Predictive factors of late biochemical recurrence after radical prostatectomy

JJCO Japanese Journal of Clinical Oncology Japanese Journal of Clinical Oncology, 2017, 47(3) 233 238 doi: 10.1093/jjco/hyw181 Advance Access Publication Date: 9 December 2016 Original Article Original Article Predictive factors of late biochemical recurrence after radical prostatectomy Takahito Negishi 1, *, Kentarou Kuroiwa 2, Yoshifumi Hori 2, Toshihisa Tomoda 3, Hiroshi Uchino 4, Noriaki Tokuda 4, Nobuki Furubayashi 1, Kei Nagase 1, Hidenori Iwai 1, and Motonobu Nakamura 1 1 Department of Urology, National Hospital Organization Kyushu Cancer Center, Fukuoka, 2 Department of Urology, Miyazaki Prefectural Miyazaki Hospital, Miyazaki, 3 Department of Urology, Oita Prefectural Hospital, Oita, and 4 Department of Urology, Saga-Ken Medical Center Koseikan, Saga, Japan *For reprints and all correspondence: Takahito Negishi, National Hospital Organization Kyushu Cancer Center, 3-3-1, Notame, Minami-ku Fukuoka-shi, Fukuoka 811-1395, Japan. E-mail: negishi.t@nk-cc.go.jp Received 31 August 2016; Editorial Decision 21 November 2016; Accepted 23 November 2016 Abstract Objective: To assess the characteristics of biochemical recurrence in the late period (>5 years after radical prostatectomy) and the differences in the predictors of biochemical recurrence in different periods, we conducted a multicenter retrospective study. Methods: We reviewed 478 men who underwent radical prostatectomy for clinically localized prostate cancer. All of the patients were followed up for at least 5 years. The cohort was then divided into three groups; no recurrence group, recurrence <5 years after surgery group and recurrence 5 years after surgery group. The background characteristics of each group were compared using the χ 2 test. A Cox multivariate regression analysis was performed to determine the predictors of biochemical recurrence in each period. Results: Biochemical recurrence occurred in 135 men. In 113 (84%) of the patients, biochemical recurrence occurred at <5 years after surgery; in 22 (16%), it occurred at 5 years after surgery. The proportion of men with a low preoperative prostate-specific antigen level was significantly larger in the latter group (P = 0.0023). A preoperative prostate-specific antigen level and a positive surgical margin were significant predictors of biochemical recurrence at <5 years after surgery (hazard ratio: 1.03 and 3.20). A positive surgical margin was also a significant predictor of biochemical recurrence at 5 years after surgery (hazard ratio: 3.03); however, a high preoperative prostate-specific antigen level was not. Conclusions: Biochemical recurrence occurred at 5 years after surgery in 16% of the patients. A positive surgical margin predicted biochemical recurrence in both the early and late periods. Key words: radical prostatectomy, biochemical recurrence, predictive factor, prostate cancer, positive surgical margin Introduction In 2013, 1.4 million men were affected with prostate cancer, which is the most common of male-specific cancer, and 290,000 men died from prostate cancer, which was the sixth leading cause of cancer death in males, in the world (1). At the same time, men with localized prostate cancer who receive appropriate treatment can expect long-term survival. Radical prostatectomy (RP), which has been performed safely since Reiner et al. reported an anatomical approach to the surgical management of the dorsal vein, remains one of the most efficacious treatment modalities (2 4). After The Author 2016. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com 233

234 Late recurrence after prostatectomy surgery, surveillance is conducted by periodic prostate-specific antigen (PSA) monitoring. PSA elevation, which is termed biochemical recurrence (BCR), occurs in 15 33% of patients (5 7). Because BCR has been reported to lead to distant metastasis and cancer death, it is necessary for men with BCR to undergo salvage therapy or to be followed up closely (5,8). While BCR usually occurs in early period (within 5 years after RP), some cases occur in the late period (>5 years after RP)(7,9). Although it has been reported that the backgrounds of patients with BCR differ according to the period after surgery in which BCR occurs (7,10), it remains controversial whether the prognosis differs according to the time of recurrence (5,6).With regard to salvage therapy for BCR,most patients in past studies about salvage therapy suffered from recurrence in the early period after BCR (11 13). It is therefore unknown whether salvage therapy for BCR in the late period is necessary or when treatment should be administered to such patients. In spite of the lack of knowledge about BCR at 5 years after surgery, because of the possibility of distant metastasis, patients should be followed up for a long period of time and there is no evidence as to when surveillance can be safely ended; however, the probability of BCR advancing to life-threatening disease might differ according to the period of recurrence (14). There are some reports about predictors of BCR at <5 years after surgery (15,16). In contrast, there have been few studies on predictors of BCR at 5 years after surgery (14). At present, RP has been widely performed for a long time, thus there are many long-term survivors undergoing RP and it is important to recognize the predictive factors of BCR, not only in the early period after surgery but also in the late period. We conducted a multicenter retrospective study to assess the different predictors of BCR in the different periods. Patients and methods We reviewed 478 men who underwent open retropubic RP for clinically localized prostate cancer without neoadjuvant hormonal therapy at four hospitals from 1999 to 2009. All of the patients were followed up for at least 5 years after RP. The four hospitals are all hub medical institutions for cancer treatment in their respective prefectures in Japan. The present study was approved by the institutional review board of the National Hospital Organization Kyushu Cancer Center. After RP, PSA measurement was performed every few months and radiographic imaging was not routinely conducted. Because this was a retrospective analysis, the follow-up protocols were not predetermined and the schedules of PSA measurement were determined by each physician. Postoperative therapy, including adjuvant or salvage treatment, was administered to 114 men during the follow-up period. In all cases, this involved radiation therapy to the prostate bed or hormonal therapy. The treatment modality was determined by the individual physicians. BCR was defined as a single PSA level of 0.2 ng/ml or the initiation of postoperative therapy for a PSA level of <0.2 ng/ml, because until 2010, the Japanese classification of prostate cancer defined BCR after RP as three consecutive PSA increases (17). Thus, some of the patients who received postoperative therapy had a PSA level of <0.2 ng/ml. The date of BCR was defined as the day when the PSA level increased to 0.2 ng/ml for the first time after RP or when salvage treatment was initiated while the PSA level was <0.2 ng/ml. If the PSA level had never been reduced to <0.2 ng/ml after RP, then the date of the surgery was defined as the date of BCR. Metastases were diagnosed based on the review of computed tomography, magnetic resonance imaging or bone scan images. The physicians performed diagnostic imaging when the PSA level increased. The Kaplan Meier method was used to estimate the BCR-free survival rate at 5 and 10 years after RP. The patients in whom BCR occurred during the follow-up period were divided into three groups. The first group contained men who did not undergo BCR; the second group contained men in whom BCR occurred at <5 years after RP; the third group contained men in whom BCR occurred at 5 years after RP. The background characteristics of each group were compared using the χ 2 test. The background characteristics of men who received postoperative therapy were compared according to their PSA level at the initiation of the additional treatment, with the χ 2 test used to compare categorical variables and Student s t-test used to compare continuous variables. A Cox multivariate regression analysis was performed to test the association of clinical or pathological parameters with BCR at <5 years and 5 years after RP. The covariates included the preoperative PSA level, the margin status, the pathological T stage and the Gleason score. Hazards ratios and the corresponding 95% confidence intervals were determined for each covariate. Two-sided P values of <0.05 were considered to indicate statistically significant differences. All of the analyses were performed using the JMP Pro software package (version 9.0.2, SAS Institute, Inc., Cary, NC, USA). Results Table 1 shows the characteristics of the 478 patients. The median follow-up period was 7.5 years (range: 5.0 15.6 years). A total of 135 patients had BCR. In 113 (84%) of the patients, BCR occurred at <5 years after RP; in 22 (16%) patients BCR occurred at 5 years after RP. The 5-year and 10-year recurrencefree survival rates were 76.3 and 67.9%, respectively (Fig. 1). The preoperative PSA level, margin status, pathological features and incidence of distant metastasis in each group are shown in Table 1. The proportion of patients in whom the preoperative PSA level was >10 ng/ml and seminal vesicle invasion was significantly larger in patients with BCR at <5 yearsafterrpthaninpatientswithbcrat 5 years after RP. There were no significant differences between the two groups in the margin status or pathological features, including extracapsular extension (EPE) and Gleason score. Postoperative therapy was administered to a significantly higher proportion of the patients with BCR at <5 years after RP. With respect to the prognosis, there was no difference between the two groups in the incidence of metastasis and no patients died of prostate cancer in this cohort. Postoperative therapy was administered to 114 men. Table 2 shows their characteristics. Thirty-eight (33%) men received postoperative therapy before the PSA level increased to 0.2 ng/ml. Their preoperative PSA and pathological features were not significantly different from those of the men who underwent postoperative therapy with a PSA level of 0.2 ng/ml (Table 2A). On the other hand, they were less favorable than those of men in whom BCR did not occur (Table 2B). The multivariate analysis revealed that a preoperative PSA level and a positive surgical margin were significant predictors of BCR at <5 years after RP (HR: 1.03 and 3.20) (Table 3A). On the other hand, among 365 patients in whom BCR did not occur for 5 years after RP, a positive surgical margin was the only significant predictor of BCR at 5 years after RP (HR: 3.03) (Table 3B). Discussion Fifteen to thirty-three of patients who undergo RP for clinically localized prostate cancer are reported to experience BCR (5 7). Although 77 94% of BCR occurs within the early (<5 years after RP), it can

Jpn J Clin Oncol, 2017, Vol. 47, No. 3 235 Table 1. A comparison of the characteristics of the patients with PSA recurrence 5 years and >5 years after surgery Characteristics Total (n = 478) BCR 5 years (n = 113) BCR >5 years (n = 22) P Patients (%) Patients (%) Patients (%) Age 60 83 (17) 18 (16) 6 (27) 0.394 61 70 279 (59) 64 (57) 12 (55) 71 116 (24) 31 (27) 4 (18) Preoperative PSA (ng/ml) 10 317 (66) 50 (44) 18 (82) 0.00230 >10, 20 118 (25) 47 (42) 2 (9) >20 43 (9) 16 (14) 2 (9) Surgical margin Positive 119 (25) 57 (50) 13 (59) 0.0989 Negative 355 (74) 56 (50) 8 (36) Unknown 4 (1) 0 1 (5) Pathological stage EPE 0 306 (64) 58 (51) 11 (50) 0.726 1 166 (35) 54 (48) 10 (45) Unknown 6 (1) 1 (1) 1 (5) SVI 0 448 (94) 101 (89) 21 (95) 0.0465 1 24 (5) 11 (10) 0 Unknown 6 (1) 1 (1) 1 (5) pn pn0 461 (96) 105 (93) 22 (100) 0.187 pn1 9 (2) 8 (7) 0 Unknown 8 (2) 0 0 Pathological Gleason score 6 117 (24) 24 (21) 4 (19) 0.103 7 253 (53) 50 (44) 14 (67) 8 10 106 (23) 39 (35) 3 (14) Unknown 2 0 0 No. of salvage therapy 114 (24) 101 (89) 13 (59) 0.0012 No. of metastasis 4 (1) 3 (3) 1 (5) 0.608 BCR, biochemical recurrence; PSA, prostate-specific antigen; EPE, extracapsular extension; SVI, seminal vesicle invasion. Figure 1. BCR-free survival curve. BCR, biochemical recurrence. also occur in the late period ( 5 years after RP) (7,9). Similarly to previous studies, BCR occurred in 135 of the 478 patients in the present cohort; in 113 (84%) cases, BCR occurred at <5 years after RP, while in 22 (16%) cases, BCR occurred at 5 years after RP. The latest case of BCR occurred 11.4 years after RP. The accumulated evidence regarding the prognosis of patients with BCR at 5 years after RP is insufficient. Loeb and Caire reported that cases in which BCR occurs at 5 years after RP are less likely to progress to metastasis or cancer death than cases in which BCR occurs at <5 years after RP (7,10). On the other hand, Boorijian reported that the time from RP to BCR was not significantly associated with the risk of systemic progression or cancer-specific mortality (5). However, this issue remains controversial. The background characteristics of patients with BCR, including the preoperative PSA, pathological stage and margin status, differ according to the timing of BCR (7,10); thus, the prognosis after BCR may differ according to the periods from RP to BCR. At any rate, BCR that occurs at 5 years after RP has the potential to progress to metastasis or cancer death, so it is necessary for patients who undergo RP to be followed up for a long period of time, and for salvage therapy to be administered in some cases. Some studies have reported that the preoperative PSA, a positive surgical margin, pathological stage and other factors are predictors of BCR (18 20). In line with previous reports, we found that the preoperative PSA and a positive surgical margin were predictors of BCR at <5 years after RP. When we investigated the patients in whom BCR occurred at 5 years after RP, a positive surgical margin remained a predictor of BCR but the preoperative PSA did not. Ahove conducted similar research and concluded that the pathological stage and Gleason score were predictors of BCR at 5 years

236 Late recurrence after prostatectomy Table 2. A comparison of the characteristics of patients undergoing postoperative therapy with PSA levels <0.2 and 0.2 ng/ml (A) and the characteristics of patients without BCR with who underwent postoperative therapy with a PSA level of <0.2 ng/ml (B) (A) Characteristics Total (n = 114) Initiation of postoperative therapy with PSA < 0.2 (n = 38) Initiation of postoperative therapy with PSA 0.2 (n = 76) P Median PSA at initiation of postoperative 0.31 (0.008 4.64) 0.075 (0.008 0.199) 0.6 (0.20 4.64) <0.0001 therapy, ng/ml (range) Median period from surgery to postoperative 472 (3 4320) 203 (3 3565) 580 (28 4320) 0.0036 therapy, day (range) Median preoperative PSA, ng/ml (range) 10.73 (3.10 104) 10.4 (3.10 104) 10.66 (4.4 58.9) 0.0987 Modality of treatment, no. (%) Hormonal therapy 78 (68) 19 (50) 59 (78) 0.0031 Radiation therapy to prostate bed 36 (32) 19 (50) 17 (22) Surgical margin, no. (%) Positive 56 (49) 22 (58) 34 (45) 0.171 Negative 58 (51) 16 (42) 42 (55) Pathological stage, no. (%) EPE 0 49 (43) 13 (34) 36 (47) 0.216 1 64 (56) 24 (63) 40 (53) Unknown 1 (1) 1 (3) SVI 0 102 (89) 31 (81) 71 (93) 0.116 1 11 (10) 6 (16) 5 (7) Unknown 1 (1) 1 (3) pn pn0 105 (92) 34 (90) 71 (93) 0.833 pn1 7 (6) 2 (5) 5 (7) Unknown 2 (2) 2 (5) Pathological Gleason score, no. (%) 6 24 (21) 10 (26) 14 (18) 0.549 7 51 (45) 17 (45) 34 (45) 8 10 39 (34) 11 (29) 28 (37) (B) Characteristics Survival without BCR (n = 343) Initiation of postoperative therapy with PSA < 0.2 (n = 38) P Median PSA at initiation of postoperative therapy, ng/ml (range) 0.075 (0.008 0.199) Median period from surgery to postoperative therapy, day (range) 203 (3 3565) Median preoperative PSA, ng/ml (range) 7.37 (0.9 88.33) 10.4 (3.10 104) <0.001 Modality of treatment (%) Hormonal therapy 19 (50) Radiation therapy to prostate bed 19 (50) Surgical margin (%) Positive 55 (16) 22 (58) <0.001 Negative 285 (83) 16 (42) Unknown 3 (1) Pathological stage (%) EPE 0 241 (70) 13 (34) <0.001 1 98 (29) 24 (63) Unknown 4 (1) 1 (3) SVI 0 326 (95) 31 (81) 0.0076 1 13 (4) 6 (16) Unknown 4 (1) 1 (3) pn pn0 336 (98) 34 (90) 0.017 pn1 1 2 (5) Unknown 6 (2) 2 (5) Pathological Gleason score (%) 6 89 (26) 10 (26) 0.296 7 189 (55) 17 (45) 8 10 64 (19) 11 (29) Unknown 1

Jpn J Clin Oncol, 2017, Vol. 47, No. 3 237 Table 3. Univariate and multivariate analyses of various factors affecting recurrence at <5 years (A) and 5 years after surgery (B) (A) n Univariate analysis P Multivariate analysis P Preoperative PSA Elevation of PSA by 1 ng/ml 478 1.03 (1.02 1.04) <0.0001 1.03 (1.01 0.97) <0.0001 Surgical margin Negative 119 1 (reference) 1 (reference) Positive 355 3.81 (2.63 5.51) <0.0001 3.20 (2.07 4.97) <0.0001 Unknown 4 Extracapsular extension Negative 306 1 (reference) 1 (reference) Positive 166 2.27 (1.56 3.29) <0.0001 1.18 (0.75 1.84) 0.473 Unknown 6 Seminal vesicle invasion Negative 448 1 (reference) 1 (reference) Positive 24 2.67 (1.35 4.75) 0.00650 1.33 (0.67 2.45) 0.408 Unknown 6 Pathological Gleason score 6 117 1 (reference) 1 (reference) 7 253 0.94 (0.58 1.55) 0.804 0.77 (0.48 1.29) 0.312 8 10 106 1.92 (1.16 3.23) 0.0108 1.32 (0.78 2.26) 0.306 Unknown 2 (B) n Univariate analysis P Multivariate analysis P Preoperative PSA Elevation of PSA by 1 ng/ml 365 1.00 (0.93 1.03) 0.872 0.98 (0.92 1.04) 0.503 Surgical margin Negative 298 1 (reference) 1 (reference) Positive 63 3.41 (1.34 8.16) 0.0117 3.03 (1.08 8.16) 0.0360 Unknown 4 Extracapsular extension Negative 252 1 (reference) 1 (reference) Positive 108 2.52 (1.04 6.06) 0.0406 1.93 (0.68 5.31) 0.212 Unknown 5 Seminal vesicle invasion Negative 347 1 (reference) 1 (reference) Positive 13 0.239 0.106 Unknown 5 Pathological Gleason score 6 93 1 (reference) 1 (reference) 7 203 2.36 (0.85 8.35) 0.105 1.89 (0.66 6.79) 0.248 8 10 67 1.17 (0.23 5.31) 0.837 1.01 (0.20 4.68) 0.990 Unknown 2 HR, hazard ratio; CI, confidence interval. after RP (14). This finding was completely different from our results. Because there are only a few reports about the predictors of late BCR, it is difficult to suggest a reason for this difference. On one hand, the Kaplan Meier curves in their study showed that the patients with a high Gleason score or a high pathological stage (such as pt3b) experienced BCR at 5 years after RP. On the other hand, in our cohort, BCR in such patients usually occurred at <5 years after RP and there were only a few cases of BCR at 5 years after RP. Similarly, a report about the prognosis of Japanese patients after RP noted that among patients with a high Gleason score or a high pathological stage, BCR usually occurred at <5 years after RP (20). Although this discrepancy may be due to racial differences, the accumulation of further evidence is necessary to discuss this problem. High preoperative PSA and a positive surgical margin have been reported to predict BCR in the early period (18,20). In our study, the preoperative PSA level was a predictor of BCR in the early period, while patients with a positive surgical margin continued to develop BCR for a long period of time. It has been demonstrated that in patients with a positive surgical margin, the tumor grade at the site of the positive margin has prognostic value and that the Gleason score of the RP specimen is associated with PSA velocity (21,22). Considering these reports, it is hypothesized that among the cases with a positive surgical margin, the cases with a high tumor grade at the site of the positive margin undergo BCR in the early period after RP because of their high PSA velocity, while those with a low tumor grade at the site of the positive margin undergo BCR in the late period because of

238 Late recurrence after prostatectomy their low PSA velocity. Although, further studies on the relationship of the tumor grade at the margin site and the period from RP to BCR are required, it is necessary to follow-up cases in which there is a positive surgical margin for a long time after RP. The present study is associated with some limitations. First, some patients received postoperative therapy when their PSA was <0.2 ng/ml. This is partly because, until 2010, the Japanese classification of prostate cancer defined BCR after RP as three consecutive increases in the PSA level regardless of the absolute value. Thus, these patients received early postoperative therapy; however, their preoperative PSA level and pathological features were as unfavorable as those of the men whose PSA levels increased to 0.2 ng/ml after RP. Moreover, their preoperative PSA level and the rates of patients with a positive surgical margin, EPE, SVI and pn were significantly higher than those of the men in whom BCR did not occur. Thus, early BCR was expected to occur in these patients and the effects of the early postoperative therapy on the results are thought to be acceptable. Second, we did not discuss overall survival because of the short follow-up period and the absence of cancer-related mortality. Although the examination of the prognosis was insufficient, the elevation of the PSA level after RP has been reported to precede clinical recurrence or distant metastasis. It is therefore reasonable to set BCR as an endpoint for exchange (8). Last, although the follow-up protocol was not predetermined due to the retrospective nature of the study, the results can be generalized because it was a multicenter study and all of the participating institutions were hub medical institutions that provide cancer treatment for their respective prefectures in Japan. Conclusion Sixteen percent of BCRs occurred at 5 years after RP. Although the preoperative PSA levels of the patients were relatively low, there was one case that advanced to metastasis. Because a positive surgical margin was the predictor of BCR at <5 and 5 years after RP, cases that involve a positive surgical margin should be followed up for a long period of time. Conflict of interest statement None declared. References 1. Fitzmaurice C, Dicker D, Pain A, et al. The global burden of cancer 2013. JAMA Oncol 2015;1:505 27. 2. Mottet N, Bastian PJ, Bellmunt J, et al. Guidelines on prostate cancer. ARNHEM: European Association of Urology, 2014. 3. Rosenfeld S. NCCN clinical practice guidelines in oncology (NCCN Guidelines ) prostate cancer. Washington: National Comprehensive Cancer Network, 2014.2014. 4. Reiner WG, Walsh PC. An anatomical approach to the surgical management of the dorsal vein and Santorini s plexus during radical retropubic surgery. J Urol 1979;121:198 200. 5. Boorjian SA, Thompson RH, Tollefson MK, et al. Long-term risk of clinical progression after biochemical recurrence following radical prostatectomy: the impact of time from surgery to recurrence. Eur Urol 2011;59:893 9. 6. Ward JF, Blute ML, Slezak J, Bergstralh EJ, Zincke H. The long-term clinical impact of biochemical recurrence of prostate cancer 5 or more years after radical prostatectomy. J Urol 2003;170:1872 6. 7. Loeb S, Feng Z, Ross A, Trock BJ, Humphreys EB, Walsh PC. Can we stop prostate specific antigen testing 10 years after radical prostatectomy? J Urol 2011;186:500 5. 8. Pound CR, Partin AW, Eisenberger MA, Chan DW, Pearson JD, Walsh PC. Natural history of progression after PSA elevation following radical prostatectomy. JAMA 2014;281:1591 97. 9. Amling CL, Blute ML, Bergstralh EJ, Seay TM, Slezak J, Zincke H. Longterm hazard of progression after radical prostatectomy for clinically localized prostate cancer: continued risk of biochemical failure after 5 years. J Urol 2000;164:101 5. 10. Caire AA, Sun L, Ode O, et al. Delayed prostate-specific antigen recurrence after radical prostatectomy: how to identify and what are their clinical outcomes? Urology 2009;74:643 7. 11. Stephenson AJ, Scardino PT, Kattan MW, et al. Predicting the outcome of salvage radiation therapy for recurrent prostate cancer after radical prostatectomy. J Clin Oncol 2007;25:2035 41. 12. MacDonald OK, Schild SE, Vora S, et al. Salvage radiotherapy for men with isolated rising PSA or locally palpable recurrence after radical prostatectomy: do outcomes differ? Urology 2004;64:760 4. 13. Trock BJ, Han M, Freedland SJ, et al. Prostate cancer-specific survival following salvage radiotherapy vs observation in men with biochemical recurrence after radical prostatectomy. JAMA 2008;299:2760 9. 14. Ahove DA, Hoffman K E, Hu JC, Choueiri TK, D Amico AV, Nguyen PL. Which patients with undetectable PSA levels 5 years after radical prostatectomy are still at risk of recurrence? Implications for a riskadapted follow-up strategy. Urology 2010;76:1201 5. 15. Graefen M, Noldus J, Pichlmeier U, et al. Early prostate-specific antigen relapse after radical retropubic prostatectomy: prediction on the basis of preoperative and postoperative tumor characteristics. Eur Urol 1999;36:21 30. 16. Catalona WJ, Smith DS. Cancer recurrence and survival rates after anatomic radical retropubic prostatectomy for prostate cancer: intermediateterm results. J Urol 1998;160:2428 34. 17. Murai M. Japanese Classification of Prostate Cancer. 3rd edn. Tolyo: The Japanese Urological Association, 2001. 18. Ploussard G, Agamy MA, Alenda O, et al. Impact of positive surgical margins on prostate-specific antigen failure after radical prostatectomy in adjuvant treatment-naïve patients. BJU int 2011;107:1748 54. 19. Rouanne M, Rode J, Campeggi A, et al. Long-term impact of positive surgical margins on biochemical recurrence after radical prostatectomy: ten years of follow-up. Scandinavian. J Urol 2014;48:131 37. 20. Yokomizo A, Murai M, Baba S, et al. Percentage of positive biopsy cores, preoperative prostate-specific antigen (PSA) level, pt and Gleason score as predictors of PSA recurrence after radical prostatectomy: a multiinstitutional outcome study in Japan. BJU int 2006;98:549 53. 21. Yossepowitch O, Briganti A, Eastham JA, et al. Positive surgical margins after radical prostatectomy: a systematic review and contemporary update. Eur Urol 2014;65:303 13. 22. Loeb S, Sutherland DE, D Amico AV, Roehl KA, Catalona WJ. PSA velocity is associated with Gleason score in radical prostatectomy specimen: marker for prostate cancer aggressiveness. Urology 2008;72: 1116 20.