Data on the detection of clinically significant prostate cancer by magnetic resonance imaging (MRI)-guided targeted and systematic biopsy

This is a data article from the original publication “Reasons for missing clinically significant prostate cancer by targeted magnetic resonance imaging/ultrasound fusion-guided biopsy” [1]. From January 2014 to April 2019 a sample collective of 785 patients with 3T multiparametric magnetic resonance imaging (mp-MRI) of the prostate and subsequent combined systematic biopsy (SB) and magnetic resonance imaging/ultrasound (US) fusion-guided biopsy (TB) was retrospectively analyzed. Prostate cancer (PCa) detection by TB and/or additional SB was analyzed.


Specifications
Radiography and radiology Specific subject area: Imaging-guided prostate cancer detection Type of data:

Value of the Data
• These data show the detection rates and accuracy of the Prostate Imaging and Reporting Data System (PI-RADS) version 2.1 classification, which is a standardized scoring system for prostate MRI assessment. As standardization in interpretation of prostate MRI is necessary, the presented data confirm that the PI-RADS v2.1 is an excellent tool for csPCa risk stratification. • The benefit of systematic biopsy (SB) additional to MRI-targeted biopsy (TB) was analysed.
As there is an on-going discussion regarding the best biopsy approach, our research suggests that additional SB adds overall a limited value. • Researchers in radiology and urology may profit from these data because the data can be used for further studies on mpMRI and MRI-guided biopsy (e.g., sample size calculation).
Radiologists and urologists can use the results for their prostate biopsy or MRI follow-up decision.

Data Description
The presented dataset supports the research article "Reasons for missing clinically significant prostate cancer by targeted magnetic resonance imaging/ultrasound fusion-guided biopsy" [1] . Of 2,418 patients referred to the Departement of Interventional and Diagnostic Radiology (at the University Hospital Duesseldorf, Germany) for mp-MRI of the prostate within the study period, 785 patients (mean age 65 ± 9 years; median PSA value 8.1 ng/ml, interquartile range 5.9 -12 ng/ml; median prostate volume 47 ml, interquartile range 33 -66 ml) received complete, subsequent MRI/US fuison-guided targeted (TB) and systematic (SB) biopsy and were finally analysed ( Fig. 1 ). Different 3T MRI scanners were used (Magnetom Trio TIM: n = 549, Skyra: n = 90, Prisma:  Table 1 . All mp-MRI were evaluated according to the Prostate Imaging and Reporting Data System (PI-RADS) version 2. PI-RADS is a standardized scoring system to interpret and judge findings in prostate MRI on a scale from 1 (highly likely benign) to 5 (highly likely suspect for a clinically significant prostate cancer; csPCa). Localization and maximum diameter of the MRI index lesion (IL) were documented. PSA values, prostate volume, and PSA density (PSA-D) are provided in the corresponding research article [1] . The overall PCa detection rate was 59% (n = 461) including 84% csPCa (n = 342; ISUP grade group ≥2). The detection rate of csPCa was the highest in patients with primary biopsy (51%) compared to patients with a secondary biopsy (34%), or patients on Active Surveillance (AS) (47%) ( Table 2 ). Among all patients, TB detected 300 csPCa (88%), whereas SB detected only 247 csPCA (72%) and more non significant (ns) PCa (ISUP grade group 1) ( Table 3 ). In 30 cases (8.8%) a csPCa was detected by SB only and in 67 (20%) cases by TB only. In 12 cases (3.5%) SB lead to a Gleason upgrade ≥ ISUP 2, TB respectively in 28 cases (8.2%). 44 additional nsPCa were detected by SB and 34 by TB ( Table 4 ). The PCa detection according to the PI-RADS classification v2.1 is shown in Table 5 . In 98% of patients in PI-RADS assessement category 5 and in 62% of patients in PI-RADS category 4 a PCa was histopathologically proven, including 85% and 44% csPCa, respectively. Therefore, a sensitivity of 94% and specificity of 54% for all PCa and 98% and 44% for csPCa was achieved     Note. -csPCa = clinical significant prostate cancer; nsPCa = non-significant prostate cancer; PCa = prostate cancer; SB = 12-core transrectal ultrasound-guided biopsy; TB = MRI/US-fusion-guided biopsy.  ( Table 6 ) . The .xls file "MRI_Biopsy_Data" contains clinical, MRI, biopsy, and histopathological variables ( Supp. document ).

Experimental Design, Materials and Methods
Patients with elevated PSA who received a 3T mp-MRI of the prostate in accordance with the current PI-RADS recommendations were enrolled in this study [ 2 , 3 ]. Patients with subsequent combined 12-core SB and TB using elastic fusion of real-time ultrasound segmentation data by UroNAV biopsy system (Philips Healthcare, Invivo Corporation, USA) were finally retrospectively analyzed. Prostate segmentation (MRI contouring of the prostate gland boundary) and lesion registration (marking the cancer suspicious regions (mCSR) in the axial T2 images for targeted biopsy with a 3D region of interest (ROI) of the total lesion and a sub-ROI of the lesion center correlating to the area with the lowest apparent diffusion coefficient (ADC) value) were performed using the DynaCAD software (Version 3 or 4, Philips Healthcare, Invivo Corporation, USA). Two targeted cores were obtained from each registered mCSR described in the mp-MRI report. All biopsy cores from TB and SB were histopathologically evaluated in accordance with the recommendations of the International Society of Urological Pathology (ISUP). CsPCa was defined as ISUP grade group 2 or higher in at least one lesion; Gleason score ≥ 3 + 4 = 7 [4] .