Follow-up of men with a PI-RADS 4/5 lesion after negative MRI/Ultrasound fusion biopsy

Magnetic resonance imaging/Ultrasound (MRI/US) fusion targeted biopsy (TB) in combination with a systematic biopsy (SB) improves cancer detection but limited data is available how to manage patients with a Prostate Imaging-Reporting and Data System (PI-RADS) ≥ 4 lesion and a negative biopsy. We evaluate the real-world management and the rate of clinically significant Prostate Cancer (csPCa) during follow-up. 1546 patients with a multi-parametric MRI (mpMRI) and a PI-RADS ≥ 3 who underwent SB and TB between January 2012 and May 2017 were retrospectively analyzed. 222 men with a PI-RADS ≥ 4 and a negative biopsy were included until 2019. For 177/222 (80%) complete follow-up data was obtained. 66/84 (78%) had an initial PI-RADS 4 and 18 (22%) a PI-RADS 5 lesion. 48% (84/177) received a repeat mpMRI; in the follow-up mpMRI, 39/84 (46%) lesions were downgraded to PI-RADS 2 and 11 (13%) to PI-RADS 3; three cases were upgraded and 28 lesions remained consistent. 18% (32/177) men underwent repeated TB and csPCa was detected in 44% (14/32). Our study presents real world data on the management of men with a negative TB biopsy. Men with a positive mpMRI and lesions with high suspicion (PI-RADS4/5) and a negative targeted biopsy should be critically reviewed and considered for repeat biopsy or strict surveillance. The optimal clinical risk assessment remains to be further evaluated.


Patients and methods
Study population. Out of a prospective database of 1546 men who underwent transrectal MRI/US TB and SB due to suspicion of prostate cancer at Charité-Universitätsmedizin Berlin between January 2012 and May 2017, we retrospectively analyzed men with a negative biopsy of PI-RADS Score ≥ 4 lesions. Follow-up data were extracted from patient files where available and by a patient questionnaire. The questions included inter alia post-biopsy patient guidance and the indication for a repeat mpMRI or biopsy. The indication was made by the treating outpatient urologist as it is the standard of care in Germany. Patient meta-data was collected in accordance with the standards of reporting for MRI-TB studies (START) checklist in a prospective database 17 . Subgroups of this cohort were included in previous analyses 17 . All patients signed a written informed consent for the intervention, data acquisition, and data appraisal. The study was performed according to the Declaration of Helsinki and authorized by the Institutional Review Board of the Charité-Universitätsmedizin Berlin.
Multiparametric imaging control. All patients received a 3-T mpMRI (Magnetom Skyra; Siemens Medical Systems, Erlangen, Germany) at Charité-Universitätsmedizin Berlin. The MRI protocol always comprised multi-planar (axial and coronal) high spatial resolution T2-weighted turbo spin-echo sequences (T2w TSE), axial T1-weighted images, axial diffusion-weighted images (DWIs; measured b-values 0.400 and 800 s/mm 2 , calculated b-value of 1 400 s/mm 2 ) and gadolinium-based dynamic contrast-enhanced (DCE) sequences. T2w imaging and DWIs were performed in all patients and DCE MRI in most patients. In compliance with the guidelines of the European Society of Urogenital Radiology (ESUR) the evaluation and validation of the mpMRI image data were performed or supervised by a team of experienced expert radiologists at Charité-Universitätsmedizin Berlin using PI-RADS version 2 (v2). Lesions initially rated using PI-RADS V1 were re-rated using V2 for the analysis. Experienced radiologists were defined by the consensus statement with a minimum number read of 1000 mpMRIs and a yearly read over 200 18 . Considering clinical routine, radiologists were not blinded to clinical data. For patients with multiple lesions, the maximal PI-RADS score was used for further analysis.

MRI/US fusion-guided TB and SB.
In accordance with the EAU guidelines at the time of the biopsy, the transrectal interventions were performed under antibiotic prophylaxis with a fluoroquinolone (ciprofloxacin) by a team of experienced urologists. MRI/US fusion-guided TB of the prostate was performed first, using the high-end US machine HiVison Preirus (Hitachi Medical Systems, Tokyo, Japan) or Aplio 500 (Toshiba, Otawara, Japan) with endocavity endfire probes (11C3, Toshiba; EUP V53 W; Hitachi Medical Systems) or a biplane probe transrectal (EUP CC531, Hitachi Medical Systems), as described previously 6 . Per PI-RADS lesion 3 targeted biopsies were obtained. Subsequently, transrectal SB was performed with 10 cores from left/right apex, left/right lateral midgland, left/right base, left/right ventral and left/right paraurethral.
All cores were potted and documented separately and were examined and analyzed by a certain team of experienced pathologists at Charité-Universitätsmedizin Berlin.
Definition of clinically significant prostate cancer. Clinically significant prostate cancer was defined as Gleason score ≥ 3 + 4 = 7 (International Society of Urological Pathology [ISUP] grade 2). Gleason scoring was based on the highest grade detected on histological analysis. Statistical analyses. Continuous variables were described using medians and interquartile ranges (IQR), whereas categorical variables were characterized using proportions. All descriptive analyses were performed using the Statistical Package for the Social Sciences (SPSS) software version 25 for Mac OS (SPSS Inc., IBM Company, Chicago, IL, USA).

Research involving human participants. Approval was obtained from the ethics committee of Charité
University Medicine Berlin. The procedures used in this study adhere to the tenets of the Declaration of Helsinki.

Results
The definitive cohort included a total number of 222 men with a PI-RADS score of 4 or 5 and a negative TB in combination with a SB. Complete Follow-up data was available for 177/222 (80%) men with a median follow-up of 45 months (IQR 35-61) ( Fig. 1 and 2).
60/84 (71%) men received a new MRI but no additional biopsy because of a downgrading of the lesion or a constant finding either in MRI and/ or because of constant clinical parameters.
The underlying reasons for no further biopsies were available in 110/177 (62%) men (Supplemental 3). In 71% (78/110) no further biopsy was warranted due to stable clinical parameters such as PSA and DRE. 12% (13/110) of men had left urological care and 7% (8/110) patients did not agree to a new biopsy despite an indication. In 79% (85/110) of the patients, the decision for no further biopsy was made by the treating outpatient urologist and in 21% (23/110) men the decision against a biopsy was based on their own opinion.

Discussion
Despite the positive impact of the mpMRI on selected men at risk for csPCa not all men with a PI-RADS 4 or 5 lesion have cancer detected in prostate biopsy and their follow-up remains unclear. Currently, there is little data on the real-life management of men after a negative TB approach.
Kinnaird et al. support that a follow-up TB is necessary if suspicious lesions are seen on the follow-up MRI as 54% of the patients had csPCa with a PI-RADS 5 lesion in their cohort, which is in line with our findings 19 . Hence, mpMRI works as a triage-test in biopsy-naïve men and is further highly reliable in excluding significant disease in men with previous negative biopsy 11,20 . Panebianco et al. underline that as 1827 (75% of 2422 in total) patients with positive mpMRI after an initial negative were diagnosed with PCa and 1103 (46%) had csPCa, respectively 21 . At the same time, a MRI may indicate false-positive findings (e.g. PROMIS: positive predictive value 51%, specificity 41%), so that a biopsy may be unnecessary 11,22 .
TB surpasses the SB in terms of detection rate in ISUP Grade > 2 in the repeat-biopsy setting 2 . In our follow-up biopsy setting the cancer detection rate (CDR) in total was 66% (21/32) whereas in a primary setting the CDR for a targeted biopsy alone and for TB in combination with SB was 63% and 74% in our institution in a similar cohort, respectively 6 .
There is little data or guidance on how to manage these men especially if a PI-RADS ≥ 4 lesion was biopsynegative. Ullrich et al. recommend a re-biopsy of a PI-RADS 4 lesion in the peripheral zone when there has been a negative biopsy before but suggest a follow-up for negatively biopsied transition zone lesions 23 . The updated German guidelines have added a recommendation that no further imaging or biopsy is indicated after a negative TB if clinical parameters are stable 22 .
Additionally, one needs to strongly consider a reevaluation of the MRI and technical issues of the TB itself may have led to false-negative biopsy results. With a broader use of the mpMRI, the known pitfalls such as reader inter-variability, imaging quality and missed TB may lead to an increase of men with negative TB and the need of standardized patient management 24 . As we found a high rate of downgrading (60% (50/84)), a "reference" radiology might give further guidance. One explanation may be prostatic inflammations which mimic cancer lesions on PI-RADS, which can disappear on further follow-up 23,25,26 .
Concordant to our findings Meng et al. established that 88/497 (18%) of patients with a PIRADS 4 or 5 lesion prior to initial biopsy had no cancer in TB 27 due to a PI-RADS downgrading in 73% of men on repeat MRI 27 . The prospective study by Meng 27 over 4 years included 497 men with a negative biopsy out of a cohort of 1595 men (overall 31% of men), whereas in our consecutive cohort only 14% (222/1546) of PI-RADS 4/5 cases presented with a negative biopsy result. This finding may be explained by our single center experience with a high primary detection rate 6 .
Persistent PI-RADS 4/5 lesions are at high risk of missed PCa, but false positive findings still exist as e.g. PCa in the transitional zone is still challenging 23,27,28 . Prostate deformation, patient movement, mismatch in imaging may be other reasons for failure in TB 24 . www.nature.com/scientificreports/ With a follow-up of 80% and its prospective documentation this study needs further long-term data and potential data of the PRECSION trial will give further insight into risk assessment of patients with a negative biopsy, either SB or TB 12 . The limitations of our study are its retrospective nature and the lack of histological  www.nature.com/scientificreports/ proof of absence of cancer in all men. A further limitation is, that the analysis presents the real-world follow up data and clinical decision were not based on a standard follow-up protocol and may therefore underestimate the absence on csPCa during follow-up.

Conclusion
Our study presents real world data on the management of men with a negative TB biopsy. Men with a positive mpMRI and lesions with high suspicion (PI-RADS4/5) and a negative targeted biopsy should be critically reviewed and considered for repeat biopsy or strict surveillance. The optimal clinical risk assessment remains to be further evaluated.

Data availability
The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.