Cancer: Harnessing AI for the development of new drugs
An international research group, coordinated by the University of Lucerne, has developed a novel method for localising and treating various forms of cancer. The new approach integrates artificial and human intelligence.
Theranostic drugs represent an innovative approach to cancer therapy. They enable the localisation of tumours, the prediction of the course of disease, its treatment and the monitoring of treatment success. Examples include radiotracers such as radioiodine for thyroid cancer, DOTATATE for neuroendocrine tumours and PSMA ligands for prostate cancer. These radiotracers are typically administered intravenously and are specifically absorbed by tumour cells. Once inside, they decay and emit radiation, which can be used to both visualise and treat tumours with high precision.
The "theranostic genome"
The field of theranostics is currently advancing at a rapid pace; however, significant gaps remain in our understanding of the available radiotracers with theranostic potential and the cancer cells that can be effectively targeted by these tracers. The study "The Theranostic Genome", recently published in the scientific journal Nature Communications and selected as an "Editors’ Highlight", addresses these knowledge gaps.
The research group has been the first to map the part of the human genome that is targetable with theranostic drugs, opening new avenues for the development of new cancer therapies. Professor Martin A. Walter explains: "The theranostic genome represents a specific part of the human genome that is crucial for both cancer diagnosis and treatment." Walter, who led the research group, is a titular professor of medical sciences at the University of Lucerne and a specialist in nuclear medicine at the Hirslanden Klinik St. Anna in Lucerne.
The theranostic genome comprises a total of 257 genes, which account for 0.4 per cent of the human genome. It includes 13 key "gene families" that play a vital role in the development of various types of tumours.
Combining human and artificial intelligence
The aim of the study was to identify potential theranostic agents for tumour treatment through an in-depth analysis of information from the human genome. To achieve this, the research group developed a novel method that uses artificial intelligence (AI) to connect various research databases, including PubMed, Gene Expression Omnibus, DisGeNET, The Cancer Genome Atlas, and the NIH Molecular Imaging and Contrast Agent Database. In total, data from over 17,000 human tissues were analysed for their affinity to theranostic drugs.
Dr Pablo Jané from the University Hospital of Geneva explains: "The targeted use of artificial intelligence allowed us to extract vast amounts of research data and compare them in the context of our research question. Human intelligence then came into play when interpreting and drawing insights from the extracted data."
The Hirslanden Klinik St. Anna, which has a long-standing partnership with the University of Lucerne, played a crucial role in conducting the research. The partnership focuses on creating optimal conditions for teaching, and for developing and implementing innovative, clinically relevant research projects. For the "Theranostic Genome" study, the clinic provided the necessary infrastructure for its implementation, among other contributions.
Diverse perspectives for cancer treatment
The study revealed that the theranostic genome is significant for a wide range of cancers. Based on this discovery, intelligent, personalised cancer therapies tailored to the unique genome of each patient could become a reality in the future. Xiaoying Xu, PhD, lead researcher at the University of Lucerne, explains: "The theranostic genome aims to tackle existing challenges in cancer treatment by more accurately identifying subpopulations of cancer, which can then be treated with customised theranostic drugs." As an example, the research group's publication demonstrates how different stages of prostate cancer can be effectively targeted using specific radiotracers.
Overall, the research group hopes that the theranostic genome will accelerate the development of targeted theranostics by reducing development cycles and costs, lowering attrition rates in drug development, and facilitating the repurposing of existing radiotracers for therapeutic applications. Beyond cancer treatment, other potential applications may include the treatment of systemic inflammatory diseases like rheumatoid arthritis.
Xiaoying Xu, Pablo Jané, Vincent Taelman, Eduardo Jané, Rebecca A. Dumont, Yonathan Garama, Francisco Kim, María del Val Gomez, Karim Gariani, Martin A. Walter
The Theranostic Genome
Nature Communications, 30 December 2024
Open Access retrieval
Contact:
Martin A. Walter, Head of the Research Group / Titular Professor of Medical Sciences at the University of Lucerne and Specialist in Nuclear Medicine at the Hirslanden Klinik St. Anna, Lucerne; martin.walter @ unilu.ch