Multiomics approach for patient stratification and novel target identification in metastatic clear cell renal carcinoma (Meet-URO 31).

Abstract

TPS498 Background: The choice of the best treatment in first line metastatic clear-cell renal cell carcinoma (mccRCC) patients is becoming an issue, since no biomarkers are available to guide the treatment allocation strategy. Recently there has been a great deal of interest in Artificial Intelligence (AI) systems and their ability to process heterogeneous data for both classification and prediction purposes. An additional fields of interest in genitourinary oncology are also liquid biopsy and radiomics. Non-invasive liquid biopsy methods are able to detect and characterize circulating cell-free DNA (cfDNA), extracellular vesicles (EV) associated RNAs and circulating tumor cells (CTCs) and to allow longitudinal evaluation of tumor evolution whereas radiomics may provide a novel approach to develop predictive tools by correlating imaging features to tumor characteristics including histology, tumor grade, genetic patterns and molecular phenotypes, as well as clinical outcomes. We hypothesize that AI can be used to integrate data obtained from radiomics, genomic and transcriptomic analysis of CT scan, neoplastic tissues and circulating cell-free DNA or microvescicle-associated RNA with the purpose of defining an optimal allocation strategy for patients with mccRCC undergoing first-line therapy and identifying novel targets in mccRCC. Methods: This is a multicenter Italian prospective translational study evaluating transriptomic, genomics and radiomics in treatment-naïve advanced ccRCC patients. Subjects will be screened to identify a total of 100 patients eligible for the study, candidate to receive first-line treatment as per investigator’s choice according to clinical practice. Tumoral tissue, plasma samples and radiological exams will be collected at baseline, at 3 months, at the time of first radiological evaluation and at disease progression (PD) to provide a comprehensive molecular profile and radiomic features extrapolation, respectively. AI systems will be used to build a genomic-radiomic profile of patients to correlate to treatment response. The planned sample size of 100 patients will allow an exploratory analysis of the prognostic and predictive performance of the "multiomic" classifier, to be subsequently validated in a larger expansion cohort of patients. Through the above research we are confident to provide proof of concept that AI is able to combine the information from genomics, transcriptomic and radiomics to provide an opportunity for a molecularly driven patients' stratification, aiming to choose the ideal first-line systemic treatment for each patient. Enrollment has already begun and the trial has enrolled 15 of the planned 100 patients. Clinical trial information: NCT05782400 .