- Статьи
- Science and technology
- Watch to monitor: Software will predict survival rates of lung cancer patients
Watch to monitor: Software will predict survival rates of lung cancer patients
Specialists from Sechenov University have created a software package that can predict the survival rate of cancer patients depending on the stage of the disease. The obtained data will help researchers and doctors to compare different scenarios of the course of the disease, choose optimal treatment options and thereby increase the life expectancy of patients. The software can also be used to develop and efficiently implement new oncologic drugs. At this stage, it has been used to build prognoses for patients with one of the most common and deadly types of lung cancer.
Prognosis in non-small cell lung cancer
Experts in mathematical modeling at Sechenov University have used the Oncomonitor software to forecast the development of the disease in patients with non-small cell lung cancer (NSCLC) depending on the stage of the disease and the types of therapy regimens used. It is the most common type of disease and is characterized by a very low survival rate. Only 28% of people with NSCLC live for more than five years. In the future, the platform is planned to be used to build prognostic models for other types of malignant tumors.
By combining survival data from various randomized clinical trials, the developers have created a tool for predicting the overall survival of patients. The new method is based on meta-analysis combined with survival models, which the program can use to make predictions, the university said.
The model takes into account the stage of the disease, the presence or absence of certain mutations or biomarkers, as well as the possibility of applying different types of therapy - from traditional surgery to combination immunotherapy. Current treatment regimens from global and Russian clinical guidelines were also integrated into the model.
This allows to make forecasts in different scenarios of the course of the disease, differing in treatment protocols, the characteristics of the studied cohort of patients and other factors, the developers said. The advantage of the methodology is that it allows to quantitatively assess the effectiveness of different drugs and treatment regimens not only in individual patient cohorts, but also in the general population.
- For example, in a study of the drug Osimertinib, its efficacy in prolonging overall survival compared to standard therapy was proven. However, in the European population, only about 15% of NSCLC patients have the mutation targeted by this drug. Accordingly, the advantage of this drug, which it showed in the clinical trial, will not be so significant in the framework of the analysis of effectiveness for the total cohort of patients," Kirill Zhudenkov, associate professor of the Department of Mechanics and Mathematical Modeling at Sechenov University, told Izvestia.
The introduction of elements of pharmacoeconomic analysis into the methodology allowed us to estimate the number of added years of life for different cohorts of patients depending on different conditions. This opens up opportunities for a more balanced decision in terms of assessing the effectiveness of new technologies in the treatment of non-small cell lung cancer, the developers believe.
Prognostic models in medicine
Models to assess the severity of the disease are an important part of modern medicine. Their creation was spurred by the coronavirus pandemic, Petr Kshnyakin, head of the laboratory of personal medical assistants at the NTI Competence Center "Bionic Engineering in Medicine" on the basis of SamSMU, told Izvestia.
- Now predictive models capable of predicting the severity of diseases are being built in various nosologies, and the presented "Oncomonitor" is a rather vivid representative of such developments. This model is clearly of high relevance. A doctor needs to have a personalized, tailored prognostic model for a particular patient," the expert said.
It is difficult to overestimate such systems, because they combine a large number of factors that are almost impossible for a doctor to assess independently, emphasized Petr Kshnyakin. Such systems make it possible to improve the quality of treatment, the quality of the patient's prognosis and to select the most personalized therapy. The development of such preferred systems is a step towards real personalized medicine.
The data obtained is more like the average temperature in the ward, says Mikhail Bolkov, a senior researcher at the Institute of Immunology and Physiology of the Ural Branch of the Russian Academy of Sciences.
- The developers take the average population and see how this or that drug affects it. It would be good if this program were applied individually. If scientists predicted that such patients with such mutations would be susceptible to certain types of therapy," the specialist told Izvestia.
The description of the methodology and operation of the digital platform "Oncomonitor" is published in the scientific journal Computational and Structural Biotechnology.
