Selective Cell Size MRI for the Diagnosis and Grading of Clinically Significant Prostate Cancer: A Feasibility Study

Clinically significant prostate cancer (csPCa) and benign prostate tissue are composed of distinct cell populations. Selective size imaging using filters via diffusion times (SSIFT) MRI, a novel technique generating cell size-weighted maps sensitive to specific cell sizes (e.g., cancer cells), has an unclear clinical value in prostate cancer. This study aimed to evaluate whether SSIFT MRI can serve as a novel imaging biomarker for prostate cancer, specifically for discriminating csPCa from clinically insignificant disease (CIS). A total of 152 men (89 with csPCa and 63 with CIS) suspected of csPCa were prospectively enrolled, with 229 lesions (93 with csPCa and 136 with CIS) delineated. SSIFT MRI data were acquired using pulsed and oscillating gradient diffusion sequences to generate incremental AUC (iAUC). We evaluated the performance of iAUC, the apparent diffusion coefficient (ADC), and prostate-specific antigen density (PSAD) in discriminating csPCa from CIS and assessing Gleason grade group (GGG). Additionally, correlations between iAUC and pathologic parameters (nuclei fraction, diameter, and cellularity) were analyzed in patients with whole histopathologic slides. Higher GGG was associated with increased iAUC values (GGGs 1-5: 3.34% ± 1.28%, 5.31% ± 2.33%, 6.18% ± 2.71%, 6.36% ± 2.23%, and 6.23% ± 1.15%, respectively). For distinguishing csPCa from CIS, iAUC exhibited the highest diagnostic performance, with AUCs of 0.91 (peripheral zone), 0.86 (transitional zone), and 0.89 (whole prostate). In per-patient analysis, the combined model incorporating iAUC, ADC_50ms, and PSAD achieved the highest diagnostic accuracy (86%, 131/152) and AUC (0.93; 95% CI: 0.88-0.97). Histopathological analysis revealed a mean prostate cancer nuclei diameter of 10.98 ± 1.81 μm, within the SSIFT filter range (10-20 μm). These findings suggest that SSIFT MRI shows promise as a novel imaging biomarker for prostate cancer, enabling noninvasive identification of csPCa and optimizing clinical decision-making.