Presentation Authors: Nicholas Chakiryan*, Ann Martinez-Acevedo, Luc Boileau, Angela Waldron, Bryan Foster, Fergus Coakley, Ryan Kopp, Christopher Amling, Jen Jane Liu, Portland, OR
Introduction: MRI fusion biopsy (Fus-Bx) of the prostate improves the overall detection rate of clinically significant prostate cancer (CSPC) compared to standard template biopsy (Std-Bx). MRI cognitive fusion biopsy (Cog-Bx) can also be used to target lesions visible on MRI, but without the need for Fus-Bx equipment. We prospectively compare detection of prostate cancer (PC) between Fus-Bx and Cog-Bx modalities.
Methods: From July 2017 - September 2018 patients with suspected PC, or who were on an AS protocol for low risk PC, underwent a multiparametric 3.0 Tesla prostate MRI. Patients with an identifiable lesion PiRADS 3 or greater were identified for both cognitive and MRI fusion biopsy. Each patient underwent targeted Cog-Bx of the suspicious lesion, immediately followed by targeted Fus-Bx using the UroNav image-guided stereotactic biopsy system, and then Std-Bx. Concordance rates between the Fus-Bx and Cog-Bx were described, and agreement between modalities was assessed using Cohen's kappa coefficient. Sensitivity was calculated for each modality against the "gold standard" defined as the highest Gleason grade group (GGG) detected using all three modalities (Std-Bx, Cog-Bx, and Fus-Bx). CSPC was defined as GGG 2 or greater. High grade prostate cancer (HGPC) was defined as GGG 4 and 5.
Results: 40 patients were included in the analysis. The mean age was 66.6 years (sd: 8.1). Median PSA was 8.41 ng/dl (IQR: 5.67-13.37). Median gland volume was 45.6cc (IQR: 34.25-74.55). Median lesion size was 1.15cm in largest dimension (IQR: 0.83-1.40), and 0.63cc in volume (IQR: 0.30-1.80). Median number of cores examined was 16 (IQR: 14-17). 32 patients (80%) had positive biopsies. 21 patients (52.5%) had CSPC. 7 patients (17.5%) had HGPC. Concordance between Fus-Bx and Cog-Bx was 85.0% for detection of any cancer, 92.5% for detection of CSCP, and 100% for HGPC. A high degree of agreement between modalities was shown with Cohen's kappa coefficient (Any cancer: 0.65; CSPC: 0.85; HGPC: 1.00). The targeted modalities had similar sensitivities for detection of any cancer (Cog-Bx: 0.75; Fus-Bx: 0.72), CSPC (CogBx: 0.95; Fus-Bx: 0.86), and HGPC (Cog-Bx: 1.00, Fus-Bx: 1.00). Sensitivity improved significantly when both Cog-Bx and Fus-Bx were analyzed together (Any cancer: 0.84; CSPC: 1.00; HGPC: 1.00).
Conclusions: Concordance rates were high between Cog-Bx and Fus-Bx. Sensitivities for each test were similar for detection of any cancer, CSPC, and HGPC. Both Cog-Bx and Fus-Bx detected CSPC and HGPC with high sensitivity. Fus-Bx did not demonstrate an objective benefit over Cog-Bx in our cohort.