Wang and others have collaborated to develop mathematical models that can predict the effectiveness of immunotherapy

0:00 a.m. Beijing time on January 5, 2021, nature-biomedical engineering published online an academic paper by Dr. Wang Andohui and Dr. Vittorio Cristini of the Methodist Institute in Houston, USA, in collaboration with scientists at the MD Anderson Cancer Center at the University of Texas, entitled "A mechanistic immunotherapy model of patient-specific quantification of immune and associated and long-term tum."
the study, researchers described how mathematical models were used to predict the effectiveness of immunotherapy in cancer patients, and explored differences in treatment effects based on drug types and cancer types.is a new generation of cancer treatments that identify and attack cancer cells by activating the patient's immune system. Compared with other therapies, this approach has the advantages of higher target killing rate and fewer side effects, which is a major advance in our research in the fight against cancer. Currently, however, this method works only for certain types of cancer and only in some patients with these types of cancer.
therefore, it is difficult for clinicians to diagnose and determine the most appropriate specific drug (or combination of drugs) for different patients, and to identify the best complementary treatments to improve the patient's immune system. We must explore and develop new ways to improve the effectiveness of immunotherapy from different perspectives, so as to provide more patients with more optimized treatment results.
The team, led by Dr. Wang and Dr. Vittorio Cristini of the Methodist Institute in Houston, worked with researchers at the University of Texas MD Anderson Cancer Center to develop a machine-based mathematical model of checkpoint inhibitor immunotherapy that can be used to predict and quantify the response of specific cancers to specific immunotherapy therapies, thereby quantifying the success rate of treatment after using a specific combination of cancer-immunotherapy drugs for individual patients.
unlike traditional, most statistical data models, which use mathematical equations based on known laws of physics and chemistry to describe complex biological systems involved in immunotherapy and related immune responses. Through model analysis, they explained why immunologic drugs vary greatly in the effectiveness of treatments for different types of cancer.
, we can't quantify certain important parameters in the patient's body related to cancer treatment (e.g., the number of immune cells, or the concentration of drugs that penetrate the tumor, etc.). This limitation is broken by mathematically linking these non-quantifiable parameters with those that can be quantified during treatment in different linear and nonlinear combinations.
, the input data for this mathematical model is the data that the patient has measured during normal treatment and does not require additional supporting data. Therefore, this model provides clinicians with a tool that can be used immediately to predict the effectiveness of treatment. In addition, this model can provide individual prediction of the immunotherapy effects of each patient, which is an exciting and innovative step towards the future of individualized medicine, and provides a new way of thinking and framework for helping physicians develop individualized treatment strategies through engineering methods.to test the model's accuracy in predicting the efficacy of immunotherapy, they first obtained data on 124 patients in four internal clinical trials of checkpoint inhibitor immunotherapy. CT and MRI scan data are available for each patient (before, after immunotherapy, treatment and later). Through the analysis, they found two important model parameters: (1) the health of the immune system in the tumor, and (2) the final killer rate of immune cells activated by immunotherapy to cancer cells. These two parameters can be combined in some form to become a parameter highly relevant to the size of the time series tumor, which also provides a unique quantitative scoring mechanism for the strength of the cancer response to the drug.
, the researchers successfully further validated their findings in 177 cancer patients in another group of checkpoint inhibitor immunotherapy (anti-CTLA4 or anti-PD1/PDL1 monotherapy). These mathematical parameters obtained from the mathematical mathematical model can be used as mathematical markers (mathematical marker, which can be understood as another type of new form of biomarker), and these mathematical markers cannot be obtained through traditional statistical, data mining, or machine learning methods. The results show that these mathematical markers can be widely applied to many combinations of cancer and immunotherapy.
research is highly innovative and practical in the field of integrated cancer immunotherapy. The researchers hope the model could eventually become an auxiliary diagnostic tool for clinicians, providing them with additional personalized predictive information about patients. This information can be combined with current standard diagnostic and predictive data to increase the success rate of immunotherapy for specific patients. (Source: Science Network)
related paper information: