Mathematical models for the digital transition in neuro-oncology: In-silico design of a clinical trial for glioblastoma.
This project aims to digitize essential aspects of the work chain in Neuro-Oncology. We intend to generate advances in knowledge and data management and exploitation to develop mathematical model-based digital technologies. We will use disruptive digital technologies based on mathematical models and digital patient twins to design a clinical trial proposal for the most frequent and lethal type of brain tumor, glioblastoma (GBM). We also plan to release a public database of GBM images and patient data of unprecedented size. Our goal is to enable better and personalized treatments for GBM patients using digital tools grounded on mathematical methods informed by data. Biomedical product development and assessment are based on long and expensive experimental processes that push development costs and times to unsustainable levels, thus stifling innovation and leading to the constantly growing cost of health-care provision. The failure rate of clinical trials under our current system is too high, and to increase their success rate we need to smarten up trial design. GBMs have a dismal prognosis (15 months overall survival) even when receiving maximal surgical safe resection followed by RT plus concomitant and adjuvant chemotherapy with temozolomide (TMZ). The number of TMZ cycles and the interval between them was selected based on results with other molecules in different cancers and there is potential for improvements using in silico approaches. Also, the digital transition in Artificial Intelligence (AI) methods in Neuro-Oncology requires large databases, currently unavailable, to be properly trained. The project is organized around three objectives. First, we will develop and fit using human data, mathematical models describing the response of GBMs to TMZ and RT. Models will be experimentally validated using animal models. The second objective is the use of the mathematical models to find the best combination scheme for chemoradiotherapy of GBMs in-silico. A clinical trial protocol based on the project ideas, i. e. an in-silico designed trial, will be ready for clinical implementation at the end of the project. The third objective is building and publishing a GBM database for mathematical and AI studies allowing for the digital transition in Neuro-oncology based on the approved public use of the applicants GBM imaging datasets. Our team includes applied mathematicians from the Mathematical Oncology Laboratory (, biologists from Instituto de Salud Carlos III and Hospital 12 de Octubre and medical doctors with backgrounds on medical oncology, radiation oncology, radiology, and neurosurgery from six hospitals. A collaborative approach is the only way to reach our goal of designing for the first time a clinical trial in silico using mathematical models, enabling the digital transition in neuro-oncology. Exploring different treatment regimes in-silico would allow for shorter development times and a more efficient use of the treatments. The change of paradigm that in-silico trial design provides could lead to improving the effectiveness of drug and medical procedures leading to health benefits and economic savings for patients and public health systems. The publication of the project database will facilitate the use of AI for automatic tumor location and delineation, image-based differential diagnosis, non-invasive tumor grading and prognosis/response assessment, etc.
Funding entity
Ministerio de Ciencia e Innovación (NextGenerationEU) (2023-2024)
Univ Castilla-La Mancha, Univ. Cadiz, Univ. Guadalajara (Mexico), H. 12 de Octubre, Instituto de Salud Carlos III, H Marqués de Valdecilla., H. de Salamanca, H. Carlos haya, Inst. Valenciano Oncología
Principal investigator
2 years