Department of Urology, Osaka Medical Center for Cancer and Cardiovascular Diseases, Osaka, Japan

JJCO Japanese Journal of Clinical Oncology Japanese Journal of Clinical Oncology, 2017, 47(1) 74 79 doi: 10.1093/jjco/hyw150 Advance Access Publication Date: 12 October 2016 Original Article Original Article Postoperative early ultrasensitive prostatespecific antigen identifies patients at risk for biochemical recurrence in margin positive prostate cancers: a single-center study Koji Hatano, Takuya Okusa, Yu Ishizuya, Yasutomo Nakai, Masashi Nakayama, Ken-ichi Kakimoto, and Kazuo Nishimura* Department of Urology, Osaka Medical Center for Cancer and Cardiovascular Diseases, Osaka, Japan *For reprints and all correspondence: Kazuo Nishimura, Department of Urology, Osaka Medical Center for Cancer and Cardiovascular Diseases, 1-3-3 Nakamichi, Higashinari-ku, Osaka 537-8511, Japan. E-mail: nisimura-ka2@mc.pref.osaka.jp Received 15 June 2016; Accepted 30 September 2016 Abstract Objective: To identify predictors for biochemical recurrence among patients with positive surgical margins (RM1) after radical prostatectomy and to examine the effect of ultrasensitive prostatespecific antigen measured early after prostatectomy on biochemical recurrence. Methods: We identified 705 patients with prostate cancer who were treated with radical prostatectomy without preoperative hormonal therapy at our institution between 2000 and 2014. The patients with RM1 who had a postoperative prostate-specific antigen <0.2 ng/ml without lymph node metastasis were evaluated for biochemical recurrence free survival. Survival rates were calculated using the Kaplan Meier method. The Cox regression model was used for multivariate analysis. The prediction of biochemical recurrence was assessed using area under the curve of the receiver operating characteristic. Results: Among the 705 patients, 190 (27%) had RM1. Biochemical recurrence was evaluated in 164 patients, excluding 26 patients who underwent adjuvant therapy with or without lymph node metastasis. With a median follow-up of 55 months, the biochemical recurrence free survival rate of the entire RM1 cohort was 78% at 2 years and 64% at 4 years. The multivariate analysis revealed that postoperative early ultrasensitive prostate-specific antigen >0.02 ng/ml was the significant risk factor for biochemical recurrence (hazard ratio 13.10). Meanwhile, the patients with postoperative early ultrasensitive prostate-specific antigen <0.01 ng/ml had a significantly lower risk for biochemical recurrence (hazard ratio 0.12). Area under the curve for the postoperative early ultrasensitive prostate-specific antigen value to predict biochemical recurrence was 0.789. Conclusions: The ultrasensitive prostate-specific antigen value measured early after prostatectomy was the potent predictor of biochemical recurrence among the patients with RM1. Key words: biochemical recurrence, positive surgical margin, prostatectomy, prostate cancer, prostate-specific antigen Introduction Radical prostatectomy is one of the major treatments for localized prostate cancer. The aim of radical prostatectomy is to improve oncological outcome without harming quality of life aspects such as urinary and erectile functions. Occasionally, however, some peripheral tumor foci are not recognized by surgeons. As a result, a positive surgical margin (RM1) is reported to occur at a rate of 6.5 38% after radical prostatectomy (1,2). Although the clinical effect The Author 2016. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com 74

Jpn J Clin Oncol, 2017, Vol. 47, No. 1 75 of RM1 on cancer prognosis still remains unclear (2), a number of studies have demonstrated that RM1 increases the risk of biochemical recurrence (BCR) after radical prostatectomy (3 9). Thus, patients with RM1 are candidates for immediate adjuvant radiation therapy (RT) to decrease BCR and provide a survival benefit compared with that of patients undergoing observation (10 12). However, several reports indicated that BCR did not occur in over half of the patients, even in the presence of RM1, over 5-year follow-up (4,7,8). This raises an issue for the treatment of patients with RM1 in which a number of patients are overtreated by immediate adjuvant RT. Thus, the predictors for BCR among the patients with RM1 may help to solve this issue by identifying the patients with potential localized failure. Among predictors of BCR, postoperative ultrasensitive prostatespecific antigen (upsa) assay is one of the potential candidates. Conventional PSA assays have been applied to detect BCR, in which the detection threshold of 0.2 ng/ml was considered adequate to detect BCR (13). The upsa was established to detect PSA with a lower limit of detection <0.1 ng/ml (14). Although the clinical usefulness of the upsa after radical prostatectomy is still unclear, upsa may have the potential to sensitively predict BCR with a lead time advantage over that of the conventional relapse definition (15). We previously reported that postoperative nadir value of upsa was a predictor of BCR among the patients with adverse pathological features (16). Recently, several reports focused on a postoperative early upsa value to predict BCR. Inoue et al. reported that the upsa value measured 3 months after prostatectomy predicted BCR among patients with localized prostate cancer (17). Recent reports also supported the use of postoperative early upsa value to identify patients who require salvage therapy among those with adverse pathological features (15,18). The postoperative early upsa may be useful to detect patients who require salvage therapy allowing adequate time for planning eventual therapy without waiting for the nadir PSA value. Here, we retrospectively evaluated the patients with RM1 after radical prostatectomy. The goal of this study was to identify predictors for BCR among patients with RM1 after radical prostatectomy. We examined the effect of upsa measurement early after prostatectomy on BCR as well as the traditional predictors for BCR. Patients and methods Patient selection We retrospectively reviewed our clinicopathological database of patients with prostate cancer who were treated with radical prostatectomy at the Osaka Medical Center for Cancer and Cardiovascular Diseases between January 2000 and December 2014. We identified 1023 patients, including 856 cases of open surgery and 167 cases of robotic surgery. The exclusion criteria of this study were patients with preoperative hormonal therapy or insufficient clinicopathological data. Among the 1023 patients, 705 were eligible for this study. RM1 was defined if cancer cells were in touch with the ink on the surface of the radical prostatectomy specimen. The patients with RM1 who had a postoperative PSA <0.2 ng/ml were further evaluated for BCR-free survival. The patients with RM1 who received adjuvant RT and/or adjuvant hormonal therapy as well as those with lymph node metastasis were excluded from the analysis. PSA follow-up For PSA follow-up, upsa was used after radical prostatectomy. The cut-off value of upsa was 0.02 ng/ml from 2000 to 2003 and 0.003 ng/ml from 2004 to 2015. The upsa kit had been changed from Elecsys (Roche, Basel, Switzerland) to Architect (Abbott, Chicago, Illinois) where the compatibility between the two kits was confirmed in our institution. upsa values were measured at least every 3 months until 2 years after surgery, every 6 months until 5 years and annually thereafter. The postoperative early upsa value was defined as the upsa value measured 3 months after prostatectomy. Evaluation The primary endpoint of this study was BCR, which was defined as a PSA value 0.2 ng/ml. Disease progression was defined as BCR, objective progression evaluated in a radiographic study or the start of salvage therapy due to the serial increase of PSA. Progression-free survival (PFS) was measured from the date of surgery to the date of the progression or the last follow-up. Survival rates were calculated using the Kaplan Meier method, and the univariate comparisons were made using the log rank test. The Cox regression model was used for the multivariate analysis. The prediction of BCR was assessed using area under the curve of the receiver operating characteristic. JMP software was used for the statistical analysis. P < 0.05 was considered statistically significant. Results Among the 705 patients, 190 (27%) showed RM1. Among the 190 patients with RM1, pelvic lymphadenectomy was performed in 189 patients excluding 1 patient with National Comprehensive Cancer Network (NCCN) low-risk group. Disease progression was evaluated in 164 patients, excluding 23 patients who underwent immediate adjuvant radiation or hormonal therapy. Five patients who had lymph node metastasis were also excluded from the analysis. The clinicopathological data of the 164 patients are shown in Table 1. Based on the preoperative PSA value, Gleason score at biopsy and clinical stage, patients were stratified according to the NCCN risk groups (19). The pathological features of the patients such as pathological stage and Gleason score at surgery are also shown in Table 1. Further evaluation of the postoperative early upsa value showed that although most of the cases had a postoperative early upsa value <0.02 ng/ml, 23 cases (14%) had a postoperative early upsa value >0.02 ng/ml. With a median follow-up of 55 months (3 181 months), the PFS of the entire RM1 cohort was 91% at 1 year, 78% at 2 years and 64% at 4 years (Fig. 1). Perioperative factors were evaluated for their relation to PFS (Table 2). Among them, NCCN risk group, Gleason score, seminal vesicle invasion and postoperative early upsa were the significant risk factors for PFS. The postoperative early upsa >0.02 ng/ml was the significant risk factor for PFS among different NCCN risk groups (Table 3). No patients with a postoperative early upsa value >0.02 ng/ml survived without BCR for more than 4 years. There was no significant difference in PFS between the initial PSA level, surgical approach (open vs robotic), RM1 status (single vs multiple) or RM1 location (apical vs nonapical) (Table 2). The multivariate analysis revealed that NCCN high-risk group (hazard ratio [HR] 2.05), Gleason score at surgery 8 and over (HR 2.47) and postoperative early upsa value >0.02 ng/ ml (HR 13.10) were the significant risk factors for PFS (Table 4). Within the follow-up period, 45 (90%) out of 50 patients who developed BCR were treated by salvage RT and/or hormonal therapy. Then, we examined if time to metastatic-free survival was

76 Ultrasensitive PSA after prostatectomy Table 1. Summary of 164 patients with RM1 eligible for analysis NCCN risk group Low 25 (15) Intermediate 105 (63) High 34 (22) Initial PSA (ng/ml) <10 99 (60) 10 20 53 (32) >20 12 (8) Type of surgery Open 128 (78) Robot 36 (22) Gleason score at surgery 6 35 (21) 7 105 (63) 8a 24 (16) Pathological T stage T2 78 (47) T3a 75 (45) T3b 11 (8) RM1 status Apex 131 (80) Base 22 (13) Lateral 39 (24) Multiple 26 (16) <0.02 141 (86) 0.02 23 (14) No of patients (%) NCCN, National Comprehensive Cancer Network; PSA, prostate-specific antigen; upsa, ultrasensitive prostate-specific antigen. Figure 1. Progression-free survival of the entire cohort of 164 patients with a positive surgical margin (RM1). influenced by the postoperative early upsa status. The patients with postoperative early upsa value >0.02 ng/ml had significant worse metastatic-free survival at 4 years compared with those who had postoperative early upsa <0.02 ng/ml (90% vs 100%, respectively; P = 0.0009). Table 2. Univariate analysis for PFS (n = 164) Variables n PFS (%) P value 1 year 2 years 4 years NCCN risk group Intermediate/Low 130 93 82 70 0.0018 High 34 82 62 44 Initial PSA (ng/ml) <10 99 88 79 69 0.5055 10 65 95 77 59 Type of surgery Open 128 92 78 65 0.7362 Robot 36 88 77 Not applicable Gleason score at surgery 7 140 94 80 68 0.0045 8 24 73 68 45 Capsular invasion Negative 82 96 82 63 0.3279 Positive 82 86 74 67 Seminal vesicle invasion Negative 153 94 81 66 0.0004 Positive 11 50 40 40 RM1 status Single 138 93 78 65 0.5765 Multiple 26 80 75 59 RM1 location Apical-only 107 93 77 63 0.9861 Non-apical 57 87 79 66 <0.02 141 98 87 76 <0.0001 0.02 23 46 25 0 PFS, progression-free survival; RM1, positive surgical margin. Table 3. Univariate analysis for PFS among different NCCN risk groups (n = 164) Variables n PFS (%) P value 1 year 2 years 4 years NCCN low/intermediate risk group <0.02 115 98 90 81 <0.0001 0.02 15 60 29 0 NCCN high-risk group <0.02 26 100 74 57 <0.0001 0.02 8 15 15 0 We further analyzed the efficacy of postoperative early upsa value. In our institution, the cut-off value of upsa was changed from 0.02 to 0.003 ng/ml in 2003. Then, 149 patients were eligible for the further analysis, excluding 15 patients who had an upsa value <0.02 ng/ml before 2003. The 149 patients were divided into three different groups according to the postoperative early upsa value: <0.01 ng/ml, 0.01 but <0.02 ng/ml, and 0.02 ng/ml. The BCR-free survival rate was significantly different between three groups (Fig. 2). Perioperative factors were evaluated for their relation to PFS with upsa cut-off value 0.01 ng/ml (Table 5). Identified risk factors for PFS were NCCN risk group, Gleason score, seminal vesicle invasion and postoperative early upsa. These were identical factors found in the evaluation of the entire RM1 cohort as shown

Jpn J Clin Oncol, 2017, Vol. 47, No. 1 77 Table 4. Multivariate analysis for PFS (n = 164) Variable HR 95% CI P value NCCN risk group (high vs 2.05 1.06 3.83 0.033 intermediate/low) Seminal vesicle invasion (positive vs 1.58 0.57 3.79 0.354 negative) Gleason score at surgery ( 8vs 7) 2.47 1.12 5.12 0.025 Postoperative early upsa ( 0.02 vs <0.02 ng/ml) 13.10 6.70 25.37 <0.0001 HR, hazard ratio; CI, confidence interval. Figure 2. Progression-free survival was evaluated according to the postoperative early ultrasensitive prostate-specific antigen (upsa) value: <0.01 ng/ml (Group 1; n = 102), 0.01 but <0.02 ng/ml (Group 2; n = 24) and 0.02 ng/ml (Group 3; n = 23). P < 0.0001. in Table 2. The multivariate analysis demonstrated that postoperative early upsa value <0.01 ng/ml significantly lowered a risk (HR 0.12) for BCR (Table 6). The area under the receiver operating characteristic curve for postoperative early upsa to predict BCR was 0.789 (Fig. 3). Discussion RM1 is reported to be a risk factor for BCR. However, over half of the patients with RM1 did not experience BCR at 4 years in this study using the BCR threshold of 0.2 ng/ml. This result was comparable with that of previous reports (4,7,8). Recently, Toussi et al. reported that the BCR threshold of 0.4 ng/ml correlated with subsequent systemic progression (20). Although it is impossible to analyze out data due to the initiation of salvage therapy, the number of patients without recurrence is likely to increase by using 0.4 ng/ml as BCR threshold. Therefore, RM1 may not always represent residual cancer cells that can cause disease recurrence. Many artifacts may influence the diagnosis of RM1 and cause false positives of the pathology (1). Furthermore, cancer cell death may be caused by tissue cautery, desmoplasia or ischemia of the resection margin (21). Table 5. Univariate analysis for PFS with upsa cut-off value 0.01ng/ml (n = 149) Variables n PFS (%) P value 1 year 2 years 4 years NCCN risk group Intermediate/Low 117 94 82 70 0.0025 High 32 81 59 39 Initial PSA (ng/ml) <10 87 87 77 64 0.9804 10 62 95 77 61 Type of surgery Open 113 92 77 63 0.8173 Robot 36 88 77 Not applicable Gleason score at surgery 7 127 94 79 65 0.0046 8 22 71 65 46 Capsular invasion Negative 73 96 80 57 0.9348 Positive 76 86 74 70 Seminal vesicle invasion Negative 138 94 80 65 0.0006 Positive 11 50 40 40 RM1 status Single 124 93 77 62 0.865 Multiple 25 79 74 63 RM1 location Apical-only 94 93 76 59 0.7772 Non-apical 55 87 78 68 <0.01 102 100 89 80 <0.0001 0.01 47 71 50 28 Among the patients with RM1, it is necessary to identify the risk factors or biomarkers for disease recurrence that may help to detect the patients who require additional therapy. In this study, the NCCN high-risk group and a Gleason score at surgery 8 were identified as the risk factors for BCR (Table 4). Several reports support these results, indicating that the advanced pathological features are likely to increase the risk of BCR for the patients with RM1 (5,6,22,23). Here, we ultimately focused on the upsa value measured early after prostatectomy because a value >0.02 ng/ml was the most potent risk factor for disease recurrence (HR 13.10) compared with NCCN high-risk group (HR 2.05) and Gleason score 8 (HR 2.47). upsa has been used by several groups to predict the risk of BCR after radical prostatectomy. Patients with an undetectable upsa value after surgery are less likely to experience recurrence (17,24 26). We previously reported that postoperative nadir upsa value was a predictor of BCR among patients with adverse pathological features such as pathological T3 and/or RM1 (16). upsa may be useful to predict early disease recurrence (27) and determine patient followup protocol. However, it is still unclear whether earlier detection of BCR translates into prolonged time to metastasis (14). Although the patients with postoperative early upsa value >0.02 ng/ml had worse metastatic-free survival in this study, the limitations include the relatively small rate of metastatic recurrence and the limited followup period. Patients with RM1 are candidates for both immediate adjuvant RT and salvage RT after BCR (28). Currently, immediate adjuvant RT is applied for the prostate cancer patients with adverse

78 Ultrasensitive PSA after prostatectomy Table 6. Multivariate analysis for PFS with upsa cut-off value 0.01ng/ml (n = 149) Variable HR 95% CI P value NCCN risk group (high vs intermediate/ 1.67 0.84 3.20 0.142 low) Seminal vesicle invasion (positive vs 2.90 1.04 6.92 0.043 negative) Gleason score at surgery ( 8vs 7) 4.45 1.78 10.91 0.002 Postoperative early upsa (<0.01 vs 0.01 ng/ml) 0.12 0.06 0.24 <0.0001 Figure 3. The receiver operating characteristic curve for postoperative early upsa to predict BCR. AUC, area under the curve. pathological features such as extracapsular extension, seminal vesicle invasion and/or RM1 (10 12). However, adjuvant RT would not be necessary for all of the patients with RM1 because over half of them did not experience recurrence without any adjuvant treatment. Furthermore, salvage RT is another treatment strategy for the patients with RM1 when they experience BCR. Several studies showed that salvage RT is more effective when initiated at a lower PSA level (29 31). Thus, upsa may be useful to promote early salvage treatment and minimize overtreatment. Our results suggested that patients with a postoperative early upsa value >0.02 ng/ml may be candidates for early salvage treatment because most of them experienced recurrence within 4 years (Table 2 and Fig. 2). However, most of the patients with a postoperative early upsa value <0.01 ng/ml survived without recurrence (Table 5 and Fig. 2). Although the timing of postoperative upsa measurement and the cut-off value for upsa are not yet clearly identified, recent reports support the use of postoperative upsa measurement to determine patients requiring salvage treatment (15,18). In conclusion, our study revealed that a postoperative early upsa value was the most potent predictor of BCR among the patients with RM1. The patients with postoperative early upsa greater than 0.02 ng/ml may be candidates for salvage therapy. Further studies are needed to determine the threshold and clinical usefulness of the upsa value after prostatectomy. Acknowledgements This work was supported by Osaka foundation for the prevention of cancer and cardiovascular diseases. Conflicts of interest None declared. References 1. Yossepowitch O, Bjartell A, Eastham JA, et al. Positive surgical margins in radical prostatectomy: outlining the problem and its long-term consequences. Eur Urol 2009;55:87 99. 2. 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. 3. Pfitzenmaier J, Pahernik S, Tremmel T, Haferkamp A, Buse S, Hohenfellner M. Positive surgical margins after radical prostatectomy: do they have an impact on biochemical or clinical progression? BJU Int 2008;102:1413 8. 4. Boorjian SA, Karnes RJ, Crispen PL, et al. The impact of positive surgical margins on mortality following radical prostatectomy during the prostate specific antigen era. J Urol 2010;183:1003 9. 5. Wright JL, Dalkin BL, True LD, et al. Positive surgical margins at radical prostatectomy predict prostate cancer specific mortality. J Urol 2010;183: 2213 8. 6. Chalfin HJ, Dinizo M, Trock BJ, et al. Impact of surgical margin status on prostate-cancer-specific mortality. BJU Int 2012;110:1684 9. 7. Mauermann J, Fradet V, Lacombe L, et al. The impact of solitary and multiple positive surgical margins on hard clinical end points in 1712 adjuvant treatment-naive pt2-4 N0 radical prostatectomy patients. Eur Urol 2013;64:19 25. 8. Sooriakumaran P, Ploumidis A, Nyberg T, et al. The impact of length and location of positive margins in predicting biochemical recurrence after robot-assisted radical prostatectomy with a minimum follow-up of 5 years. BJU Int 2015;115:106 13. 9. Hashimoto T, Yoshioka K, Nagao G, et al. Prediction of biochemical recurrence after robot-assisted radical prostatectomy: analysis of 784 Japanese patients. Int J Urol 2015;22:188 93. 10. Bolla M, van Poppel H, Collette L, et al. European Organization for Research and Treatment of Cancer. Postoperative radiotherapy after radical prostatectomy: a randomized controlled trial (EORTC trial 22911). Lancet 2005;366:572 8. 11. Thompson IM Jr, Tangen CM, Paradelo J, et al. Adjuvant radiotherapy for pathologically advanced prostate cancer: a randomized clinical trial. JAMA 2006;296:2329 35. 12. 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;181:956 62. 13. Cookson MS, Aus G, Burnett AL, et al. Variation in the definition of biochemical recurrence in patients treated for localized prostate cancer: the American Urological Association Prostate Guidelines for Localized Prostate Cancer Update Panel report and recommendations for a standard in the reporting of surgical outcomes. J Urol 2007;177:540 5. 14. Tilki D, Kim SI, Hu B, Dall Era MA, Evans CP. Ultrasensitive prostate specific antigen and its role after radical prostatectomy: a systematic review. J Urol 2015;193:1525 31. 15. Kang JJ, Reiter RE, Steinberg ML, King CR. Ultrasensitive prostate specific antigen after prostatectomy reliably identifies patients requiring postoperative radiotherapy. J Urol 2015;193:1532 8. 16. Yoshida T, Matsuzaki K, Kobayashi Y, et al. Usefulness of postoperative nadir prostate-specific antigen value by ultrasensitive assay as a predictor of prostate-specific antigen relapse for pathological T3 or positive surgical

Jpn J Clin Oncol, 2017, Vol. 47, No. 1 79 margins after radical prostatectomy for prostate cancer. Int Urol Nephrol 2012;44:479 85. 17. Inoue H, Nishimura K, Yamaguchi S, Nonomura N, Hara T. Prostatespecific antigen measured 3 months after radical prostatectomy as a new predictor of biochemical recurrence. Int J Clin Oncol 2015;20:171 5. 18. Vesely S, Jarolim L, Duskova K, Schmidt M, Dusek P, Babjuk M. The use of early postoperative prostate-specific antigen to stratify risk in patients with positive surgical margins after radical prostatectomy. BMC Urol 2014;14:79. 19. Mohler J, Bahnson RR, Boston B, et al. NCCN clinical practice guidelines in oncology: prostate cancer. J Natl Compr Canc Netw 2010;8:162 200. 20. Toussi A, Stewart-Merrill SB, Boorjian SA, et al. Standardizing the definition of biochemical recurrence after radical prostatectomy what prostate specific antigen cut point best predicts a durable increase and subsequent systemic progression? J Urol 2016;195:1754 9. 21. Grossfeld GD, Tigrani VS, Nudell D, et al. Management of a positive surgical margin after radical prostatectomy: decision analysis. J Urol 2000; 164:93 9. discussion 100. 22. Viers BR, Sukov WR, Gettman MT, et al. Primary Gleason grade 4 at the positive margin is associated with metastasis and death among patients with Gleason 7 prostate cancer undergoing radical prostatectomy. Eur Urol 2014;66:1116 24. 23. Karl A, Buchner A, Tympner C, et al. The natural course of pt2 prostate cancer with positive surgical margin: predicting biochemical recurrence. World J Urol 2015;33:973 9. 24. Nakamura M, Hasumi H, Miyoshi Y, et al. Usefulness of ultrasensitive prostate-specific antigen assay for early detection of biochemical failure after radical prostatectomy. Int J Urol 2005;12:1050 4. 25. Chang SL, Freedland SJ, Terris MK, et al. Freedom from a detectable ultrasensitive prostate-specific antigen at two years after radical prostatectomy predicts a favorable clinical outcome: analysis of the SEARCH database. Urology 2010;75:439 44. 26. Malik RD, Goldberg JD, Hochman T, Lepor H. Three-year postoperative ultrasensitive prostate-specific antigen following open radical retropubic prostatectomy is a predictor for delayed biochemical recurrence. Eur Urol 2011;60:548 53. 27. Arai Y, Okubo K, Aoki Y, Maekawa S, Okada T, Maeda H. Ultrasensitive assay of prostate-specific antigen for early detection of residual cancer after radical prostatectomy. Int J Urol 1998;5:550 5. 28. Freedland SJ, Rumble RB, Finelli A, et al. American Society of Clinical Oncology. Adjuvant and salvage radiotherapy after prostatectomy: American Society of Clinical Oncology clinical practice guideline endorsement. J Clin Oncol 2014;32:3892 8. 29. Stephenson AJ, Shariat SF, Zelefsky MJ, et al. Salvage radiotherapy for recurrent prostate cancer after radical prostatectomy. JAMA 2004;291:1325 32. 30. 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. 31. King CR. The timing of salvage radiotherapy after radical prostatectomy: a systematic review. Int J Radiat Oncol Biol Phys 2012;84:104 11.