Artificial intelligence is establishing itself as a key tool in precision medicine and is already enabling progress towards safer and more personalised cancer treatments. A team of ten researchers from the University of Alicante (UA), the Miguel Hernández University of Elche (UMH), and the Alicante Institute for Health and Biomedical Research (Isabial), led by researcher María José Prieto Castelló, is working on a pioneering project that aims to adapt chemotherapy doses to each patient’s genetic profile in order to reduce toxicity through the use of AI.
Approximately 25% of cancer patients receive chemotherapy. The exact percentage depends primarily on whether the treatment approach is curative or palliative and the type of tumour. For solid tumours, surgery remains the primary treatment.
The use of big data and artificial intelligence techniques is one of the project’s cornerstones. The UA research group led by Juan Carlos Trujillo Mondéjar will develop mathematical models and predictive algorithms capable of analysing large volumes of clinical and genetic data to determine the optimal drug dosage based on the patient’s individual characteristics. This approach will improve clinical decision-making and advance towards a more precise and efficient data-driven medicine.
Along with Castelló and Trujillo, the team is completed by Paula Castro, Andrés Corno, Irene García, Dolores Marhuenda, Manuel Marco, Alejandro Mate, Pedro Pernías and Alexander Sánchez.
Consultations and hospitalisations due to treatment effects cost 500 euro per patient
The research, which has received 5,000 euro from the network projects in which the UA, UMH and Isabial collaborate, focuses on colon and breast cancer and is aimed at identifying genetic markers (patterns) that allow for personalised adjustment of the dose of certain cancer drugs, thus reducing the side effects associated with the treatment.
The study focuses on capecitabine, a drug widely used in the treatment of solid tumours such as colorectal cancer and breast cancer, which exhibits high variability in patient response. Approximately 25.5% of those receiving this treatment develop severe toxicity, and around 1.6% may die as a result of adverse effects.
Key protein
The research falls within the field of clinical pharmacogenetics, which studies how genetic variations influence drug response. Specifically, the team will analyse variants in the gene that codes for the enzyme cytidine deaminase (CDA), a key protein in the activation of capecitabine within the body. Alterations in this enzyme can modify how the drug is metabolised, increasing the risk of toxicity.
These genetic variants are also very common, as nearly 60% of the European population has one of them.
The project is being developed through an observational, multicentre and prospective study with patients from different hospitals in the province of Alicante, such as those in Elda, General de Elche, Sant Joan d’Alacant or Alcoy. Genetic variants, enzyme activity, and the appearance of adverse effects during the first cycles of treatment are analysed, with the aim of establishing relationships that allow for the prediction of drug toxicity.
Dosage guide
Based on this data, the researchers will develop a personalised dosing clinical guide and a bioinformatics tool that will generate automatic reports to facilitate the work of healthcare professionals. In this way, the doctor will be able to have a recommendation based on scientific evidence without having to manually interpret the genetic information, which will contribute to improving the safety and effectiveness of the treatment.
The trial focuses on identifying genetic patterns in breast and colon cancer, and personalising therapies. The impact of the project is not limited to the clinical field, but also has an important economic aspect.
Currently, the adverse effects associated with capecitabine generate an average cost of about 500 euro per patient due to consultations, hospitalisations and related treatments. According to the research team’s estimates, implementing pharmacogenetic screening could reduce these costs by up to 50%, representing an approximate saving of €20,000 per year per hospital. Extrapolated to the entire healthcare system, this saving could reach €5 million per year, contributing to greater sustainability of the national health system.
Four years
The project is planned to last four years and is already in an advanced stage, with over one hundred patients recruited and analysed. At this stage, the research does not involve any additional interventions for the participants, as it is based on samples obtained through routine clinical practice and patient follow-up.
Although the current objective is to generate scientific evidence on the role of these genetic variants as biomarkers, the results could lay the foundation for future clinical trials in which dose adjustment based on the genetic profile is applied directly in clinical practice.
In this regard, the researchers emphasise that this work represents a decisive step towards a more preventive, personalised and patient-centred healthcare model, in which the combination of genetics and artificial intelligence allows for safer, more effective treatments tailored to each individual.
Castelló, who is a researcher at UMH and Isabial, concludes by summarising that “pharmacogenetics and artificial intelligence are advancing towards safer and more effective cancer treatments.”
