Abstract
Aim: To investigate if targeted prostate biopsy (TBx) has superior performance to standard untargeted biopsy (SBx) in determining the optimal agreement between biopsy and surgical Gleason Score (GS). Patients and Methods: An analysis of our institutional longitudinal database identified 683 consecutive patients who underwent either SBx (18-20 standardized transrectal ultrasound peripheral/transitional zone cores) or TBx alone (4-6 cores for each multiparametric magnetic resonance suspicious lesion, Prostate Imaging Reporting and Data System [(PI-RADS)≥3] after a previous negative first SBx. A total of 246 consecutive patients with diagnosis of prostate cancer (117 SBx and 129 TBx diagnoses) who underwent robot-assisted radical prostatectomy between January 2014 and December 2015, were enrolled. The concordance of biopsy GS to pathological GS, as well as the association between categorical variables [age, digital rectal exam (DRE), TNM, PI-RADS], were analyzed by Fisher's exact test. Results: Prostate cancer was diagnosed in 32.0% of the SBx group and in 49.3% of TBx. The rate of correctly classified, up-graded and down-graded GS was 53.8% vs. 91.5%, 39.3% vs. 7.8% and 6.8% vs. 0.8% for SBx and TBx, respectively (p<0.001). The GS concordance rates for SBx and TBx cohorts were: 14.3% vs. 41.7% for GS 6, 61.0% vs. 83.8% for GS 3+4, 56.3% vs. 75.0% for GS 4+3, 27.3% vs. 100% for GS 8 and 80% vs. 100% for GS 9, respectively. Conclusion: TBx ensured a higher of accuracy of prostate cancer detection and a better performance in discriminating significant from insignificant prostate cancer, when compared to SBx. TBx significantly reduced the risk of GS up-/down-grading at radical prostatectomy for all histopathological categories. This is a notable advance in the selection of candidates for active surveillance.
Prostate cancer (PCa) exhibits a considerable biological variability which hampers accurate prediction of disease aggressiveness. Gleason score (GS) assignment is crucial for patient counselling and treatment decision-making, particularly when proposing less aggressive treatment alternatives such as active surveillance or focal treatment (1).
However, it has been reported that nearly a third of patients with PCa will have significant GS up-grading between systematic biopsy (SBx) and radical prostatectomy (RP) (2-4). These studies also suggested great variability in GS up-grading, especially in cases of biopsy GS 3+3 cancer, which was found to be up-graded to GS 3+4 on RP specimen in 20% to 66% of cases (5).
Pathological results in men who were initially followed-up with active surveillance, showed organ-confined disease and favourable Gleason grading in the majority of cases; anyway, the proportion of unfavourable outcomes could not be neglected in this group. GS up-grading could be one of the most relevant causes of risk re-classification during follow-up (6). Moreover, the preoperative anticipation of histological prognostic features at RP could affect the therapeutic approaches to localised PCa, such as preservation of neurovascular bundles and the stratification of patient risk for positive surgical margins.
Aiming at more accurate biopsy, in the last years, the research has been focused on multiparametric magnetic resonance imaging (mp-MRI), not only ensuring an higher accuracy for PCa detection, but also better performance in discriminating significant from insignificant PCa (7-8). Several studies demonstrated that mp-MRI targeted biopsies (TBx) (cognitive or visually guided) are feasible and ensured a significantly higher median detection rate than random biopsies (after a prior negative one), allowing physicians to direct prostate biopsies to suspicious lesions rather than randomly (9-13).
The aim of this study was to investigate if prostate TBx has performance superior to SBx in determining the optimal agreement between biopsy and surgical GS.
Materials and Methods
An analysis of an institutional longitudinal database (San Luigi Hospital, Orbassano, Italy) identified 683 consecutive patients who underwent either SBx (n=381) or TBx (n=302) after a previous negative first SBx (12 samples) between January 2014 and December 2015. No patient underwent both SBx and TBx procedures. All men had ongoing suspicion of PCa (elevated or rising PSA or suspicious digital rectal exam (DRE) and only in TBx one or more detectable lesions in mp-MRI.
Among them, 271 patients (39.6%) had diagnosis of PCa: 32.0% (n=122) among those undergoing SBx, and 49.3% (n=149) among those undergoing TBx. Later, 246 consecutive patients (90.7%; 117 and 129 patients in the SBx and TBx groups, respectively) underwent laparoscopic robot-assisted RP at our centre and were included in this study. RP was accompanied by standard pelvic lymph node dissection, due to their class of risk.
Cancer familiality was considered positive for men with two or more affected first-degree relatives (father or brother).
Due to the observational nature of this research and according to Italian law (Agenzia Italiana del Farmaco-AIFA, Guidelines for observational studies, 20 March 2008), no formal institutional Ethics Committee/Institutional Review Board approval was needed.
Prostate mp-MRI. All consecutive patients, having either elevated/rising PSA or suspicious DRE, underwent mp-MRI according to the European Society of Urological Radiology (ESUR) guidelines, PI-RADS version 1 score was used for each reported lesion (9). The mp-MRIs were carried out at three centres with a 1.5-T scanner using a 32-channel phase array coil or 4-channel phase array coil combined with an endorectal coil. The prostate and seminal vesicle anatomy was assessed on T2-weighted (T2W) images in the axial, coronal and sagittal planes. T1 Fast spin-echo axial images were generated to identify areas of intraprostatic haemorrhage and to evaluate the pelvic nodes and bones. Functional information was obtained by diffusion weighted imaging (DWI) and dynamic contrast enhanced (DCE) MRI. DWI was performed using axial echo planar imaging sequences at different b-values.
Three experienced radiologists analyzed the mp-MRI findings. Lesions were scored using PI-RADS and rated on a score from 1 to 5 in each of the three MRI sequences. Each lesion was given an overall total score (3-15 points). A PI-RADS grade from 1 to 5 was assigned for each patient determined by the respective total score. The sizes of the detected lesions were measured with a free-hand region of interest (ROI) on axial T2W images (mm2) and each ROI was furthermore localized into the 27 ROI models (9). Since we preferred to biopsy even the men belonging to the ‘indeterminate’ PI-RADS 3 subgroup, mpMRI was considered positive if final PI-RADS was 3 or more, and negative if less than 3.
Prostate biopsy and pathology. For the SBx group, 18-20 standardized transrectal ultrasound (TRUS) peripheral and transitional zone biopsy cores were taken. For TBx, 4-6 cores (depending on lesion diameter: 4 cores when >7 mm, 6 when ≤7 mm) for each mp-MRI-suspicious lesion (PI-RADS≥3) were taken using the BioJet™ fusion system (D&K Technologies GmbH, Barum, Germany). The gland and the ROI were contoured, and the prostate contour was fused in real time with the TRUS image. TRUS was performed using a Hawk Ultrasound scanner 2102 EXL with a side-fire probe (BK Ultrasound Europe, Herlev, Denmark) which was fixed to a mechanical stepper during the entire procedure. Two position sensors attached to this stepper, for longitudinal and rotation movements, transmitted the exact position of the probe to the fusion system. All the patients underwent a fusion biopsy in an ambulatory setting under local anaesthesia, and were fixed in a lithotomy position. Biopsies were performed via a transrectal or transperineal approach depending on the location of the ROI (posterior or anterior, respectively) using a disposable 18-G biopsy gun with a specimen size of 18-22 mm (Bard Medical, Covington, GA, USA).
Pathological staging was performed according to the seventh edition of the TNM Classification of Malignant Tumours (14). RP specimens were evaluated using serially 3 mm sectioned whole-mount specimens according to the Stanford protocol (15) and primary and secondary GS were assigned by an experienced uropathologist, blinded to the mp-MRI results, according to the 2005 consensus conference of the International Society of Urological Pathology definitions (16).
Statistical analysis. The concordance of biopsy GS with pathological GS, as well as the association between categorical variables (age, digital rectal examination, TNM, capsule/ vesicle/ neuronal/ vascular/ endolymphatic invasion; PI-RADS), were analysed by the Fisher's exact test; the Mann–Whitney and Kruskal–Wallis tests were used for continuous variables. All results for continuous variables are expressed as the median (range). All reported p-values were obtained by the two-sided exact method, at the conventional 5% significance level. Data were analysed as of February 2016 by R 3.2.3 (R Foundation for Statistical Computing, Vienna-A, http://www.R-project.org).
Results
Two hundred and forty-six patients underwent robot-assisted RP and were enrolled in the trial; in 117 patients, PCa was diagnosed at SBx, while in 129 at TBx. Main clinical and pathological characteristics of the patients' cohort are reported in Tables I and II. Overall, their median age was 65 (44-80) years, median preoperative PSA 8.3 (1.0-52.0) ng/ml and median prostate volume 41 (19-153) ml. DRE was suspicious for PCa in seven (2.8%) patients, all in the SBx group. Most patients (70.7%) had PSA ≤10 ng/ml, 70% had clinical stage T1c and 71.2% of patients had a biopsy GS 6 or 7 (3+4).
Clinical patient characteristics.
Pathological patient characteristics.
In SBx and TBx groups, median age and PSA were: 65 (45-80) vs. 67 (44-78) years, and 8.6 (3.6-52.0) vs. 6.0 (1.0-48.0) ng/ml, respectively (p<0.001 for both comparisons). In the TBx cohort, most patients (76.6%) had one suspected lesion at MRI, being PI-RADS grade 4 in 54.2% of cases.
No significant differences were found when comparing the two groups (SBx and TBx) for median prostate volume (p=0.223) and pathological stage (p=0.436), while the median tumor volume was larger in the SBx cohort (p=0.01).
Biopsy and surgery Gleason score (GS) pattern of change for standard untargeted biopsy (SBx) and targeted prostate biopsy (TBx).
We observed a strong direct association between PI-RADS and surgical GS (p=0.008); among those with GS 3+3, 83.3% had PI-RADS 3, while of those with Gleason ≥8, 80.0% had PI-RADS ≥4. The prevalence of pathological GS 3+3 was around four-fold in SBx cases vs. TBx ones (35.9 vs. 9.3%). On RP specimen analysis, cancer was multifocal in 141 (57.3%) patients. Overall, 231 patients (93.9%; 106 and 125 in SBx and in TBx cohort) had a clinically significant PCa according to definition, including 99 (40.2%) with locally advanced disease (pT3N0) and 16 (6.5%) with positive lymph nodes.
The overall GS agreement between SBx or TBx and RP is reported in Figure 1. The rate of correctly classified GS was significantly lower for SBx than TBx, (p<0.001). The rates of pathological GS underestimation (up-grading) and overestimation (down-grading) were also significantly higher by SBx than TBx (p<0.001). The correct classification rates for all histopathological GS at RP by SBx and TBx are reported in Table III.
Biopsy exact match, up-graded and down-graded for each Gleason sum category in SBx and TBx cohorts are reported in Figure 2A and B, respectively.
The primary and secondary Gleason grade pattern was determined accurately in 85 (72.6%) and 111 cases (86.0%, p=0.018), and in 56 (47.9%) and 103 patients (79.8%, p<0.001) for the SBx and TBx groups, respectively.
The rates of primary Gleason grade 4 were quite comparable between SBx and TBx (86.7% vs. 86.5%, respectively), conversely for secondary grade 4, TBx showed better performance (59.6% vs. 84.5%).
In the TBx cohort, the PI-RADS score was not associated to any GS change pattern: among PI-RADS 3/4/5 subgroups, the proportion of men having a stable GS were 73.1, 81.0, and 87.0%, respectively (p=0.132).
Rate of agreement between biopsy and surgery Gleason score (GS) by Gleason sum categories in the standard untargeted biopsy (Sbx) (A) and targeted prostate biopsy (TBx) (B) cohorts.
Discussion
Many studies demonstrated that in patients undergoing RP, the pathological GS may significantly be up- or down-graded from that of the biopsy sample (17-21). Patients having a significant Gleason up-grading have been found to have outcomes more similar to their pathological rather than their biopsy GS (22). Inaccurate Gleason grading is mainly caused by the restricted inter-observer reproducibility among pathologists and, more importantly, by the sampling error of untargeted SBx (23, 24). For this reason, considering SBx with prostate random sampling as the reference standard in PCa diagnostic could be an important limitation.
Rate of agreement between Gleason score (GS) by standard untargeted biopsy (SBx) and targeted prostate biopsy (TBx).
Indeed, in addition to other important flaws of SBx (low overall cancer detection rate and overdiagnosis of clinically insignificant disease, in ≤50% of men screened on an individual basis) (25), untargeted Bx technique produces an incorrect GS classification in about 30% of biopsies (26-27). Some authors attempted to address this issue by proposing ‘saturation’ biopsies, considering that increased tissue sampling would improve the accuracy of GS evaluation. However, even the most invasive saturation biopsy protocols using >50 template-guided transperineal cores still misclassify GS in almost one third of the cases (20).
In the present study, we demonstrated that TBx not only increases PCa diagnosis, but it is also even more accurate than SBx in estimating the correct GS based on the RP specimen. Moreover, the rates of underestimation and overestimation in prediction of pathological GS were clearly inferior for TBx vs. SBx: in particular, GS up-grading was 7.8% vs. 39.3% respectively.
Our present results are in accordance with previous studies, whereas in men undergoing RP after SBx, the GS is up-graded in 30-50% of cases (based on evaluation of the entire RP specimen) (2). Men eligible for active surveillance are particularly at risk when GS down-staging occurs (3). GS up-grading from biopsy to surgery could be one of the most relevant reasons for risk reclassification during active surveillance follow-up (5).
We demonstrated that TBx may reduce the risk of Gleason up-grading on RP for all histopathological GS, comprising 3+3. TBx had a significantly reduced false-positive GS 3+3 rate, around three-fold compared to SBx: a notable advance in the selection of active surveillance candidates.
Walton Diaz et al. reported that stable findings on mp-MRI were associated with GS stability in patients with GS 6 PCa on active surveillance and could potentially reduce the number of unnecessary biopsies in men undergoing active surveillance (28).
For GS down-grading, one-third of our SBx patients with Gleason 8 were down-graded to a lesser disease, while none were down-graded in the TBx group. This is of particular importance for patients with a biopsy GS of 8 or more, since they are often excluded from surgery and advised to undergo radiation therapy with long-term androgen deprivation.
Lanz et al. investigated the accuracy of TBx for GS determination: even if TBx had good performance on GS prediction in more than two-thirds of men, its incorrect classification appeared to be dependent on lesser accuracy of secondary Gleason grade (29). In our research, TBx demonstrated a better prognostic accuracy both for primary (86.0%) and secondary (79.8%) Gleason grade when compared to SBx.
Difficulties in determining secondary Gleason grade pattern in the Lanz et al. series were related to the determination of grade 4 rate in the biopsy targeted foci (29). In the current study, we observed that primary Gleason grade 4 rates were quite comparable between SBx and TBx (86.7% vs. 86.5%, respectively); for secondary grade 4, TBx showed a better performance (59.6% vs. 84.5%). Attaining correct determination at higher than grade 4 is quite critical, as poorer prognosis of GS 4+3 vs. 3+4 was proven in two studies (30, 31).
Moreover, Lanz et al. reported a higher GS determination on TBx for both low Gleason and high PI-RADS score cancer (29). In our experience, predictive accuracy of TBx for GS 3+3 was above mentioned. We observed a strong direct association between PI-RADS and surgical GS: allowing for the detection of the index lesion more accurately in cases of high grade PI-RADS, TBx might increase the agreement between biopsy and pathological GS.
Our study presents several limitations. Firstly, the study sample size was small, albeit one of largest population to be evaluated for Gleason concordance between SBx/TBx and RP. Secondly, the data were collected prospectively but the analyses were carried out retrospectively. Thirdly, the study did not include any discussion of costs and logistics: given increasing healthcare costs, these issues are of key importance and further studies should be advised. Fourthly, PI-RADS v1 and not the more recent v2 was used, potentially affecting the diagnostic accuracy, especially for anterior ROIs. Finally, visible and not visible tumors on MRI would make comparison of the TBx and SBx results difficult: at the same time, our aim was to evaluate the ideal biopsy/surgery GS agreement between two different biopsy techniques, not the rate of PCa diagnosis.
Due to the better TBx accuracy in GS estimation, MRI should be mandatorily performed before RP not only to guide the therapeutic approaches for localised PCa (as for the preservation of neurovascular bundles and the risk stratification for positive surgical margins), but also to select candidates for active surveillance more appropriately. Moreover, if prospective controlled trials on larger patient cohorts will confirm our findings, then mp-MRI could play a critical role even during AS follow-up, e.g. in order to reduce the number of repeat prostate biopsies. Considering that patient counselling and treatment decision-making are based on the biopsy GS determination, further studies should investigate other potential genomic or biological factors that might increase biopsy GS concordance.
- Received July 6, 2016.
- Revision received July 23, 2016.
- Accepted July 27, 2016.
- Copyright© 2016 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
