Accuracy of three-dimensional prostate cancer map based on MR/US fusion biopsy
Kazumi Kamoi, Koji Okihara, Fumiya Hongo, Yoshio Naya, Naitoh Yasuyuki, Iwata Atsuko, Terukazu Nakamura, Tsuneharu Miki
Department of Urology Kyoto Prefectural University of Medicine
Introduction: The aim of this study was to develop 3D-cancer map depending on both MRI findings and histology by MR-US fusion biopsy, and to evaluate its accuracy of predicting eligibility for focal cryotherapy by comparing with radical prostatectomy (RP) specimen.
Methods: A cognitive fusion biopsy (n = 74) or a MRI-US fusion biopsy (n = 67) was used for PCa diagnosis before RP. Cognitive fusion biopsies were performed using 2 target cores and 8 systematic cores. MR-US fusion biopsies were performed using Urostation (Koelis, France) with 2 targeted cores and 6 cores from conventional sextant sites. 3D-cancer maps were created and compared with the RP specimens in terms of the primary Gleason grades and the number of significant tumors. The MRI-estimated tumor volume and exptraprostatic extension were also compared with those in the RP specimens. The indication of focal treatment according to the whole mount section analysis was arbitrarily determined as tumor volume less than 3ml, Gleason sum of 7 or less, and no extraprostatic extension.
Results: The primary and secondary Gleason grades in the index lesion, the number of significant tumors on 3D-cancer map, findings of pT3 disease were better matched with the correspondences to the RP specimens in the MR-US fusion group than cognitive fusion group (kappa values of 0.31 vs. 0.25, 0.64 vs. 0.37, and 0.40 vs. 0.30, respectively). The MRI-estimated tumor volume correlated better with the volume of index tumor in patients underwent MRI-US fusion biopsy than cognitive fusion biopsy (Pearson’s correlation coefficients of 0.73 vs. 0.31). The sensitivities and specificities of predicting focal therapy indication by 3D-cancer map were 95% (20/21) and 80% (37/46) in the MR-US fusion group, and 88% (21/24) and 72% (36/50) in the cognitive fusion group, respectively.
Conclusions: MR-US fusion biopsy has significant potential to detect the pathological grade and size of index tumor which enables tailored treatment depending on the oncological characteristics.