Derek William Cool MD, PhD, Presenter: Patent agreement, Eigen

Xuli Zhang BSc, Abstract Co-Author: Nothing to Disclose

Cesare Romagnoli MD, Abstract Co-Author: Nothing to Disclose

Walter Matthew Romano MD, Abstract Co-Author: Nothing to Disclose

Jonathan Izawa, Abstract Co-Author: Nothing to Disclose

Aaron Fenster PhD, Abstract Co-Author: License agreement, Eigen

Prostate MRI’s high sensitivity to early stage prostate cancer (PCa) permits targeted biopsy as an alternative to the current non-targeted systematic TRUS biopsy. Biopsy needle guidance under TRUS is the most economical option, but spatial correspondence of MRI findings with TRUS is non-trivial. The accuracy of sampling MRI lesions under 2D & 3D TRUS and 3D TRUS fusion is quantified.

Three experts performed simulated biopsies on 12 patients (Pros. Vol=37±11g, PSA=9.0±5.1ng/ml) with a total of 15 Bx-confirmed PCa MRI lesions (0.8±0.8g) who were selected from 90 patients undergoing MRI-3D TRUS fusion biopsy. Two core-biopsies were targeted at each MR lesion using 2D TRUS, 3D TRUS and MRI-3D TRUS fusion for needle guidance, using a validated Bx simulator. Additional single core-biopsies were directed toward common regional targets defined for all patients. Regions included: left-anterior transition zone (TZ), left mid-gland, right apex and right base. The simulated Bx core locations were compared to the original MRI to evaluate the sampling accuracy.

The 15 PCa tumors were distributed as follows: Base=4, Mid-gland=5, Apex=6, with 7 lesions along the anterior prostate. Only 44±4% & 51±14% of tumors were sampled with 2D and 3D TRUS, respectively, compared to 98±4% with fusion biopsy. Bx sampling rates were not significantly different between anterior and posterior tumors for any modality. The Bx sampling errors for the regional targets were significantly higher (p<0.01) for 2D (8.1±4.5mm) and 3D TRUS (8.8±4.7mm), as compared to MR-TRUS fusion (1.5±2.6mm). These errors correspond to respective predicted detection rates of 28%, 21% & 94% if each target were a tumor of 0.5g (smallest clinically significant size) and 67%, 71% & 98% for a 4g tumor. Typically, Bx cores targeted at the base and anterior TZ were over-rotated toward the base, while Apex & Mid-gland cores erred toward the apex. Targeting errors are likely due to marked orientation differences between the axial MRI and the standard axial TRUS Bx images, which ranged from 50o-90o.

Biopsy of prostate MRI lesions under 2D or 3D TRUS without MR-fusion may be inaccurate and lead to a falsely negative biopsy.

Targeting biopsies at suspicious lesions seen on prostate MRI using 2D or 3D TRUS without image fusion is likely not accurate enough to adequately detect many significant prostate cancers.

Cool, D, Zhang, X, Romagnoli, C, Romano, W, Izawa, J, Fenster, A, Simulator-based Comparison of 2D, 3D and Fusion 3D Transrectal Ultrasound (TRUS) Needle Guidance Accuracies for Biopsy (Bx) of Prostate MRI Lesions.  Radiological Society of North America 2013 Scientific Assembly and Annual Meeting, December 1 - December 6, 2013 ,Chicago IL. http://archive.rsna.org/2013/13023150.html