Clinical Validation of Transrectal Multiparametric Ultrasound (PCaVision) for Detecting Clinically Significant Prostate Cancer: A European Head-to-Head Comparison With MRI in Primary Diagnosis and Active Surveillance Populations

Rationale The incidence of prostate cancer (PCa) has increased over the years. To diagnose PCa, histopathological confirmation is required. Different diagnostic pathways are currently available. Systematic biopsies have long been the cornerstone in the diagnostic work-up of men suspected of prostate cancer. However, systematic biopsies can lead to under-diagnosis of clinically significant prostate cancer (csPCa), and each biopsy is associated with the risk of infection and other side effects.

In the magnetic resonance imaging (MRI) pathway, targeted biopsies are only performed when suspicious lesions are detected on MRI. The MRI pathway purposely detects fewer clinically insignificant prostate cancers (ciPCa), but has an increased sensitivity for csPCa and improved localization accuracy of suspicious regions. The MRI-based strategy is now recommended as the first-line investigation. However, reported sensitivities and specificities for MRI vary widely between studies, which can be attributed to differences in MRI equipment, study design, reference standard quality, and inter-observer variability. Moreover, MRI has limited availability and is a time-consuming and expensive imaging modality.

Transrectal ultrasound (TRUS), which is widely available, more cost-effective, and familiar to urologists, may offer a valid alternative. In an ultrasound-based diagnostic pathway, 3D contrast-enhanced ultrasound (CEUS) combined with contrast ultrasound dispersion imaging (CUDI) focuses on detecting angiogenetic changes in the microvascular architecture to localize lesions suspicious for PCa, followed by targeted biopsies for histological confirmation.

PCaVision is a software package designed to support the diagnosis of csPCa by integrating 3D B-mode, 3D Shear Wave Elastography (SWE), and 4D CEUS scans. The PCaVision algorithm was trained on a cohort of 252 patients using prostatectomy pathology as the reference standard, and 83 "negative" patients (no suspicious MRI lesions or positive prostate biopsies). Internal validation showed a sensitivity and specificity of 0.82 and 0.82, respectively. These results led to a first prospective clinical investigation in the Netherlands to demonstrate non-inferiority compared to the MRI-based pathway (NCT06281769). That initial prospective trial aimed to compare the two diagnostic pathways in biopsy-naïve patients under tightly controlled conditions (e.g., 3T MRI, transperineal biopsies, and cognitive or fusion targeting using MIM software). These conditions, however, are not generalizable across Europe, where 1.5T and 3T MRI are used interchangeably, both transperineal and transrectal biopsy approaches are employed, and various fusion systems are in use. Additionally, the previous study excluded patients in active surveillance (AS) and those with prior negative biopsies, who represent a substantial portion of the demand for PCa diagnosis and biopsy guidance.

The objective of the European head-to-head trial described in this protocol is to compare the diagnostic accuracy of two different imaging pathways for detecting csPCa in a broader, more generalizable European setting: (1) the PCaVision-targeted biopsy pathway and (2) the MRI-targeted biopsy pathway. The trial aims to demonstrate the non-inferiority of the PCaVision pathway compared to the MRI pathway in two cohorts: (1) biopsy-naïve and prior negative patients and (2) patients undergoing active surveillance. A fully paired design will be used. The updated PCaVision version 1.1 will be employed in this study, incorporating enhancements to reduce unusable scans and improve diagnostic accuracy and user experience.