Radiol Oncol 2005; 39(1): 37-47. Is quadrant biopsy adequate as first-line sampling scheme in men likely to have non-organ-confined prostate cancer: comparison to extended biopsy protocol Zoran Brnić 1, Petar Anić 1, Slavko Gašparov 2, Nikola Radović 3, Damir Kučan Željko Vidas 3, Žarko Zeljko 3, Petar Lozo 5, Vesna Ramljak 6 1Department of Diagnostic and Interventional Radiology, 2Department of Pathology, 3Department of Urology; University Hospital Merkur, Zagreb, Croatia; 4Poliklinika Lozo, Department of Ultrasound, Zadar, Croatia; 5University Hospital for Tumors, Zagreb, Croatia Background. While extensive prostate biopsy (PB) in the patients with early prostate cancer (PC) provides better sensitivity and more precise tumour staging, in the patients with advanced PC, it is virtually only a confirmation of malignancy. The purpose of our study was to find out whether the quadrant prostate biop-sy (QPB) provides a sufficient first-line pathological evaluation in the patients likely to have advanced PC, and whether the reduction of core number impairs the competence of PB through missing quantitative his-tology information. Methods. We studied 84 men who underwent PB and classified into groups »H« (highly-) and »L« (low likely to have advanced PC). Pathological results of 5-12 cores PB and simulated QPB were retrospectively compared, particularly for the presence of PC, tumour volume, Gleason score (GS), and the presence of highgrade prostatic intraepithelial neoplasia (HGPIN). Results. The PC detection rate was not impaired in group H, but dropped significantly in group L, while the percentage of positive cores was insignificantly changed in group H (p=0.39), but significantly decreased in group L (p=0.04) due to the sampling scheme reduction. No HGPIN was missed with QBP in group H, while 2 HGPIN were missed in group L. Insignificant GS changes resulted in both groups as a consequence of the limitation to QPB. Conclusions. QPB is an appropriate first-line scheme in the patients with advanced PC as the information lost due to the core number reduction is mainly not critical for patient management. Key words: prostatic neoplasms - pathology; biopsy, needle; prostate-specific antigen Received 26 July 2004 Accepted 9 August 2004 Correspondence to: Zoran Brnić, MD, PhD, University Hospital "Merkur", Department of Diagnostic and Interventional Radiology, Zajčeva 19, 10000 Zagreb, Croatia. Phone: +385 98 199 12 26; Fax: +385 1 243 14 14; E-mail: zoran.brnic@zg.htnet.hr Introduction Standard sextant prostate biopsy (PB) is proved to be of limited sensitivity in prostate cancer (PC) detection. An increase of the num-ber of tissue cores per PB session improves the PC detection rate,1-5 and contributes to a 38 Brnić z et al / Quadrant biopsy in men with advanced prostate cancer better preoperative staging accuracy.6-8 Recently, many urologists abandoned biopsy-ing hypoechoic focal lesions, and focused on systematic sampling of the gland with as much cores as possible. Although even he extensive PB was proved to be relatively safe, discomfort and minor complications occur in many patients;9-11 it is therefore sensible to avoid them if possible. In the patients with pre-sumed high tumour burden, with regard to PSA level, suspicious digital rectal examination (DRE) or transrectal ultrasound (TRUS), and suspicion of metastases,12-15 it does not seem reasonable to take large number of cores in initial PB because PC has very probably spread all over the gland volume, and exact as-sessment of intraprostatic tumour distribution is of minor importance. As these men are at high risk to have a non-organ-confined (NOC) PC, they are rarely candidates for radical prostatectomy (RP); hence only histological confirmation of the diagnosis of prostatic ma-lignancy is virtually needed. There is scant lit-erature16 dealing with the possibility to reduce the PB protocols when an extensive sampling is not strictly necessary in order to spare the invasiveness of the procedure and its costs. We hypothesized that the extensive firstline PB is redundant in the patients with pre-sumed high tumour burden, and that the quadrant PB (QPB) can fulfil the task of preopera-tive pathological evaluation in the patients likely to have NOC PC. The purpose of the present study was to investigate whether the reduction of core number from 5-12 to 4 in the patients in whom PSA and/or clinical evaluation indicate high likelihood to have NOC PC, would impair the diagnostic com-petence of PB through missing clinically relevant information usually obtained by this procedure. Patients and methods Patients We retrospectively studied 84 consecutive pa-tients (mean age 71.8 years, range 50-89) in whom systematic PB was performed during one-year period. The men were previously untreated for PC and biopsied for the first time. The patients were classified into two study groups according to serum PSA, DRE and TRUS findings, the factors which can Table 1. Selection criteria for stratifying the patients into two categories according to probability of the presence of advanced (non-organ confined) PC. Number of patients in each sub-category is given. Likelihood for the PSAa level TRUSb and DREc finding Number of presence of advanced PC patients Low <4 ng/mL TRUS suspect or DRE suspect 5 Low 4-10 ng/mL TRUSP non-suspect and DRE non-suspect 24 High <4 ng/mL TRUSP suspect and DRE suspect 4 High 4-10 ng/mL TRUSP suspect and/or DRE suspect 12 High >10 ng/mL Irrespective of TRUS and DRE-finding 39 Total 84 a PSA test (Elecsys 1010, Roche Diagnostics GmbH, Manheim, Germany) was done prior to any prostate manipulation, to avoid false positive findings; no patients in our series had acute prostatitis (possible cause of elevated PSA); mean prostate size was similar in groups H and L. b TRUS was considered suspicious of malignancy if hypoechogenic sector or nodule in peripheral zone was de-tectable, if the prostate was inhomogeneous without zonal discrimination, or if unsharp prostate margins or infiltration of extraprostatic tissues was seen. c DRE was considered suspicious of malignancy if considerable irregularity of the prostate surface, »rocky hard« induration/nodule or considerable asymmetry is detected on palpation. Abbreviations: PC = prostate cancer, PSA = prostate-specific antigen, TRUS = transrectal ultrasound, DRE = digital rectal examination Radiol Oncol 2005; 39(1): 37-47. Brnić z et al / Quadrant biopsy in men with advanced prostate cancer 39 predict the PC burden in a patient.12-15 The group more likely to have NOC PC (high tu-mour burden) is assigned »H«, while other patients (with presumed low tumour burden) are classified in »group L«. The selection cri-teria for the groups are listed in Table 1. Prostate biopsy protocol We perform US-guided PB by transrectal ap-proach, routinely taking 6-8 tissue cores from <50 cm3 of prostate glands, and 8-12 cores from >50 cm3 of glands at the first-line PB. Six cores are taken from the very lateral parts of peripheral zone at the base, mid-gland and apex bilaterally, followed by additional cores from the posterolateral parts of peripheral zone, similarly to protocols used in the stud-ies.17,18 The number of cores intended to be taken in a particular patient is dependent ex-clusively upon the prostatic size, irrespective of the parameters of suspicion for PC. However, we occasionally reduce the number of cores ad hoc if bleeding from haemorrhoids occurs, or on the patient’s demand due to pain. We usually obviate more medial cores, as also those less expected to be positive.1,2 As a consequence of such approach, the material in the present study consists of 5-12 cores per biopsy session. specimens was unaware of the aims of this study. Methods Pathological report for the entire set of PB samples (5-12 cores) was available for each patient. We verified each individual tissue core whether it was positive for PC, and whether high-grade prostatic intraepithelial neoplasia (HGPIN) was present. Gleason score (GS) was determined on the basis of the complete 5-12 PB set. With the evidence of pathohistology of the complete set of biopsies (5-12 cores) for each patient, we simulated the situation as if only 4 biopsy cores would have been taken (quadrant PB, QPB). In the hypothetical biopsy scheme the apical and medial cores were eliminated. For each patient, we retrospectively compared the patho-logical results of the actual complete 5-12 PB with the presumed results of QBP. The two compared sampling schemes are shown in Figure 1. The following relevant pathologic parameters were considered in comparison of the two PB schemes: presence of PC in the prostate, presence of HGPIN and percentage of positive cores. Additionally, for the purpose of the study only, the same pathologist (G.S.), unaware of previously reported GS, Equipment and technique HP ImagePoint ultrasound system (Hewlett-Packard Company, Andover, MA, USA) with 5.0-7.5 MHz sector endorectal probe and plastic biopsy needle guide was used to assist PB, performed with spring-loaded Bard Magnum device (Bard Urological Division, Covington, GA, USA) coupled with 18-G-nee-dles. A new needle was taken for every 3 - 4 tissue cores. Biopsy cores from different sites of the prostate were submitted for analysis in individually labelled separate containers,19 and core sites were charted on a dedicated form. Pathologist (G.S.) who analysed the Figure 1. The distribution of biopsy sites in the two compared sampling schemes: o + x = 5-12 PB scheme, x = QPB scheme. Radiol Oncol 2005; 39(1): 37-47. 40 Brnić z et al / Quadrant biopsy in men with advanced prostate cancer determined GS for each patient on the basis of the set of 4 cores, which matched the sites of QBP. GS was then compared with that de-termined from 5-12 PB. T-test was used in sta-tistical analysis. Oral and written informed consent was ob-tained from each patient before PB, and information on possible complications of sys-tematic PB was given. Our study did not in-fluence the patient management in any way, as QPB was only an imaginatively rather than really applied procedure. Local Medical Ethics Committee approved this investiga-tion. Results The mean age of patients was 71.8 years (range, 50-89), and did not differ significantly between the groups H and L (71.2 vs. 72.9 years, p=0.86). The mean PSA for the whole series was 36.3 ng/mL (range, 0.03-346); sig-nificant difference (p=0.0002) was observed between group H (52.9 ng/mL, range 1-346) and group L (5.8 ng/mL, range 0.03-9). Prostate volume ranged 16-192 cm3, and did not differ significantly between the groups H and L (63.2 vs. 63.7 cm3, p=0.19). GS deter-mined from the 5-12 BP and QPB material are given in Table 2. Positive correlation between GS and PSA (c=0.39), and between GS and the percentage of positive cores (c=0.53) was shown in the 5-12 PB material. GS deter-mined from the QPB material did not differ significantly from GS determined from the material of 5-12 PB, either for the whole se-ries, or for each particular study groups H (p=0.13) and L (p=0.12), with a maximum in-dividual difference of 2 points. In one L-pa-tient, GS=4 was overgraded as GS=6 from the QPB material, while in one H-patient GS=8 was undergraded as GS=6 also from the QPB material. In 22/46 (47.8%) patients, GS de-fined by both PB sets was identical, while in 20/46 (43.5%), it was undergraded, and in 2/46 (4.4%) overgraded by 1 point by QPB. Overall results of 5-12 PB and QBP for groups H and L are shown in Tables 3 and 4. Total number of cores taken from 84 patients was 605. Median number of cores per PB was 8 (range 5-12). Of all cores, 54.5% were positive for PC: 69.3% in group H, and 14.1% in group L. The percentage of positive cores in the 5-12 PB and QPB material in both compared study groups are shown in Tables 5 and 6. In 19/84 (22.6%) patients, all cores in the 5-12 PB material were positive for malignancy (1 patient in group L, 18 patients in group H). The num-ber of patients with PC detected in only one tissue core in both PB materials is shown in Tables 3 and 4. Pathological results allowing for the following parameters, presence of PC, presence of HGPIN, and percentage of positive cores for two different PB schemes are shown in Tables 5 and 6. The data, which Table 2. Gleason scores determined from 5-12 PB and QPB material. Gleason score 5-12 PB QPB Difference observed Overall series median 6 median 6 0 mean 6.59 mean 6.32 p=0.13 (NS) range 3-9 range 3-9 Group H median 7 median 6 -1 point mean 6.79 mean 6.36 p=0.13 (NS) range 4-9 range 3-9 Group L median 4 median 5 +1 point mean 4.50 mean 5.50 p=0.12 (NS) range 3-5 range 4-6 Abbreaviations: PB=prostate biopsy, QPB=quadrant prostate biopsy, NS=non-significant Radiol Oncol 2005; 39(1): 37-47. Brnić z et al / Quadrant biopsy in men with advanced prostate cancer 41 would have been missed if only QPB were done, are given in the last column. Discussion The two fundamental shortages of systematic PB - its sampling error and invasiveness, lie in reciprocity: sampling extensiveness decreas-es sampling error at the cost of higher patient discomfort and postbiopsy morbidity. The recent tendency to increase the number of cores per one PB session is based on the evidence that an extensive sampling yields a higher PC detection rate and staging accura-cy.2-4,6-8,17,19,20 To balance the diagnostic yield and risk, the PB protocol needs to be individ-ualized for each patient according to his PSA level, TRUS and DRE findings, prostate volume, age and life expectancy. In the present Table 3. Overall results of 5-12 PB outcome: the number of patients in each study group with respect to the pres-ence of PC and PIN Patient group 5-12 PB positive for PC >1 core positive 1 core positive 5-12 PB negative for PC Total no PIN HGPIN+LGPIN L H Total 3 43 46 10 3 13 4+7 5+3 9+10 29 55 84 Abbreviations: PB=prostate biopsy, PC=prostate cancer, LGPIN = low-grade prostatic intraepithelial neoplasia, HGPIN = high-grade prostatic intraepithelial neoplasia Table 4. Overall results of QPB outcome: the number of patients in each study group with respect to the presence of PC and PIN Patient group QPB positive for PC >1 core positive 1 core positive QPB negative for PC Total no PIN HGPIN+LGPIN L H Total 1 42 43 15 6 21 2+5 5+2 7+7 27 57 84 Abbreviations: PB=prostate biopsy, PC=prostate cancer, LGPIN = low-grade prostatic intraepithelial neoplasia, HGPIN = high-grade prostatic intraepithelial neoplasia Table 5. Comparative results of different PB schemes in group H Parameter analysed Presence of PC Presence of PIN Percentage of positive cores 5-12 PB QPB Missed with QPB 44 5 HGPIN 3 LGPIN 69.3% 44 5 HGPIN 2 LGPIN 63.1% 0 0 1 LGPIN 6.2%, p=0.39 (NS) Abbreviations: PB=prostate biopsy, QPB=quadrant prostate biopsy, HGPIN=high-grade prostatic intraepithelial neoplasia, LGPIN=low-grade prostatic intraepithelial neoplasia, NS=non-significant Table 6. Comparative results of different PB schemes in group L Parameter analysed 5-12 PB QPB Missed with QBP presence of PC 8 presence of PIN 4 HGPIN 7 LGPIN percentage of positive cores 14.1% 5 2 HGPIN 5 LGPIN 9.8% 3 2 HGPIN 2 LGPIN 4.3%, p=0.04 Abbreviations: PB=prostate biopsy, QPB=quadrant prostate biopsy, HGPIN=high-grade prostatic intraepithelial neoplasia, LGPIN=low-grade prostatic intraepithelial neoplasia, NS=non-significant Radiol Oncol 2005; 39(1): 37-47. 42 Brnić z et al / Quadrant biopsy in men with advanced prostate cancer study, we focused our attention to the pa-tients with high laboratory and clinical suspi-cion of advanced PC. Despite the presump-tive diagnosis of PC, most of these men have to undergo PB to obtain the tissue diagnosis before treatment with androgen ablation. A very accurate staging is mostly not critical for the therapy, which is rarely radical. Extensive sampling protocols do not seem to be reason-able in the first-line BP in such patients, be-cause the confirmation of prostate malignan-cy and orientation on tumour biology could be reached even with few biopsy cores, and unnecessary discomfort, risks and costs may be avoided by such an approach. Cancer detection rate (sensitivity) can de-crease due to PB scheme reduction for two reasons: overall sampling density reduction and eliminating the gland areas, in which PC is frequently located, from sampling. The im-pact of sampling density on the sensitivity of PB is well known.2-5,17,20 It is particularly ex-pressed in the patients with negative DRE and TRUS, and PSA<10 ng/mL,3,4 correspon-ding to patient population similar to our group L. Thus, the strategy of reducing PB protocol does not seem convenient for the men with presumed low tumour burden because only the extensive sampling provides a proper sensitivity for the early detection of a po-tentially curable malignancy.4 This is concor-dant to our results: using QBP in group L, 3 of 8 PC would have remained undetected, which is a considerable drop of sensitivity. How-ever, in the men with presumed high tumour bur-den, the risk that PC will remain undetected with limited number of cores is little because their tumour is probably not small, and can-cer-free areas in prostate are less likely to ex-ist. Moreover, as many PC are predominantly infiltrating rather than only expansive, even the finding of cancer-free cores does not war-rant that this part of the prostate is not in-volved. Our results reassured these assump-tions: QPB would detect malignancy in all 44 men with PC detected with 5-12 PB protocol in group H, with neither significant nor in-significant PC missed due to sampling scheme reduction. Other authors similarly showed that PC detection rate is less affected by the core number increase in the patients with PSA>10 ng/mL, while significantly im-proved in those with PSA<10 ng/mL.1,3,20 Aus et al. showed that the reduction of sextant PB protocol to QPB resulted in the decrease of sensitivity for PC by only 4% in the patients with elevated PSA and positive DRE and/or TRUS.1 Damiano et al have recently demon-strated that the reduction of 14- to 8-cores regimen resulted in only 3.1% lower PC de-tection rate, and concluded that 8-cores PB may be appropriate as initial PB for general male population.16 A question arises whether, in the series larger than our, some H-patients positive for PC on 5-12 PB would have appeared negative on QPB. As a rule, any patient with high sus-picion of PC and negative initial PB have to be rebiopsied, and rebiopsy need to be more extensive than the first-line PB.2 In such a way, a part of false-negative H-patients on QPB will be correctly detected as positive. Thus, adhering to QPB as the first-line BP in H-pa-tients, we spare discomfort and costs in at least 98% of positives on initial QBP, paying the price of rebiopsy in <2.2% (theoretically 1 of ?45) false negatives on QPB. This »price« is considerably lower than unnecessary extensive sampling in every H-patient, which yields no clinically relevant information for these patients. Excluding different prostate areas from sampling will yield in the same sensitivity de-crease because the likelihood to be an origin of PC varies. We eliminated medial biopsies in our reduced PB because the medial cores are less frequently positive for PC than the lateral ones,1,2,4,17 and the lateral parts of pe-ripheral zone can be sampled by transrectal approach more efficiently than the medial parts. As the biopsy needle passes more or-thogonally across the posteromedial periph- Radiol Oncol 2005; 39(1): 37-47. Brnić z et al / Quadrant biopsy in men with advanced prostate cancer 43 eral zone and more longitudinally through the lateral parts of peripheral zone, the lateral tissue cores almost completely include pe-ripheral zone, while the medial cores usually include also a considerable part of transition-al zone. Finally, it is our impression that medial passes cause bleeding more frequently than lateral ones and that they are more painful. It seems harder to argument apical biopsies elimination from the first-line PB protocol. Quite a lot of PC are localized near the midline at the prostate apex,2,17 which may remain undetected after the exclusion of apical biopsies from the first-line PB. Two apical biopsies added to 2 middle lobar lateral biopsies increase the sensitivity by 13%,1 and most tumours missed on the initial BP were located just in apico-dorsal region.2 Nevertheless, the apex-directed PB have su-perior sensitivity compared with the sextant PB in the patients with PSA<10 (comparable to our group L), but the sensitivity was lower than in the sextant PB in the patients with PSA>10 (comparable to our group H).21 This may indicate that the sampling of the apex is unavoidable only for the detection of early stage PC, while less critical when an advanced PC is more probable. Therefore, when searching for PC in general population or population similar to our group L, it would not be advisable to obviate apical cores. In H-patients, however, even if originated in the apex, PC would probably have infiltrated into the majority of the gland, with positive basal and mid-gland cores, and would possibly be NOC. Obviating the apical cores in such pa-tients would consequently not be critical for PC detection rate, as we have confirmed in our results. »One-core« prostate tumours. In men with limited life expectancy, it is important to determine whether T1c PC is clinically signifi-cant and needs treatment at all. As tumour significance is related to its volume (>0.5 mL), hence to the number of cores that contain neoplastic tissue,22 PC detected as only one positive core may be insignificant. A dilemma arises of how many biopsies should be per-formed to increase the overall PC detection rate without over-diagnosing clinically in-significant neoplasms,5,22 and whether a less extensive sampling decreases that risk. In our group L, 5 of 8 tumours were detect-ed as »one-core tumours« on 5-12 PB. On QPB, 1 of 5 »one core CP« would have been missed. In low range PSA patients, many PCs are detected by chance, being not responsible for patient’s clinical presentation, and missing such an insignificant PC is not detrimental, particularly if PC tumour is of low aggressive-ness. However, in group L, 2 of 3 significant CP would have also been missed on QPB - a considerable drop of overall PC detection rate. In group H, the number of »one-core PC« increased from 1 to 2. One PC with GS=8, which was detected in 2 adjacent cores on 5-12 PB, would have become »one-core PC« on QPB. This patient would have been managed similarly, irrespective of the number of positive cores, due to its high GS. Even if some insignificant PC would remain undetected on QPB in group H (if our series were larger), this would not be a serious shortcoming as such small PC is not likely to be responsible for clinical and laboratory presentation, which, indeed, prompted PB in group H. Such a small PC could be detected in many H-patients on an extensive rebiopsy, and its sig-nificance would be estimated from the com-plete set of cores. Organ confinement of the tumour. The ability of pre-treatment variables to identify the pa-tients with organ-confined PC (OCPC) is a challenging issue. The presence of extrapro-static extension (EPE) is a feature of T3-stage, unfavouring radical treatment. The tumour volume is an important independent predict-ing parameter of the margin status and dis-ease progression after RP, and underestima-tion of tumour volume may result in overindi-cation of RP. Number, percentage and bilat-erality of positive cores in PB are valuable Radiol Oncol 2005; 39(1): 37-47. 44 Brnić z et al / Quadrant biopsy in men with advanced prostate cancer predictors of tumour volume, EPE and prog-nosis.5-8,19 Ipsilateral EPE is more likely, as the number of positive biopsies on that side increases, while the patients with >3 and bi-laterally positive cores had greater likelihood of EPE.8 It was demonstrated that such quantitative histology data are especially valuable in the men with presumed low tumour burden (similar to our group L), thus better predict-ing the final pathological stage.19 Therefore, the information on the percentage of positive cores in PB must not be sacrificed in any re-duced sampling scheme, particularly in group L in which RP is often considered as a treat-ment option. In only one patient in group H, PC was de-tected with one instead of ?2 positive cores, and in only one patient with 2 instead of 3 positive cores, as a consequence of scheme reduction to QPB. This does not preclude the use of QPB as the first-line PB scheme in H-patients. Grossklaus et al. compared <6 vs. >6 cores PB and concluded that the reduction of core number could impair the PC detection rate, but not other information, particularly the percentage of positive cores and bilaterality of PC.5 In our study, the percentage of positive cores decreased significantly (14.1% to 9.8%, p=0,04) in group L, but insignificantly in the whole series (54.5% to 45.1%, p=0.17), and in group H (69.3% to 63.1%, p=0.39), due to the limitation to QPB. Maximum individ-ual differences in the percentage were 20% and 25% in two H-patients, respectively. Therefore, considering the parameter »per-centage of positive cores«, the use of QPB as the first-line PB scheme is not appropriate in group L, while acceptable in group H. Although the conclusions by Grossklaus et al5 and ours are similar for the overall series, the patient populations are not quite compa-rable as Grossklaus et al studied two different groups of men with different sampling schemes, while we compared two PB schemes on the same bioptic material. The association of high-grade prostatic in-traepithelial neoplasia (HGPIN) on PB specimen with concurrent invasive PC next to it or else-where in the gland is evident,23 and a signifi-cant proportion of patients with HGPIN de-tected on initial PB will be found to have PC on repeat PB.24,25 Thus, the identification of HGPIN on PB is an imperative as it may prompt further search for coexistent or sub-sequent invasive PC in the patient.23 The number of cores with HGPIN was an inde-pendent predictor of the risk for PC.25 It was shown that the extensive »five-region PB« de-tected significantly more HGPIN compared to the sextant BP.24 Thus, the reduction of core number in PB can decrease the HGPIN detection sensitivity, which can in conse-quence decrease the CP detection rate. The detection of HGPIN on the first-line PB is an imperative, particularly in L-patients, in whom the finding of HGPIN may be decisive for rebiopsy; had this information been missed, a number of rebiopsies would not have been ordered and early PCs could have remained undetected. QPB would miss 2 HG-PIN lesions in 21 PC negative L-patients (9.5%); thus, QPB is not appropriate as the first-line PB regimen in L-patients. On the contrary, none HGPIN was missed with QPB in group H. In larger H-population, some HG-PIN theoretically could have been missed, but this lack would not have been critical, as every PC-negative H-patient has to be rebiop-sied also for reasons other than HGPIN, mainly for persistent clinical and biochemical suspicion. Proper estimation of Gleason score (GS) from PB specimens is essential in making treatment decision as high GS precludes rad-ical treatment even when CP seems to be or-gan-confined.26 GS determined from a PB specimen may be discordant to that deter-mined from a surgical specimen.27-30 GS as-signed to PB material were identical to RP specimen in 51-67% of cases, greater in 4-15%, and lower in 22-54%. The magnitude of Radiol Oncol 2005; 39(1): 37-47. Brnić z et al / Quadrant biopsy in men with advanced prostate cancer 45 discrepancy was directly related to the quan-tity of tissue in PB specimen, being greater among specimens with GS<7 than among those with higher GS.27,28 GS defined by 18-cores BP speciment exactly matched that of surgical specimen in 37-57% of cases,27,29 be-ing within the interval of +1 point in 93% of cases.29 Undergrading is particularly precari-ous as it may lead the clinician to underesti-mate falsely the true biological potential of PC and to proceed to RP in the patient with great likelihood to have NOC PC; of most concern are the patients with GS>6 detected as GS<6 on QPB. As predisposing factors for errors in histological grading by needle PB were limited core length and limited number of biopsy cores.30 PB is to be repeated when low-grade PC was initially diagnosed on only limited quantities of neoplastic tissue to reduce the risk of underestimation the GS. In our study, GS of group H was not significantly in-fluenced by the core number limitation to 4, but the accuracy was decreased in group L (more low grade PC-higher grading error ac-cording to).28 In group H, in which less pa-tients may be 50 ng/mL indicates PC with a positive prediction value of 98.5%.12 Another advantages of re-duced PB are higher safety for performance on an out-patient basis, less patients’ anxiety for future PB, lower time consumption and workload to pathologists, lower costs (one needle per patient), and lower risk of seeding tumour cells. Our study may have limitations. Although PSA level and DRE finding may, to some ex-tent, indicate statistical risk of PC in a de-fined population,2-14 our study groups H and L were defined arbitrarily, with the aim of sorting out patients with significantly differ-ent likelihood to have PC. We can estimate that the likelihood of L- and H-patients to have PC is <30% and >60%, respectively.13 These rates are, however, only for orienta-tion, as our classification does not match strictly the criteria.13,14 Thus, H- vs. L- classi-fication is a provisional tool for rapid estima-tion of the likelihood for the presence of PC and tumour burden, but not an attempt to Radiol Oncol 2005; 39(1): 37-47. 46 Brnić z et al / Quadrant biopsy in men with advanced prostate cancer stage the tumour or give a prognosis. It is aimed only to serve for identifying the pa-tients who might benefit from PB reduction. We conclude that in the patients with high likelihood to have NOC PC, the reduction of the number of cores does not impair the overall sensitivity, and minimally changes the staging accuracy. In the patients likely to have an early PC, the reduction of the num-ber of cores significantly impairs the overall sensitivity. QPB can be an appropriate firstline sampling scheme in H-patients, as the information lost due to the core number reduc-tion is mainly not critical for the patient management; a more extensive PB is necessary for other patients, for proper sensitivity and stag-ing accuracy. References 1. Aus G, Bergdahl S, Hugosson J, Lodding P, Pihl CG, Pileblad E. Outcome of laterally directed sextant biopsies of the prostate in screened males aged 50-66 years. Implications for sampling order. Eur Urol 2001; 39:655-61. 2. Chon CH, Lai FC, McNeal JE, Presti JC Jr. Use of extended systematic sampling in patients with a prior negative prostate needle biopsy. J Urol 2002; 167:2457-60. 3. Nava L, Montorsi F, Consonni P, Scattoni V, Guazzoni G, Rigatti P. Results of a prospective randomised study comparing 6, 12, 18 transrectal ultrasound guided sextant biopsies in patients with elevated PSA, normal DRE and normal pro-static ultrasound. J Urol 1997; 157(Suppl): 59A. 4. Babaian RJ, Toi A, Kamoi K, Troncoso P, Sweet J, Ewans R, et al. A comparative analysis of sextant and an extended 11-core directed biopsy strategy. J Urol 2000; 163: 152-7. 5. Grossklaus DJ, Coffey CS, Shappell SB, Jack GS, Cookson MS. Prediction of tumour volume and pathological stage in radical prostatectomy speci-mens is not improved by taking more prostate needle-biopsy cores. BJU Int 2001; 88: 722-6. 6. Grossklaus DJ, Coffey CS, Shappell SB, Jack GS, Chang SS, Cookson MS. Percent of cancer in the biopsy set predicts pathological finding after prostatectomy. J Urol 2002; 67: 2032-5. Radiol Oncol 2005; 39(1): 37-47. 7. Borboroglu PG, Amling CL. Correlation of positive sextant biopsy locations to site of positive surgical margins in radical prostatectomy specimens. Eur Urol 2001; 39: 648-53. 8. Tigrani VS, Bhargava V, Shinohara K, Presti JC Jr. Number of positive systematic sextant biopsies predicts surgical margin status at radical prostate-ctomy. Urology 1999; 54:689-93. 9. Djavan B, Waldert M, Zlotta AR, Dobronski P, Seitz C, Remzi M, et al. Safety and morbidity of first and repeat transrectal ultrasound-guided prostate needle biopsies: results of the prospective European Prostate Cancer Detection Study. J Urol 2001; 166: 856-60. 10. Peyromaure M, Ravery V, Messas A, Toublanc M, Boccon-Gibod L. Pain and morbidity of an extensive prostate 10-biopsy protocol: a prospective study of 289 patients. J Urol 2002; 167: 218-21. 11. Naughton CK, Ornstein DK, Smith DS, Catalona WJ. Pain and morbidity of transrectal ultrasound guided prostate biopsy: a prospective randomized trial of 6 versus 12 cores. J Urol 2000; 163: 168-71. 12. Gerstenbluth RE, Seftel AD, Hampel N, Oefelein MG, Resnick MI. The accuracy of the increased prostate specific antigen level (greater than or equal to 20 ng./ml.) in predicting prostate cancer: is biopsy always required? J Urol 2002; 168: 1990-93. 13. Potter SR, Horniger W, Tinzl M, Bartsch G, Partin AW. Age, prostate-specific antigen, and digital rectal examination as determinants of the proba-bility of having prostate cancer. Urology 2001; 57: 1100-4. 14. Eastham JA, May R, Robertson JL, Sartor O, Kattan MW. Development of a nomogram that predicts the probability of positive prostate biopsy in men with an abnormal digital rectal examina-tion and a prostate-specific antigen between 0 and 4 ng/mL. Urology 1999; 54: 709-13. 15. Yamamoto T, Ito K, Ohi M, Kubota Y, Suzuki K, Fukabori Y, et al. Diagnostic significance of digital rectal examination and transrectal ultrasonogra-phy in men with prostate-specific antigen levels of 4 ng/mL or less. Urology 2001; 58: 994-8. 16. Damiano R, Autorino R, Perdona S, De Sio M, Oliva A, Esposito C et al. Are extended biopsies really necessary to improve prostate cancer detec-tion? Prostate Cancer Prostatic Dis 2003; 6: 250-5. 17. Gore JL, Shariat SF, Miles BJ, Kadmon D, Jiang N, Wheeler TM, et al. Optimal combinations of sys-tematic sextant and laterally directed biopsies for the detection of prostate cancer. J Urol 2001; 165: 1554-9. Brnić z et al / Quadrant biopsy in men with advanced prostate cancer 47 18. Ravery V, Goldblatt L, Royer B, Blanc E, Toublanc 30. Ruijter E, van Leenders G, Miller G, Debruyne F, M, Boccon-Gibod L. Extensive biopsy protocol im- van de Kaa C. Errors in histologic grading by pro-proves the detection rates of prostate cancer. J static needle biopsy specimens: frequency and Urol2000; 164: 393-6. predisposing factors. J Pathol2000; 192: 229-33. 19. Tombal B, Tajeddine N, Cosins JP, Feyaerts A, Opsomer R, Wese FX, et al. Does site-specific la-belling and individual processing of sextant biop-sies improve the accuracy of prostate biopsy in predicting pathological stage in patients with T1c prostate cancer? BJU Int 2002; 89: 543-8. 20. Aus G, Ahlgren G, Hugosson J, Pedersen KV, Rensfeldt K, Soderberg R. Diagnosis of prostate cancer: optimal number of prostate biopsies relat-ed to serum prostate-specific antigen and findings on digital rectal examination. Scand J Urol Nephrol 1997; 31: 541-4. 21. Brossner C, Madersbacher S, Bayer G, Pycha A, Klingler HC, Maier U. Comparative study of two different TRUS-guided sextant biopsy techniques in detecting prostate cancer in one biopsy session. Eur Urol 2000; 37: 65-71. 22. Furuya Y, Fuse H, Nagakawa O, Masai M. Preoperative parameters to predict tumor volume in Japanese patients with nonpalpable prostate cancer. Int J Clin Oncol 2002; 7: 109-13. 23. Ellis WJ, Brawer MK. Repeat prostate needle biop-sy: who needs it? J Urol 1995; 153: 1496-8. 24. Rosser CJ, Broberg J, Case D, Eskew LA, McCullough D. Detection of high-grade prostatic intraepithelial neoplasia with the five-region biop-sy technique. Urology 1999; 54: 853-6. 25. Kronz JD, Allan CH, Shaikh AA, Epstein JI. Predicting cancer following a diagnosis of highgrade prostate intraepithelial neoplasia on needle biopsy: data on men with more than one follow-up biopsy. Am J Surg Pathol 2001; 25: 1079-85. 26. Manoharan M, Bird VG, Kim SS, Civantos F, Soloway MS. Outcome after radical prostatectomy with a pretreatment prostate biopsy Gleason score of >/=8. BJU Int 2003; 92: 539-44. 27. Djavan B, Kadesky K, Klopukh B, Marberger M, Roehrborn CG. Gleason scores from prostate biopsies obtained with 18-gauge biopsy needles poorly predict Gleason scores of radical prostatec-tomy specimens. Eur Radiol 1998; 33: 261-70. 28. San Francisco IF, DeWolf WC, Rosen S, Upton M, Olumi AF. Extended prostate needle biopsy im-proves concordance of Gleason grading between prostate needle biopsy and radical prostatectomy. J Urol 2003; 169: 136-40. 29. King CR, Long JP. Prostate biopsy grading errors: a sampling problem? Int J Cancer 2000; 90: 326-30. Radiol Oncol 2005; 39(1): 37-47.