Presentation Authors: Niklas Harland*, Tübingen, Germany, Giorgio Ivan Russo, Catania, Italy, Joachim Mathes, Steffen Rausch, Eva Neumann, Sascha Kaufmann, Arnulf Stenzl, Jens Bedke, Tübingen, Germany, Stephan Kruck, Pforzheim, Germany
Introduction: Despite the growing evidence supporting MRI in prostate cancer (PC) diagnosis, there is still a lack of fast, cheap and accurate imaging tools to rule out significant PC (csPC). Consequently, we developed a new imaging method based on computed tomographic (CT) ultrasound (US) supported by artificial neural network analysis (ANNA).
Methods: 202 consecutive patients in primary or re-biopsy setting with visible mpMRI (axial 3mm slices) lesions (PIRADS V2) were scanned and recorded by robotic CT-US (axial 0.5mm slices) as part of robot-assisted elastic mpMRI-TRUS fusion biopsy (RA-FB). Only significant index lesion (Gleason Score â‰¥3+4) verified by biopsy and whole-mount pathology after radical prostatectomy (RP) were retrospectively analyzed. Their visibility was reevaluated by two blinded investigators by gray-scale US and ANNA.
Results: In the whole cohort, csPC was detected in 105 cases (52%) by RA-FB. Whole-mount histology after RP was available in 44 cases (36%). In this subgroup mean PSA level was 8.6 ng/ml, mean prostate volume was 33 cc and mean tumor volume was 0.5 cc. Median PIRADS of index lesions was 4 and median Gleason Group was 3 based on biopsy and final histology. pT3 was found in 7 cases (16%). Index lesions were visible in gray-scale US in 25 cases (57%). Use of ANNA tool we identified index lesions in 30 cases (68%) while combining CT-US-ANNA we detected index lesions in 35 patients (80%).
Conclusions: First results of multiparametric computed ultrasound imaging showed promising detection rates, using grayscale and ANNA, in patients with csPC selected by RA-FB. In the near future, precise axial US imaging with reduced slice thickness and availability of complete US prostate scans for ANNA have the potential to revolutionize PC diagnosis.