Comparison of prostate cancer detection rates in patients undergoing MRI/TRUS fusion prostate biopsy with two different software‐based systems. Issue 2 (3rd November 2021)

• Record Type: Journal Article
• Authors: Hanske, Julian; Risse, Yannic; Roghmann, Florian; Pucheril, Daniel; Berg, Sebastian; Tully, Karl H.; von Landenberg, Nicolas; Wald, Jan; Noldus, Joachim; Brock, Marko
• Is Part Of: Prostate. Volume 82:Issue 2(2022)
• Page Start: 227
• Page End: 234
• Publication Date: 2021-11-03
• Subjects: comparison -- diagnosis -- magnetic resonance imaging -- prostate cancer -- software‐based -- targeted biopsy; Prostate -- Diseases -- Periodicals; 616
• DOI: 10.1002/pros.24264
• Languages: English
• ISSNs: 0270-4137

Abstract

Abstract: Background: Magnetic resonance imaging (MRI)‐targeted prostate biopsy is a routinely used diagnostic tool for prostate cancer (PCa) detection. However, a clear superiority of the optimal approach for software‐based MRI processing during biopsy procedures is still unanswered. To investigate the impact of robotic approach and software‐based image processing (rigid vs. elastic) during MRI/transrectal ultrasound (TRUS) fusion prostate biopsy (FBx) on overall and clinically significant (cs) PCa detection. Methods: The study relied on the instructional retrospective biopsy data collected data between September 2013 and August 2017. Overall, 241 men with at least one suspicious lesion (PI‐RADS ≥ 3) on multiparametric MRI underwent FBx. The study protocol contains a systematic 12‐core sextant biopsy plus 2 cores per targeted lesion. One experienced urologist performed 1048 targeted biopsy cores; 467 (45%) cores were obtained using rigid processing, while the remaining 581 (55%) cores relied on elastic image processing. CsPCa was defined as International Society of Urological Pathology (ISUP) grade ≥ 2. The effect of rigid versus elastic FBx on overall and csPCa detection rates was determined. Propensity score weighting and multivariable regression models were used to account for potential biases inherent to the retrospective study design. Results: In multivariable regression analyses, age, prostate‐specific antigen (PSA), and PIRADS ≥ 3 lesion were related to higher odds of finding csPCa. Elastic software‐based image processing was independently associated with a higher overall PCa (odds ratio [OR] = 3.6 [2.2–6.1], p < 0.001) and csPCa (OR = 4.8 [2.6–8.8], p < 0.001) detection, respectively. Conclusions: Contrary to existing literature, our results suggest that the robotic‐driven software registration with elastic fusion might have a substantial effect on PCa detection.