Clinical Feasibility of MRI-guided In-Bore Prostate Biopsies at 0.55T

Abstract

Objective In-bore MRI-guided biopsy allows direct visualization of suspicious lesions, biopsy needles, and trajectories, allowing accurate sampling when MRI-ultrasound fusion biopsy is not feasible. However, its use has been limited. Wide-bore, lower-field, and lower-cost scanners could help address these issues, but their feasibility for prostate biopsy is unknown. The purpose of our study was to evaluate the feasibility of in-bore MRI-guided prostate biopsy using a large-bore (80cm), 0.55T scanner. Materials and Methods Nineteen participants (68 ± 10 years) with suspected prostate cancer (PCa) were recruited for this Institutional Review Board (IRB) approved study (May 2023 -October 2024). Prebiopsy diagnostic scans and intra-procedural T2-weighted images were used for lesion localization. PSA levels, lesion sizes, cancer detection rates, positive core volume percentage, ISUP (International Society of Urological Pathology) grade groups (GG), positive volume cores, skin to target distances, and procedure durations were reported. Results Seventeen participants underwent biopsies (four transrectal, thirteen percutaneous). Two participants were excluded. Twenty lesions (mean size 1.95 ± 1.29 cm) were biopsied which showed various GG cancers (GG1, GG2, GG3, GG4, and GG5), with positive cores ranging from 10%-100%. 20% of the lesions were benign. Compared with the previous biopsy results, 11.7% of participants had a GG upgrade, 17.6% had an upgrade in positive core volume, 17.6% had negative biopsies and 47% of biopsy-naïve participants had new cancer detections. No upgrade was observed in 5.8% cases. One new cancer was detected near a hip prosthesis due to reduced imaging artifacts. Average total procedure time was 77 ± 21 minutes for transrectal and 74 ± 22 minutes for percutaneous biopsies, with times to first core at 45 ± 15 and 53 ± 14 minutes, respectively. Conclusion Identifying and accurately targeting suspicious prostate lesions is feasible using a 0.55T MRI scanner.