Discussion
The Gleason grading system has been shown to be the most important determinant of tumor aggressiveness, disease outcome, and mortality from prostate cancer (19). Gleason Score discordance can confound optimal treatment allocation of patients diagnosed with prostate cancer and place them at risk of worse oncological outcomes. MRI targeted biopsies, which utilise previously taken MRI images of suspected cancer and fuse them with real-time ultrasound images, are reported to be able to detect high grade cancers better than systematic biopsies and TRUS (6-9, 20,21). However, several recent studies have reported a mismatch in the biopsy and radical prostatectomy Gleason Score (10,11, 22). Despite the improved detection of clinically significant cancers with MRI targeted biopsies, considerable disparity exists in the literature regarding MRI targeted biopsy’s ability to better detect RP pathology in biopsy-naive men over conventional TRUS biopsies (10,11,22). We performed this study in order to investigate if targeted prostate biopsy has a performance superior to untargeted biopsy in determining the optimal agreement between biopsy and surgical Gleason Score. Our findings suggest that in the overall patient cohort, combined TBx+SBx provides improved detection rates particularly for higher-grade disease over either systematic or MRI-targeted biopsy or TRUS-Bx alone. However, superiority in significant cancer detection did not reflect accurate Gleason Grade Group in the whole prostate gland at final surgical pathology due to selective sampling of higher grade areas within an otherwise low grade cancer. MRI targeted biopsy was better in the prediction of the result of final histopathological analysis for ISUP 1 and ISUP 4 than systematic biopsies or TRUS-Bx. Our analysis of biopsy techniques according to ISUP Grade groups showed increased downgrading in MRI fusion biopsy and increased upgrading in TRUS-Bx.
In this cohort of patients, combined TBx+SBx biopsy provided more accurate diagnosis than MRI targeted or systematic biopsy alone. Consistent with earlier studies, we found higher cancer detection rates on TBx when compared to SBx (7,9, 23). However, 13 out of 127 prostate cancers (10.2%) were missed by TBx alone. Missed cancers by TBx may reflect the underlying limitation of mpMRI which might result from factors such as PCa not visible on mpMRI, varying inter-reader agreement on the mpMRI results and/or missing the target lesion in biopsy.
Our results have shown that MRI targeted biopsies are significantly better for the detection of clinically significant PCa (ISUP 2 or higher) than TRUS-Bx. Almost half of the cancers detected by TRUS-Bx were in ISUP grade 1 category. When TBx alone is compared to SBx alone for the detection rate of significant prostate cancer diagnosis, overall targeted biopsy alone tends to detect more significant PCa but this superiority does not reach statistical difference. In the present study, SBx had superior performance to TRUS-Bx in determining high grade cancers. This could be a result of the cognitive fusion bias within the SBx that the urologist performing the SBx was aware of the localization of the suspicious lesion on mpMRI. In our study, if a pure TBx strategy omitting SBx is applied, this will lead to missing significant prostate cancer in 11 patients. Our results support the notion that in order to obtain the most accurate assessment of the entire prostate gland especially for patients at risk of significant disease, SBx remains necessary, in addition to the TBx due to limitations of mpMRI performance/reading and of precision during lesion targeting.
In this study, combined TBx+SBx showed significantly better concordance than any biopsy method alone in ISUP Grade 1. Also, combined TBX+SBx prostate biopsy  resulted in significantly less pathologic upgrading in ISUP 1 grade group as compared with TRUS-Bx at prostatectomy (38.7% vs 62.1%). Men under active surveillance in ISUP Grade 1 group diagnosed by TRUS-Bx are at significantly greater risk of subsequent reclassification by confirmatory biopsies. It has been reported that up to 40% of men potentially suitable for AS had unfavorable disease at RP and these high rates of adverse pathologic findings which can underestimate aggressiveness of the disease (24-26). Upgrading is particularly a great concern in the context of active surveillance. Pathologic upgrading has been shown to be associated with adverse outcomes, including higher rates of biochemical recurrence (27,28). A recent study by our group has shown that men suitable for AS, but elected immediate RP, proved to have a GS upgrade rate of 30.6% and a pathological upstaging rate of 13.2% (29). In the present study, our results clearly suggest that TRUS-Bx based active surveillance decision should be questioned unless confirmatory mpMRI TBx+SBx is done. In the active surveillance scenario, it is imperative to decrease the risk of missing a high-grade disease and delaying a radical treatment, providing more confidence to the urologist and the patient with conservative management of PCa. Based on our data, multiparametric MRI TBx+SBx is the best available strategy to stratify men to AS with reliable eligibility and should be included in the AS protocols for a more accurate grading of PCa.
Among all biopsy methods, the lowest concordance was achieved in ISUP Grade 4. Combined TBx+SBx showed significantly better concordance than TRUS-Bx in ISUP Grade 4. At this group, TRUS-Bx showed increased upgrading (47.1%) but in contrast TBx+SBx showed increased downgrading (54.6%). Downgrading may be due to oversampling that small foci of Gleason 5 detected at biopsy but not documented at radical prostatectomy (4). In ISUP Grade 4, SBx was better than TRUS-Bx. This, again, shows that inadequate diagnostic biopsy sampling is high in TRUS-Bx and most of the SBx are not blind and affected by MRI cognition.
Our study has demonstrated that TRUS-Bx concordance was superior for ISUP 2,3 and 5 Grade Groups than TBx+SBx. In all ISUP groups except 1 and 4, TRUS-Bx showed better performance in predicting final pathology with higher concordance. MRI Targeted biopsies are more prone to downgrading and TRUS biopsies are more upgraded. In the present study, downgrading was higher in TBx+ SBx biopsy group compared to conventional TRUS-Bx. Downgrading may occur for several reasons, including over-sampling of the very small (less than 5%) foci of Gleason pattern 4 or 5 cancer (4, 30). There is a risk of oversampling with increased number of targets registered by radiologist and increased number of biopsy cores taken at discretion of the biopsy performer. A very small foci of Gleason pattern 4 or 5 tumor might be missed by the routine radical prostatectomy pathologic examination or might be noted as the presence of a tertiary Gleason pattern if these small tumors are detected. It is reported that the issue of accounting for tertiary grade patterns is significant as can be seen in almost 20% of RP specimens (4). An increased number of target lesions on mpMRI and thus,an increased number of core biopsy per lesion may cause needle biopsy to sample a tertiary higher grade pattern in the RP, which is not recorded in the standard GS reporting, resulting in apparent overgrading on the needle biopsy. Our results suggest that if the physician is assigning the patient as high risk group solely based on biopsy pathology, there would be a risk of overtreating quite a few patients. Having increased the percentage of surgical pathology downgrading, MRI fusion biopsy implies a high rate of overdiagnosis of Gleason sum 8 scores on biopsy, potentially leading to suboptimal treatment strategies and patient distress.
Although conventional TRUS-Bx is a blind, nontargeted and not always anatomically systematic biopsy, our findings portray that TRUS-Bx is superior but not yet significant over SBx in accurate prediction of final pathology. However, there is a difference of 7.6% and 22.4 % for ISUP 1 and 4 grades between the schemes respectively favoring SBx. This may be a result of cognitive impact of MRI findings that the unblind nature of systematic biopsies after mpMRI were probably contaminated by MRI findings and not even described to be so, in practice it influences and orients biopsy.
The major limitations of our study are its retrospective nature and analysis. Another limitation is that TBx data was obtained from 5 different centers performed by 7 different urologists using different MRI and MRI-US Fusion devices. Failure of mpMRI fusion biopsy due to mpMRI incorrect image registration or mismatch of image planes, inaccurate sampling and intralesional Gleason Score heterogeneity may have impacted our results. Another important limitation is that there was no centralized pathological examination, multicentric pathological examinations by uropathologists at respective centers. However, our data reflects the real-life nationwide picture.
In conclusion, our results have shown that MRI-targeted biopsies are significantly better for the detection of clinically significant PCa (ISUP 2 or higher) than TRUS-Bx. However, superiority in significant cancer detection appears at least in part as a result of oversampling of higher grade areas within an otherwise low grade cancer and may not reflect accurate Gleason Grade Group in the whole prostate gland at final surgical pathology due to increased downgrading. There is a risk of overtreatment if biopsy Gleason Group was taken as a unique parameter for treatment decision. TRUS-Bx has superior concordance with radical prostatectomy for ISUP 2,3,5 grade groups, but increased upgrading should also be taken into account leading to undertreatment. MRI Targeted biopsies combined with systematic biopsies has the greatest concordance in ISUP Grade 1 which is of utmost importance for active surveillance indicating that active surveillance decision should be based on MRI targeted + systematic biopsies either at initial diagnosis or in confirmatory biopsy.