the efficiency of drugs on a variety of patients.

One of the results of the technological revolution in the era following the decoding of the human genome is a new approach affecting everything concerning the development of medicines, which will be based on a better understanding of the working of the human body and the changes caused by the progression of the disease. For instance, in case of cancer, the chemotherapy medications used are based on the conservative trial and error method, which tests if and in what concentrations different substances affect the development of the illness. However different tumors react in different ways to medication and in addition the reaction to the same medication may differ from patient to patient.

The formation of a cancerous growth in a patient is a complicated process. The process progresses from genetic mutations in a single cell to a tumor developing within a tissue and this then affects all the body’s systems. Each stage is in itself complex, but what really complicates the whole issue is the multifarious interactions of the different processes which include, among other things, the specific personal reaction of each particular patient.

How can we make sense of this enormous body of data and, from all this complexity, find a connection between the illness, the physiology of the patient and its impact on the disease on the one hand, and the different treatment options on the other that will help us to choose the best treatment for the patient?

In the last few years a new approach has emerged towards the selection of the best treatment for the patient. It translates the data into mathematics and programming language, thus creating a virtual patient which can be used for testing of various virtual medications. The computer checks all the data and determines which treatment best suits the given conditions, the dosage of the medication to be given and the chances for success in the case of a particular patient.

One of the leading experts in this new field is the bio-mathematician, Prof. Zvia Agur. In 1999, she founded the Institute for Medical Bio-Mathematics and a commercial company Optimata Ltd.. The goal was to develop methods and tools for the prediction of the best treatment for serious illnesses. “Today most of the pharmaceutical developments are based on the trial and error method. We believe that if we succeed in building a computer model that describes the physiological, pharmacological and pathological characteristics of the relationship between the medication and the patient, then we will be able to accurately predict the short- term and long-term effects of a specific treatment regime. Thus we will be able to replace trial and error with a more accurate method that will enable the administration of treatments that are less toxic and more effective”, says Agur.

How does Agur’s method work? Let’s take for example a model that examines the effect of an anti-cancer treatment. The first stage is to create a virtual patient that simulates all the processes associated with cancer. All the data about the tumor that were examined are put into the virtual patient: the process of its growth, the creation of blood vessels, etc. In addition, the model simulates the most important physiological processes that characterize the tumor. “We study all known biological information about a particular disease, and when we understand the biological systems concerned, we translate the knowledge into mathematical equations that are converted into computer language. The advantage of bio-mathematics is the ability to extract from the complex data the essence that determines the dynamics of the system and to describe it accurately and briefly. Our model describes the patient using hundreds of mathematical equations and we work on the assumption that most biological processes are similar for all human beings, even though they may progress at different rates in each individual”, explains Agur.

“We can describe a universal cancer patient. However, we also have the technological ability to put into our model the specific data of each patient which influence the rate of development of the disease, such as the size of the tumor, the hematological status and even the patient's genetic composition. We have the ability to apply all this information to our virtual patient and to test the effect of the treatment on him personally”, says Agur.

The next stage is to build a model that simulates the medication and its administration. The computer tests how the medication disseminates throughout in the body, which cells react with it and how they do so. The new data are also translated into mathematical equations and converted into computer language.

Now, when Optimata scientists have data regarding both the patient's biological system and the behavior of the medication, they go one step further and investigate the impact of the various treatment protocols – such as dosage, administration timing and frequency, administration method (by injection or tablet) – on the virtual patient model. The company has developed a technology that combines data regarding the virtual patient as well as the medication and simulates an infinite number of potential treatment protocols. This technology can predict the medication's efficacy/toxicity trade-offs and its effect on various patient populations, right down to an individual patient. Lastly the system calculates the preferred treatment method.

An example of this is an investigation into the effects of Thrombopoietin, which was being developed to prevent thrombocytopenia – a condition in which the patient suffers from life-threatening hemorrhages caused by, amongst other things, chemotherapy treatment. However, during the drug's development, it was discovered that Thrombopoietin promotes the production of antibodies in the patient's circulation that can lead to a serious deterioration of his condition. Development of this drug was therefore discontinued.

Optimata researchers created a virtual patient "suffering" from thrombocytopenia and "treated" the patient with Thrombopoietin. This simulation made it possible to predict the number of platelets in the patient's blood on a daily basis. During and after treatment, laboratory tests indicated that these predictions were extremely accurate. In the next stage, using technology it developed, the company proposed an alternative treatment regimen that required half the dosage of the medication and produced the desired outcome without its serious side-effects.

In order to validate these predictions, Optimata cooperated with Prof. Gerard Wachmacher of Rotterdam University, a world expert on this medication, who carried out controlled experiments on mice and monkeys. These experiments proved the efficacy of the treatment regimen proposed by Optimata scientists and, most importantly, proved that when Thrombopoietin is applied according to the prediction obtained from the virtual patient, it did not lead to a toxic immune reaction in the patient.

The results, published in the prestigious "British Journal of Haematology", provide first-of-its-kind proof of the accuracy and efficacy of biomathematics-based patient models. These results have paved the way for various pharmaceutical companies to continue developing the medication and thereby to significantly improve the condition of thrombocytopenia patients.