A prediction model based on MRI developed in Barcelona to advance early detection in prostate cancer

Prostate cancer (PCa) is a heterogeneous disease with a wide variability, therefore, it is essential to provide a personalized approach for early detection, disease status (indolent or aggressive) and prediction of treatment response. Clinically significant prostate cancer (csPCa) refers to tumors that are likely to grow and spread, potentially leading to adverse outcomes such as metastasis or mortality if left untreated.
Researchers from Vall d´Hebron Research Institute (VHIR, Barcelona) developed a Magnetic Resonance Imaging Prostate Model (MRI-PM) by building upon earlier advancements in prostate cancer detection. This marked a significant step toward enhancing the accuracy of prostate cancer diagnosis. Despite prior analysis in PCa populations, MRI-PMs have not been evaluated by the Prostate Imaging-Reporting and Data System (PI-RADS) categories, leaving their clinical utility uncertain. Radiologists use the Prostate Imaging Reporting and Data System (PI-RADS) to report how likely it is that a suspicious area is a clinically significant cancer. PI-RADS scores range from 1 (most likely not cancer) to 5 (very suspicious).

This development offered a refined approach to select candidates for biopsy, harnessing insights gained from previous discoveries. Through a comprehensive cohort study involving over 2,400 men from Barcelona, Spain, this MRI-PM demonstrated promising results in detecting csPCa, showcasing continued progress in the field of prostate cancer screening.

Understanding the Barcelona Model

A new MRI-PM for csPCa was developed to improve and individualize the selection of candidates for prostate biopsy in the metropolitan area of Barcelona, Spain. The developed MRI-PM was able to detect 95% of existing csPCa, avoiding 40% of prostate biopsies in our overall population.
The development and validation of the Barcelona MRI-PM has yielded remarkable findings, suggesting a significant breakthrough in prostate cancer diagnosis. The MRI-PM demonstrated its effectiveness in detecting csPCa in a study conducted in the metropolitan area of Barcelona. The detection rate of csPCa saw a notable increase from 36.9% to 40.8% between two cohorts, indicating a level of robustness of the MRI-PM across diverse patient populations. The findings also underscore the potential clinical utility of the Barcelona MRI-PM across different risk categories. This tool further enhances its applicability in clinical practice, facilitating external validation and optimization of MRI-PMs based on specific risk profiles.
The Barcelona MRI-PM presented good performance on the overall population; however, its clinical usefulness varied regarding the PI-RADS category. The analysis regarding PI-RADS categories shows that the developed MRI-PM outperforms in PI-RADS < 3. That is why novel and better models are needed in order to improve the detection of csPCa while avoiding unnecessary biopsies.

Purpose of FLUTE Project in improving Prostate Cancer diagnosis.

In the era of artificial intelligence (AI) and big data, federated Learning (FL) has emerged as a transformative approach, allowing collaborative learning across distributed datasets while preserving privacy, and in this context Flute project will aim to build a better model based on the data obtained in the Barcelona predictive model. These models advance a new phase in PCa diagnosis, offering a tailored approach that considers individual risk profiles. By leveraging imaging techniques and innovative predictive models, the Barcelona MRI-PM presents a promising solution to the challenges associated with traditional biopsy approaches. This model is now under validation and refinement within the FLUTE project framework, to increase robustness across Europe. This novel approach holds the promise of improving early detection rates and reducing unnecessary interventions in the fight against prostate cancer.

Olga Méndez, Berta Miró, Juan Morote

Fundacio Hospital Universitari Vall d'Hebron