Project Details
Projekt
Artificial Intelligence based Multiomics for Personalized Outcome Prediction after Radiation Therapy in Primary Prostate Cancer Patients
Applicant
Dr. Simon Spohn
Subject Area
Nuclear Medicine, Radiotherapy, Radiobiology
Term
since 2024
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 546330039
Prostate cancer (PCa) is the most common malignancy in men in Europe. Advances in diagnostics, such as multiparametric magnet resonance tomography (mpMRI) and targeted biopsies, combined with prostate-specific antigen screening and longer life expectancies, have increased the detection of high-risk PCa requiring radical treatments. This trend poses financial and personnel challenges to healthcare systems and demands optimal treatment strategies. Current treatment decisions are based on clinicopathological features like histopathological differentiation, PSA levels, and tumor extension (TNM-classification). Radiotherapy (RT), with or without androgen deprivation therapy, is a curative option. However, up to 20% of patients with intermediate or high-risk PCa experience relapses within five years. The recurrence diagnosis and salvage treatments can lead to impaired quality of life and significant costs. Unfortunatenly, present risk classification systems do not adequately reflect individual tumor biology and aggressivents, which is reflected by hetergoeneous treatment responses. Thus new biomarkers are warrented to better understanding the PCAs' proteogenomic, metabolic, and histopathological characteristics and thus tailor personalized treatments. The AIMPORT project aims to develop robust and applicable multiomic signatures as prognosticators for outcome after definitive RT for patients with intermediate and high-risk PCa using artificial intellenge (AI) approaches. The clinical backbone consits of a randomized controlled phase III multicneter trial on personalized RT based on functional imaging (HypoFocal-SBRT trial), which is funded by the „Decade againts Cancer“. High dimensional data from these patients will be concatenated in different aims: (i) Genomic signatures including known genomic classifiers, (ii) petroemoic signatures (iii) histomic signatures and (iv) radiomic signatures will be validated as prognosticators after definitive RT. Considering in advantage of multi-omic analyses, integrative statistics and AI-based model building will be performed to identify invasive and non-invasive biomarkers for superior risk stratification. These new biomarkers will facilitate identifying patients who are at high or low risk to experience relapse after definitive RT. Through this improved risk stratification, we aim to prevent over- and under-treatment in order to increase quality of life and to reduce the risk of disease progression and accompanying comorbidities. The generated data will contribute to the design of further investigations and clinical trials validating the predictive capacity of biomarkers and thus paving the way to personalized medicine. These goals are of tremendous importance for patients, as defined by the European prostate cancer patient’s advocacy movement (EUROPA-UOMO), which supports the AIMPORT-Project.
DFG Programme
Research Grants
