Jesús García-Donas, head of the Gynecological and Genitourinary Tumors Unit of HM CIOCC Madrid
An Artificial Intelligence (AI) study applied to ovarian cancer makes it possible to advance in the knowledge of predictive and prognostic molecular biomarkers of this type of tumor, which will contribute to a better understanding of the evolution of the disease and, therefore, to define diagnoses personalized, more precise and efficient. The Clara Campal Comprehensive Oncology Center (HM CIOOC), in collaboration with the Juan Carlos I University, has designed and developed this important collaborative work.
Until now, certain genetic alterations had been determined as predictive and prognostic biomarkers in ovarian cancer, such as inactivating mutations in the BRCA1 and BRCA2 genes, but they are insufficient to understand the global evolution of the disease. Thus,
HM CIOCC and the Juan Carlos I University have collaborated to carry out this multicenter observational research focused on the identification of biomarkers with potential impact on clinical practice.
Jesús García-Donas, head of the Gynecological and Genitourinary Tumors Unit at HM CIOCC Madrid and co-author of the study, explains that “currently, they know that cancer is a complex disease, in whose evolution not only genetic alterations are important, but also The conditions of the microenvironment, the regulation of gene expression and, of course, the conditions of the person who suffers from it also have an influence.This is forcing them to change the previous approach, somewhat simplistic, in which a tumor was conceived as a mere sum of mutations and it is essential to integrate multiple data in order to understand the neoplasm and predict its evolution”.
More than 300 patients
To carry out this study, the clinical and genomic data of 300 patients with advanced ovarian cancer have been entered with the aim of establishing a relationship between them and the variables that determine the progression of the disease, which represents “a novel line of work in which they are fine-tuning artificial intelligence algorithms capable of integrating genomic data with the clinical and pathological characteristics of the disease in order to address it in a complex and comprehensive way. We think that this approach could give us a closer view of reality than the classic approaches, focused on specific punctual alterations”, indicates García-Donas.
In this sense, the AI algorithms identified patterns common to those cases that responded well to treatment versus those that were resistant. If confirmed in independent cohorts, we would be facing a new and promising line of work in which our accuracy when predicting the evolution of a case could increase exponentially.
Regarding the predictive and prognostic role of certain variables, the results confirm that subjecting patients with a high tumor burden at the time of diagnosis to neoadjuvant treatment followed by maximum effort surgery favors the reduction of said burden. . The association between these variables leads to longer survival in accordance with previous data in the literature.
Several previous studies have shown the relationship between certain genetic alterations (BRCA1/2 and RAD51C) and the evolution of the disease. The application of Artificial Intelligence has made it possible to define other gene correlations that must be confirmed by increasing the number of data, that is, of cases to be sequenced. García-Donas considers that “this work is pioneering in ovarian cancer and will require a coordinated effort between medical oncologists, basic researchers and computer engineers to achieve a real impact on patient management.”
Agreement with Microsoft
The participation of HM Hospitales in this study is an example of the clear commitment that, since the beginning of the year, the Group has been making for the application of Artificial Intelligence to cancer research. As a result of the agreement signed with Microsoft, work is being done on the analysis of diagnostic images and on the automation of data collection from clinical histories with a double objective: to guide clinical activity and support research, accelerating the development of new procedures and medications.
For García-Donas, “working in the field of artificial intelligence applied to health is always a difficult challenge”. “Currently, once we have established the best study methodology, we want to advance in the prediction of response to specific therapies so that we can personalize the treatment of patients using artificial intelligence.” In this way, he is confident that “when If we are able to understand advanced-stage disease, we will be able to apply the same methodology to early disease where the chances of cure are much greater.It is probably in this context that the impact of this technology could decisively change the course of the disease”, he ends.
Although it may contain statements, data or notes from health institutions or professionals, the information contained in Medical Writing is edited and prepared by journalists. We recommend the reader that any questions related to health be consulted with a health professional.