Biochemical recurrence rate in patients with positive surgical margins at radical prostatectomy with further negative resected tissue

Transcription

1 . JOURNAL COMPILATION 2009 BJU INTERNATIONAL Urological Oncology BIOCHEMICAL RECURRENCE RATE WITH POSITIVE SURGICAL MARGINS AT RP WITH NEGATIVE RESECTED TISSUE RABBANI et al. BJUI BJU INTERNATIONAL Biochemical recurrence rate in patients with positive surgical margins at radical prostatectomy with further negative resected tissue Farhang Rabbani, Kinjal C. Vora, Luis Herran Yunis, James A. Eastham, Bertrand Guillonneau, Peter T. Scardino and Karim Touijer Urology Service, Department of Surgery, Memorial Sloan-Kettering Cancer Center, New York, NY, USA Accepted for publication 5 June 2009 Study Type Therapy (case series) Level of Evidence 4 OBJECTIVE To determine the biochemical recurrence (BCR) rate in patients with positive surgical margins (PSMs) on the prostate specimen who have additional negative tissue resected from that site (M+ ), compared to patients with negative margins (M ) and those with persistent PSM (M+), as those with PSM at radical prostatectomy (RP) are at greater risk of BCR, and in some instances where suspicious tissue is noted in the prostate bed or when frozensection analysis shows PSM, additional tissue is resected from the suspect site of the PSM. PATIENTS AND METHODS Between January 1999 and June 2007, 4217 consecutive patients underwent open or laparoscopic RP with no previous radiotherapy or hormonal therapy. The median (interquartile range) follow-up was 37.4 ( ) months. RESULTS Pathological organ-confined (OC) cancer was pre in 2901 men, of whom 2659 had M, 216 had M+, and 26 had M+. Extracapsular extension (ECE) alone with no seminal vesicle or lymph node involvement was pre in 843 men, of whom 657 had M, 174 had M+ and 12 had M+. For patients with OC cancer, the 36-month actuarial BCR-free probability was 97.9% (95% confidence interval ) for M, vs 89.0 ( )% for M+ vs 100% for M+. For patients with ECE, the 36-month actuarial BCR-free probability was 83.7 ( )% for M vs 73.7 ( )% for M+ vs 90.0 ( )% for M+. The main limitation of the study was its retrospective nature, with the reason for resection of additional tissue not always well documented. CONCLUSIONS While the few patients with PSMs and further negative resected tissue limited the statistical analysis, it would appear that in these patients the disease behaves as in those with negative margins. KEYWORDS pathology, prostatectomy, prostatic neoplasms, frozen sections, margin INTRODUCTION In contemporary series, positive surgical margins (PSMs) occur in 11 38% of patients undergoing radical prostatectomy (RP) [1]. PSMs have been associated with a greater risk of biochemical recurrence (BCR), with hazard ratios of in recent series [1]. Given that randomized trials have reported a benefit for adjuvant radiotherapy after RP in patients with PSMs [2,3], it is imperative to accurately define which patients have a true PSM repreing inadequate surgical extirpation of the tumour. While the presence of prostate cancer cells at the inked specimen margin is the generally accepted definition of a PSM in most publications [4 7], some authors have considered patients with negative biopsies of the prostatic bed to have negative SMs (NSMs) [8]. However, there is variability in the definition of a PSM, even among experts. Epstein [9] defined a PSM as occurring in cases where tumour appears to be cut across, extending to the inked margin of resection that has an irregular roughened surface, while cases in which the tumour only focally extends to the inked margin or extends to an inked margin that has a smooth rounded configuration at scanning magnification are considered equivocal. Bong et al. [10] reported that despite similar criteria for evaluating SMs, PSMs and BCR rates in patients with PSMs can vary at different institutions even for the same surgeon. In one of the earliest studies, Ponthieu et al. [11] reported that a PSM was discovered in eight of 66 patients who had intraoperative frozen-section analysis (FSA), in six of whom the PSM was rendered negative by further resection of periprostatic tissue. They reported that the PSA level after RP was reduced in the six patients whose margins had been rendered negative; however, seven of the eight patients received adjuvant radiotherapy. However, despite studies [12 14] reporting the ability of intraoperative FSA to lower the PSM rate, there are limited data on the prognostic significance of rendering a PSM negative by resection of further periprostatic tissue, in any large study. We sought to determine if patients whose PSMs are rendered negative with further resection of periprostatic tissue JOURNAL COMPILATION 2009 BJU INTERNATIONAL 104, doi: /j x x 605

2 RABBANI ET AL. have a decreased BCR rate, more consistent with those patients with NSMs. PATIENTS AND METHODS Between January 1999 and June 2007, 4600 consecutive patients underwent RP (open retropubic, RRP, or laparoscopic, LRP) at our institution. We excluded 383 patients who had previous radiotherapy (104), chemotherapy (51) or hormonal therapy (264) or in whom no pathological Gleason score was assigned due to treatment effects seen on the RP specimen or stage pt0 (174). The remaining 4217 patients are the subject of this study. Institutional Review Board approval was obtained for the study. The pathology and operative reports of all patients with PSMs were reviewed. In 98 patients who had PSMs on the specimen, further periprostatic tissue was resected from the prostatic bed due to concern about residual cancer on visual inspection of the prostate specimen or prostatic bed, or in cases where FSA of the prostate specimen during RP revealed a positive or close margin. All surgical specimens were processed with 3 5 mm step-sectioning, with the entire prostate submitted. A PSM was defined as tumour cells pre at the inked margin. NSMs are designated M, PSMs as M+, and PSMs on the prostate specimen with the final resected periprostatic tissue being negative at the site of the PSM with no other site of PSMs are designated M+. BCR was defined as a PSA level of 0.10 ng/ml after RP and a confirmatory increase, or a PSA level of 0.10 ng/ml and receiving salvage treatment (e.g. radiotherapy, hormonal manipulation, chemotherapy). The statistical significance of differences in proportions was tested using the chi-square test or Fisher s exact test where appropriate. Kaplan-Meier analysis with the log rank test was used to compare BCR-free probability in those with M vs M+ vs M+, stratifying for pathological stage (organ-confined, OC) vs extracapsular extension alone (ECE) without seminal vesicle or lymph node involvement. The BCR-recurrence free probability was calculated at 36 months with 95% CI calculated using Greenwood s formula. RESULTS Table 1 summarizes the clinical characteristics of the 4217 patients; the median Variable Median (IQR) or n (%) Surgical approach RRP 3218 (76.3) LRP 999 (23.7) Age, years 59.1 ( ) Ethnicity: Caucasian 3738 (88.6) African-American 293 (6.9) Other/unknown 186 (4.4) Body mass index, kg/m 2 * 28.1 ( ) Clinical stage T (62.7) T (34.4) T3 114 (2.7) Tx 8 (0.2) Gleason score: (59.4) (33.2) (7.4) Preoperative PSA level, ng/ml 5.4 ( ) Pathological stage pt (70.4) pt (27.7) pt4 80 (1.9) Nodal status N (88.9) N (4.7) Nx 268 (6.4) OC disease 2901 (68.8) ECE alone 843 (20.0) PSMs Overall 586 (13.9) Apex 219 (5.2) Bladder neck 52 (1.2) Posterior 280 (6.6) Anterior 118 (2.8) Pathological Gleason score (38.3) (54.9) (6.8) RP specimen weight, g 50.0 ( ) (interquartile range, IQR) patient age was 59.1 ( ) years. Table 1 also summarizes the intraoperative and pathological characteristics of the 4217 patients. Pathological OC cancer was pre in 2901 men, of whom 2659 had M, 216 had M+ and 26 had M+. ECE alone with no seminal vesicle or lymph node involvement was pre in 843 men, of whom 657 had M, 174 had M+ and 12 had M+. The breakdown of patients according to pathological stage and margin status is shown in Fig. 1. Further periprostatic tissue was from the prostatic bed in 55 patients TABLE 1 The clinical characteristics before and during RP, and the pathological characteristics *missing in 235 patients; missing in 174 patients. with PSM and OC disease, and 43 with PSM and ECE alone. In these 98 patients, tissue was for routine analysis only in 24 patients and for FSA with or without additional tissue for routine analysis in 74 (Fig. 1). FSA was used for the prostate itself in 26 patients. The site of further tissue resection in the 98 patients is preed in Table 2, with the number of patients rendered M by further tissue resection at that site, stratified by pathological stage. A greater proportion of patients with OC disease and M+ had tissue for FSA vs. 606 JOURNAL COMPILATION 2009 BJU INTERNATIONAL

3 BIOCHEMICAL RECURRENCE RATE WITH POSITIVE SURGICAL MARGINS AT RP WITH NEGATIVE RESECTED TISSUE FIG. 1. Analysis of the 4217 patients according to pathological stage and margin status. All patients N = 4217 Positive margins N = 586 Negative margins N = 3631 Organ-confined N = 242 ECE alone N = 186 SVI or LNI or pt4 N = 158 Organ-confined N = 2659 OC M- Further tissue N = 55 No further tissue N = 187 Further tissue N = 43 No further tissue N = 143 ECE alone N = 657 ECE M- = negative OC M+, N = 26 (2 routine, 24 frozen) positive or other PSM N = 29 (11 routine, 18 frozen) = negative ECE M+, N = 12 (2 routine, 10 frozen) positive or other PSM N = 31 (9 routine, 22 frozen) SVI or LNI or pt4 N = 15 on prostate N = 10 on prostate N = 3 OC M+ N = 216 on prostate N = 8 on prostate N = 5 ECE M+ N = 174 TABLE 2 The site of further tissue resection by pathological stage Group Site Bladder neck NVB/PL margin Apex/ urethra DVC/anterior margin OC (55) With further tissue resection, N Rendered M by further resection, 1/13 13 (52) 10 (48) 3/11 n (%) or n/n ECE alone (43) With further tissue resection Rendered M by further resection 0 9 (50) 3/13 0 DVC, dorsal vein complex; PL, posterolateral. routine analysis than had patients with persistent PSMs despite further resected tissue (M+), i.e. 24/26 for OC M+ vs 18/29 for OC M+ (P = 0.008, chi-square test) and 10 of 12 for ECE M+ vs 22/31 for ECE M+ (P = 0.34, Fisher s exact test). In the 29 patients with OC disease with persistent PSMs, 18 had additional tissue for FSA, while 11 had tissue for routine pathology analysis. Fifteen of the 18 OC M+ patients having FSA had tissue from a site other than the PSM, while in three the final tissue was not negative (one with final tissue as routine, one as FSA and one negative on FSA that was positive on permanent section). Nine of the 11 OC M+ patients who had further tissue as a routine section had tissue from a site other than the PSM, while in two the final tissue was not negative. In the 31 patients with ECE with persistent PSMs, 22 had additional tissue for FSA while nine had tissue for routine pathology analysis. Eighteen of the 22 ECE M+ patients having FSA had tissue from a site other than that of the PSM, while in four the final tissue was not negative (one as FSA, three with final tissue as routine). Eight of the nine ECE M+ patients who had further tissue as a routine section had tissue from a site other than the PSM, while in one the final tissue was not negative. For patients with OC cancer, the mean (95% CI) 36-month actuarial BCR-free probability was 97.9 ( )% for M vs 89.0 JOURNAL COMPILATION 2009 BJU INTERNATIONAL 607

4 RABBANI ET AL. ( )% for M+ vs 100% for M+ (Fig. 2A). For patients with ECE, the 36-month actuarial BCR-free probability was 83.7 ( )% for M vs 73.7 ( )% for M+ vs 90.0 ( )% for M+ (Fig. 2B). DISCUSSION Given the adverse prognostic significance of PSMs [1], surgeons strive to achieve NSMs while maximizing sparing of the neurovascular bundles (NVBs) at RP where possible, to achieve the trifecta [15]. Despite the availability of preoperative nomograms to predict ECE [16], surgeons are sometimes confronted with significant desmoplastic reaction around the NVBs at the time of RP, which might be attributed to ECE of cancer [9], leading to NVB resection or a capsular incision in cases of attempts at nerve sparing. BCR rates in patients with capsular incision are higher than in those with OC marginnegative disease or focal ECE with NSMs, and comparable to those in patients with focal ECE with PSM or established ECE with NSMs [17]. As such, in cases of capsular incision into the prostate that are recognized during RP, surgeons often resect additional tissue to achieve a NSM. While several investigators have reported on use of intraoperative FSA to render surgical margins negative [11 14,18 25], there is a paucity of data on the prognostic implications of this approach, with few studies evaluating BCR rates in those with PSMs that have been rendered negative [13,21]. An early study on the use of FSA during RP to evaluate SM status and render otherwise PSMs negative by further resection reported lowering of PSA levels in those whose margins had been made negative [11]. However, this was a small study with 66 patients, in which six of eight patients with a PSM on FSA had further tissue resected until the margin was rendered negative, while the two others did not undergo RP; the assessment of postoperative PSA level was confounded by use of adjuvant radiotherapy in seven of the eight patients, as well as an unstated duration of follow-up. Cangiano et al. [13] reported no cases of BCR in nine patients having a positive posterolateral FSA with subsequent wide resection of the NVB, compared to five of 39 with negative FSA (P = 0.94). Lepor and Kaci [21] reported that all 16 patients with <5% cancer on the final apical margin had an undetectable serum PSA level at a mean follow-up of 19.1 months, compared to one of five of patients having a final apical margin biopsy showing 5% cancer developing BCR at a mean follow-up of 24 months. Kübler et al. [26], reporting on 77 patients undergoing perineal RP with documented invasion of the prostatic apex, found a mean (SD) 36-month BCR-free survival rate of 55.9 (14.9)% for patients with positive apical biopsies compared to 78.7 (5.3)% for those with negative biopsies (P = 0.023). These studies are consistent with the pre study that suggests that the BCR rate is lower in those with PSM that have been rendered negative than in patients with PSM. As such, while these patients with PSM who have been rendered negative are coded as having PSM in RP databases for the purposes of quality assurance and refining the technique of RP, for the purposes of adjuvant therapy, perhaps these patients with PSM rendered negative should be considered to have NSMs. In the pre series, 98 patients who had a PSM on the specimen had further tissue resected; 38 of the 98 (39%) had no cancer in the final tissue resected, with no other site of PSM, with the margins thus being rendered negative with resection of additional tissue. The 98 patients having further tissue resected repre only 2.3% of the entire cohort treated during the period of the study; thus the 38 margins rendered negative repre 0.9% of the entire cohort, effectively reducing the PSM rate in the entire series from 13.9% to 13.0%. The studies reporting on the effect of resection of additional tissue on PSM are summarized in Table 3 [11 14,18 25]. The rate of PSMs rendered negative in the reported studies range from 0.9% in the pre series to 21%, with the relative reduction in PSM of %. Given that the pre study suggests an improvement in BCR in those in whom the PSM is rendered negative, there is a potentially significant reduction in biologically active PSM and secondarily of BCR with a strategy of resection of additional tissue in cases where there is concern about a capsular incision or residual prostate tissue, or in cases where nerve-sparing has been used and FSA of the prostate specimen reveals cancer at the margin. It has been reported that the yield of FSA is too low to justify its routine practice [21,25]. Tsuboi et al. [25] reported that the sensitivity of FSA is 41.8%; only 23 of 55 patients with PSM on the permanent section had the PSM identified on FSA. In the 23 patients with the FIG. 2. The BCR-free probability in patients with: A, OC disease; and B, ECE alone, by margin status. A Biochemical recurrence-free probability B Biochemical recurrence-free probability Patients with Organ-Confined Cancer Margin Status 0.5 OC M+ OC 0.4 M OC M p < Follow-up, months Patients with Extracapsular Extension Margin Status 0.5 ECE M+ ECE 0.4 M ECE M p < Follow-up, months PSM identified on FSA, 17 (74%) were rendered margin-negative with resection of additional tissue, while in the 32 with a PSM, only 14 (44%) were rendered margin-negative with resection of additional tissue, as the remaining 18 had a PSM at a different site from that of the FSA. In their 760 patients, 259 had FSA, to render 31 men marginnegative. Nevertheless, this repres a 29.5% relative reduction in the PSM rate, from 13.8% to 9.7%. In the pre series, as seen in Fig. 1, 34 of 74 (46%) patients who had tissue for FSA were rendered margin-negative with the resection of additional tissue, compared to four of 24 (17%) who had the tissue for routine permanent section (P = 0.011, chi-square test). While this difference is statistically significant, our study only evaluated the effect of resection of further tissue on those with a PSM, rather than a prospective evaluation of the effect on all patients 608 JOURNAL COMPILATION 2009 BJU INTERNATIONAL

5 BIOCHEMICAL RECURRENCE RATE WITH POSITIVE SURGICAL MARGINS AT RP WITH NEGATIVE RESECTED TISSUE TABLE 3 Potential reduction in PSMs with resection of additional tissue at RP in reported series Ref Site of additional resection Type of RP No. of patients Patients with further resection PSM in series PSM rendered negative Adjusted PSM* [11] All RRP (100) 9 (13.6) 6 (9.1) 3 (4.5) 66.7 [12] PL/NVB RRP (13.7) 118 (19.4) 30 (4.9) 88 (14.5) 25.4 [13] PL/NVB RRP (33.8) ns 9 (6.3) ns [14] All (NVB) RRP (42.3) 46 (19.2) 12 (5.0) 34 (14.2) 26.1 [18] PL/NVB LRP (100) 33 (33) 21 (21) 12 (12) 63.6 [19] Bladder neck RRP (100) 13 (25.5) 3 (5.9) 10 (19.6) 23.1 [20] All LRP (100) 25 (12.6) 15 (7.6) 13 (6.6) 60.0 [21] Apex RRP (93.6) ns 19 (3.8) ns [22] Bladder neck RRP (100) ns 2 (2) ns [23] Apex RRP (100) 25 (26.3) 2 (2.1) 23 (24.2) 8.0 [24] All RRP (100) 58 (22.4) ns ns [25] All RRP (34.1) 105 (13.8) 31 (4.1) 74 (9.7) 29.5 Pre All Both (2.3) 586 (13.9) 38 (0.9) 548 (13) 6.5 Relative reduction in PSM ns, not stated; PL, posterolateral; *including further resected tissue. (including those with NSMs) who had undergone resection of additional tissue. Including the patients with NSMs would reduce the percentage of patients who potentially benefit from a strategy of intraoperative FSA. Indeed, Lepor and Kaci [21] reported that the yield of intraoperative biopsy of the bladder neck and NVB/lateral pedicle is too low to justify it in routine practice. In the pre study, about half of the subset of patients with PSM with either OC cancer or ECE alone having further tissue resection in the region of the NVB/ posterolateral margin were rendered marginnegative, as were about half of the subset of patients with PSM and OC cancer having further tissue resection at the apex/urethra. In cases of capsular violation or where there is concern about gross residual tissue, resection of additional tissue might be beneficial, as the BCR rate is improved in patients where the final resected tissue is negative for cancer. Our study has several limitations. As it is a retrospective evaluation, the reason for resection of additional tissue was not always well documented in the operative notes. Second, the study evaluates only patients who had a PSM on the specimen who had resection of additional tissue, and as such is not designed to evaluate the yield or adverse consequences of a strategy of FSA, which is best done prospectively. Thus, while most patients rendered margin-negative had further resection from the NVB/posterolateral margin, or in the case of OC cancer, from the apex/urethra, the denominator of all patients (including those with NSMs on the specimen) was not evaluated in this study to determine the yield of further tissue resection at the various sites, or the consequences in terms of erectile function or incontinence. The relatively few patients who had a PSM and had additional tissue resection limits the statistical analysis, so that while statistical significance was not achieved for the difference in BCR between those with PSM and those with PSM rendered negative, the Kaplan-Meier curves suggested a trend that might be able to be confirmed with a larger study. Insofar as a PSM is used to designate those who have residual cancer in the prostate bed, our study suggests that those with PSMs that have been rendered negative might have disease that behaves like that with NSMs, and perhaps should be considered as such when considering potential adjuvant therapy. The BCR rate in patients with a PSM that has been rendered negative with resection of additional tissue is comparable to that in patients with NSMs and lower than that in patients with persistent PSMs, suggesting a potential benefit to resection of additional tissue in cases where there is a concern of PSM on the RP specimen. ACKNOWLEDGEMENTS Supported by the Sidney Kimmel Center for Prostate and Urologic Cancers. CONFLICT OF INTEREST None declared. REFERENCES 1 Yossepowitch O, Bjartell A, Eastham JA et al. Positive surgical margins in radical prostatectomy: outlining problem and its long-term consequences. Eur Urol 2008; 55: Van der Kwast TH, Bolla M, Van Poppel H et al. Identification of patients with prostate cancer who benefit from immediate postoperative radiotherapy: EORTC J Clin Oncol 2007; 25: Thompson IM, Tangen CM, Paradelo J et al. Adjuvant radiotherapy for pathological T3N0M0 prostate cancer significantly reduces risk of metastases and improves survival: long-term followup of a randomized clinical trial. J Urol 2009; 181: Stamey TA, Villers AA, McNeal JE, Link PC, Freiha FS. Positive surgical margins at radical prostatectomy: importance of the apical dissection. J Urol 1990; 143: Ohori M, Wheeler TM, Kattan MW, Goto Y, Scardino PT. Prognostic significance of positive surgical margins in radical prostatectomy specimens. J Urol 1995; 154: Srigley JR. Key issues in handling and reporting radical prostatectomy JOURNAL COMPILATION 2009 BJU INTERNATIONAL 609

6 RABBANI ET AL. specimens. Arch Pathol Lab Med 2006; 130: Rabbani F, Bastar A, Fair WR. Site specific predictors of positive margins at radical prostatectomy: an argument for risk based modification of technique. J Urol 1998; 160: Tewari A, Srivasatava A, Menon M, Members of the VIP Team. A prospective comparison of radical retropubic and robot-assisted prostatectomy: experience in one institution. BJU Int 2003; 92: Epstein JI. Evaluation of radical prostatectomy capsular margins of resection. The significance of margins designated as negative, closely approaching, and positive. Am J Surg Pathol 1990; 14: Bong GW, Ritenour CW, Osunkoya AO, Smith MT, Keane TE. Evaluation of modern pathological criteria for positive margins in radical prostatectomy specimens and their use for predicting biochemical recurrence. BJU Int 2009; 103: Ponthieu A, Delgrande J, Granger F, Bonneau HP, Ivaldi A. Le contrôle histologique extemporané des marges dans la prostatectomie pour cancer. J Urol (Paris) 1993; 99: Eichelberg C, Erbersdobler A, Haese A et al. for the management of intraoperatively detected palpable tumor lesions during nerve-sparing scheduled radical prostatectomy. Eur Urol 2006; 49: Cangiano TG, Litwin MS, Naitoh J, Dorey F, dekernion JB. Intraoperative frozen section monitoring of nerve sparing radical retropubic prostatectomy. J Urol 1999; 162: Goharderakhshan RZ, Sudilovsky D, Carroll LA, Grossfeld GD, Marn R, Carroll PR. Utility of intraoperative frozen section analysis of surgical margins in region of neurovascular bundles at radical prostatectomy. Urology 2002; 59: Bianco FJ Jr, Scardino PT, Eastham JA. Radical prostatectomy: long-term cancer control and recovery of sexual and urinary function ( trifecta ). Urology 2005; 66: Ohori M, Kattan MW, Koh H et al. Predicting the presence and side of extracapsular extension: a nomogram for staging prostate cancer. J Urol 2004; 171: Chuang AY, Nielsen ME, Hernandez DJ, Walsh PC, Epstein JI. The significance of positive surgical margin in areas of capsular incision in otherwise organ confined disease at radical prostatectomy. J Urol 2007; 178: Fromont G, Baumert H, Cathelineau X, Rozet F, Validire P, Vallancien G. Intraoperative frozen section analysis during nerve sparing laparoscopic radical prostatectomy: feasibility study. J Urol 2003; 170: Nakamura K, Kasraeian A, Anai S, Pendleton J, Rosser CJ. Positive surgical margins at radical prostatectomy: importance of intra-operative bladder neck frozen sections. Int Braz J Urol 2007; 33: Dillenburg W, Poulakis V, Witzsch U et al. Laparoscopic radical prostatectomy: the value of intraoperative frozen sections. Eur Urol 2005; 48: Lepor H, Kaci L. Role of intraoperative biopsies during radical retropubic prostatectomy. Urology 2004; 63: Lepor H, Chan S, Melamed J. The role of bladder neck biopsy in men undergoing radical retropubic prostatectomy with preservation of the bladder neck. J Urol 1998; 160: Shah O, Melamed J, Lepor H. Analysis of apical soft tissue margins during radical retropubic prostatectomy. J Urol 2001; 165: Fasolis G, Degiuli P, Lancia M et al. Periprostatic tissues intraoperative frozen section during retrograde radical retropubic prostatectomy. Arch Ital Urol Androl 2006; 78: Tsuboi T, Ohori M, Kuroiwa K et al. Is intraoperative frozen section analysis an efficient way to reduce positive surgical margins? Urology 2005; 66: Kübler HR, Szukala SA, Madden JF et al. Apical soft tissue biopsies predict biochemical failure in radical perineal prostatectomy patients with apical cancer involvement. Prostate Cancer Prostatic Dis 2007; 10: 72 6 Correspondence: Farhang Rabbani, Urology Service, Department of Surgery, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY, 10065, USA. rabbanif@mskcc.org Abbreviations: BCR, biochemical recurrence; (L)(R)RP, (laparoscopic) (retropubic open) radical prostatectomy; (P)(N)SM, (positive) (negative) surgical margin; NVB, neurovascular bundle; OC, organ-confined; ECE, extracapsular extension; FSA, frozen-section analysis; IQR, interquartile range. 610 JOURNAL COMPILATION 2009 BJU INTERNATIONAL