Untreated Gleason Grade Progression on Serial Biopsies during Prostate Cancer Active Surveillance: Clinical Course and Pathological Outcomes

Untreated Gleason Grade Progression on Serial Biopsies during Prostate Cancer Active Surveillance: Clinical Course and Pathological Outcomes A. A. Hussein,* C. J. Welty,* N. Ameli,* J. E. Cowan, M. Leapman,* S. P. Porten,* K. Shinohara* and P. R. Carroll, From the Department of Urology and UCSF Helen Diller Family Comprehensive Cancer Center, University of California San Francisco, San Francisco, California (AAH, CJW, NA, JEC, ML, SPP, KS, PRC), and Department of Urology, Cairo University, Cairo, Egypt (AAH) Purpose: We describe the outcomes of patients with low risk localized prostate cancer who were upgraded on a surveillance biopsy while on active surveillance and evaluated whether delayed treatment was associated with adverse outcome. Materials and Methods: We included men in the study with lower risk disease managed initially with active surveillance and upgraded to Gleason score 3þ4 or greater. Patient demographics and disease characteristics were compared. Kaplan-Meier curve was used to estimate the treatment-free probability stratified by initial upgrade (3þ4 vs4þ3 or greater), Cox regression analysis was used to examine factors associated with treatment and multivariate logistic regression analysis was used to evaluate the factors associated with adverse outcome at surgery. Results: The final cohort comprised 219 men, with 150 (68%) upgraded to 3þ4 and 69 (32%) to 4þ3 or greater. Median time to upgrade was 23 months (IQR 11e49). A total of 163 men (74%) sought treatment, the majority (69%) with radical prostatectomy. The treatment-free survival rate at 5 years was 22% for 3þ4 and 10% for 4þ3 or greater upgrade. Upgrade to 4þ3 or greater, higher prostate specific antigen density at diagnosis and shorter time to initial upgrade were associated with treatment. At surgical pathology 34% of cancers were downgraded while 6% were upgraded. Cancer volume at initial upgrade was associated with adverse pathological outcome at surgery (OR 3.33, 95% CI 1.19e9.29, p¼0.02). Conclusions: After Gleason score upgrade most patients elected treatment with radical prostatectomy. Among men who deferred definitive intervention, few experienced additional upgrading. At radical prostatectomy only 6% of cases were upgraded further and only tumor volume at initial upgrade was significantly associated with adverse pathological outcome. Key Words: watchful waiting, neoplasm grading, disease progression, prostatic neoplasms Abbreviations and Acronyms AS ¼ active surveillance BXn ¼ biopsy number after initial upgrade GS ¼ Gleason score PCa ¼ prostate cancer PSA ¼ prostate specific antigen PSAD ¼ PSA density PSM ¼ positive surgical margins RP ¼ radical prostatectomy Accepted for publication January 15, 2015. Study received institutional review board approval. Supported by the U.S. Department of Defense Prostate Cancer Research Program (W81XWH-13-2-0074 and W81XWH-11-1-0489). * Nothing to disclose. Financial interest and/or other relationship with Abbott Labs. Correspondence: Department of Urology, University of California, San Francisco, 1600 Divisadero St., Box 1695, San Francisco, California 94143-1695 (telephone: 415-353-7098; FAX: 415-353-9932; e-mail: pcarroll@urology. ucsf.edu). Financial interest and/or other relationship with NIH, Department of Defense, Myriad Genetics, AUA Update, Genomic Health and Medivation/Astellas. See Editorial on page 8. IN the contemporary PSA era the widespread use of PSA screening has contributed to an estimated reduction in prostate cancer mortality of 20% to 50%. 1,2 However, this has been at the expense of over diagnosis and subsequent overtreatment of cancers that may not have posed a significant threat to patients. Currently, although the American Urological Association and Editor s Note: This article is the second of 5 published in this issue for which category 1 CME credits can be earned. Instructions for obtaining credits are given with the questions on pages 268 and 269. 0022-5347/15/1941-0085/0 THE JOURNAL OF UROLOGY 2015 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARCH,INC. http://dx.doi.org/10.1016/j.juro.2015.01.077 Vol. 194, 85-90, July 2015 Printed in U.S.A. www.jurology.com j 85

86 GLEASON GRADE PROGRESSION AND OUTCOMES National Comprehensive Cancer Network recommendations have been designed to limit over detection, neither prevent it. 3,4 The long natural history of PCa and detection of potentially indolent tumors have led to the development of active surveillance as an initial management strategy. AS for low risk disease allows postponing or avoiding the possible morbidities and adverse quality of life consequences that may be associated with treatment. Disease progression is actively monitored while maintaining an opportunity for cure. With time, more than a third of patients will be reclassified as at higher risk and pursue treatment. 5,6 In most cases reclassification to higher risk is due to upgrading at repeat biopsy. 7 Most cohorts use grade progression to Gleason scores 3þ4 or higher as a trigger for intervention and grade progression on serial biopsy is strongly associated with time to treatment. 8 It remains unclear which parameters, including grade progression, truly identify patients who need immediate, active treatment. Although most recommend active treatment in case of progression on repeat biopsies, some patients elect to continue on AS. In this context we describe the characteristics, followup and outcomes of patients who continued on AS in spite of grade progression on surveillance biopsy or sought treatment. PATIENTS AND METHODS Study Cohort This was a retrospective observational study of patients from the University of California at San Francisco Urologic Oncology Database conducted under institutional review board approval. We included consented men followed for at least 6 months after diagnosis with lower risk localized PCa (GS 3þ3 or lower at diagnosis and clinical T1 or T2 tumor) who were initially managed with AS. Additional inclusion criteria were a minimum of 6 cores taken at diagnostic biopsy, 33% or less positive and single core positivity 50% or less. We further restricted the analysis to include patients who further upgraded to GS 3þ4 or greater while on AS (fig. 1). AS Protocol The recommended AS regimen at our institution consisted of digital rectal examination, PSA testing at 3-month intervals and transrectal ultrasound guided prostate biopsies generally performed at 12 to 24-month intervals. Repeat biopsies included at least 14 cores taken from all sextants and included anterior gland sampling. Intervention was offered to men who experienced significant clinical or biopsy progression beyond the inclusion criteria. Statistical Analysis Patient demographics and disease characteristics were compared using means, medians and contingency tables with p values based on t-statistics, Wilcoxon signed rank test and chi-squared tests. Treatment after initial upgrade and adverse outcome (defined as having stage pt3a or greater and/or positive lymph nodes) among the subset of patients who underwent RP were used as the response variables. As PSM may be surgeon and technique dependent, a separate model for PSM as an adverse outcome was performed. We defined initial upgrade on any surveillance biopsy as an increase from GS 3þ3 or lower at diagnosis to 3þ4 or greater. We then categorized men with initial upgrade into 2 groups: 3þ4 and 4þ3 or greater. The first biopsy after initial upgrade was referred to as BX1, the second biopsy after initial upgrade as BX2, and so forth. The initial upgrade was used as the primary independent variable. Other covariates included in the model were age, PSAD, prostate volume and CAPRA (Cancer of the Prostate Risk Assessment) score risk classification at diagnosis, PSA before last biopsy or RP, time from diagnosis to initial upgrade, time from initial upgrade to RP and cancer volume (percentage of positive total cores and maximum of a single core positive) at initial upgrade. We also examined subsequent GS changes when compared to the initial upgrade defined as a further upgrade, downgrade, no change or benign. Biochemical failure was defined as achieving a PSA of 0.2 ng/ml or greater on 2 consecutive post-rp tests. We used Kaplan-Meier curves to estimate the probability of being treatment-free stratified by initial upgrade groups 3þ4 and 4þ3 or greater, and the log rank test to determine whether the difference between the 2 groups was statistically significant. Time was censored if the treatment event had not been observed for an individual. Cox proportional hazard regression analysis was used to examine factors associated with treatment. Multivariate logistic regression analysis was used to evaluate the effect of covariates on the probability of having adverse outcome at RP. Statistical analyses were performed using SASÒ software version 9.3. RESULTS A total of 525 men met the initial inclusion criteria. Mean age at diagnosis was 61.6 years (range 42 to 82) with median PSA 5.2 ng/ml (IQR 4.0e7.2). Most men were diagnosed with clinical stage T1 (70%) and low CAPRA score (91%). Median followup after diagnosis was 59 months (IQR 37e89). The final cohort for analysis comprised 219 (42%) men who experienced a GS upgrade on followup biopsy, of whom 150 (68%) were upgraded to 3þ4 and 69 (32%) upgraded to 4þ3 or greater. There were 26 men who had no subsequent biopsy or treatment at time of analysis (17 with pattern 3þ4 and 9 with 4þ3 or greater). Of these, 17 were followed with PSA and/or ultrasound and the remaining 9 did not have a followup visit after upgrading. Median time to the initial upgrade was 23 months (IQR 11e49). Demographics and tumor characteristics are shown in the supplementary table (http://jurology.com/). Among 133 men who had an initial upgrade to 3þ4, 43% (57) stayed on AS and underwent BX1. On

GLEASON GRADE PROGRESSION AND OUTCOMES 87 Figure 1. Natural history of 219 patients on AS who started with GS 3þ3 and upgraded to 3þ4 or higher on surveillance biopsy. UCSF, University of California, San Francisco. UODB, Urology Oncology Database at UCSF. BX1, 7 men experienced further upgrade, while 23 were downgraded. Of those who underwent BX1 53% (23) remained on AS and underwent BX2, while 2 additional men were upgraded and 14 were downgraded. Twelve men had BX3 (2 were upgraded, 5 were downgraded) and 6 underwent BX4 (4 were downgraded) (fig. 2). Among the 60 men who had an initial upgrade to 4þ3 or greater 18% (11) stayed on AS and had BX1, of whom 3 were further upgraded. Four men continued on AS to BX2, of whom 1 was further upgraded and only 1 had BX3 (fig. 2). Of the total cohort of 219 patients 74% (163) eventually sought treatment, 69% (113) of whom elected RP while 28% (46) received radiation based therapy, 1% (2) received cryotherapy and 1% (2) received androgen deprivation therapy. Treatment data were incomplete for 2 patients and, therefore, were excluded. Mean age at treatment was 66 years (range 48.0 to 86.4). Median time to treatment was 28 months (IQR 16e52) and did not differ by extent of Gleason upgrade (3þ4vs4þ3 or greater) at initial biopsy. Of the men who sought RP 64% had GS 3þ4 and 36% had 4þ3 or greater as the highest grade before treatment. Among the 163 men treated 77% (125) underwent intervention soon after the initial upgrade while 23% (38) were treated after a further period of AS. The treatment-free survival at 5 years was 22% after initial upgrade to 3þ4 and 10% after upgrade to 4þ3 or greater (log rank p¼0.0015, fig. 3). Cox proportional hazard regression analysis, modeling time to treatment, showed that men with

88 GLEASON GRADE PROGRESSION AND OUTCOMES Table 1. Cox proportional hazard regression analysis modeling time to treatment after initial GS upgrade to 3þ4 or higher Covariate HR 95% CI p Value Age at diagnosis 1.00 0.98e1.03 0.97 PSA before last biopsy or treatment (ng/ml/ml) 1.03 0.85e1.24 0.79 PSAD at diagnosis (ng/ml 2 ) 1.43 1.04e1.97 0.03 Prostate vol at diagnosis (cc) 0.65 0.40e1.03 0.07 Time to initial upgrade 0.99 0.98e0.997 0.004 Initial upgrade category (4þ3 or higher vs 3þ4) 1.93 1.34e2.78 0.0004 CAPRA risk at diagnosis (intermediate vs low risk) 1.60 0.92e2.78 0.09 Figure 2. Changes in Gleason grade (relative to initial upgrade category) with surveillance biopsies among cases upgraded to pattern 3þ4 vs upgraded to 4þ3 or higher. UG, upgrade. DG, downgrade. initial upgrade to 4þ3 or greater were more likely to be treated compared to 3þ4 (HR 1.93, 95% CI 1.34e2.78, p¼0.0004). Higher PSAD at diagnosis (HR 1.43, 95% CI 1.04e1.97, p¼0.03) and shorter time to initial upgrade (HR 0.99, 95% CI 0.98e0.99, p¼0.004) were also associated with treatment (table 1). In a separate Cox model downgrading on surveillance biopsies was associated with lower risk of treatment (HR 0.09, 95% CI 0.03e0.27, p <0.0001). At RP Gleason score remained the same for 60%, was downgraded for 34% and upgraded for 7% compared to the highest preoperative GS (table 2). With regard to adverse outcome at RP, pt3 was observed in 27% of men in the 3þ4 group and 50% of the 4þ3 or greater group. Of the men 21% (16) in the 3þ4 group had PSM, including 9 with pt2 and 7 with pt3 disease. Of these, 50% were unifocal (range 1 to 2 locations) and the median length was 2 mm. Among cases upgraded to 4þ3 or greater 24% (8) had PSM, of which 3 had pt2 and 5 had pt3 Figure 3. Kaplan-Meier curves of treatment-free survival among 219 men who experienced initial Gleason upgrade to 3þ4 vs 4þ3 or higher (log rank p¼0.0015). disease (all were multifocal, range 2 to 4 locations) with median length of 4.5 mm. Only 1 patient in the 4þ3 or greater group had positive lymph nodes. At mean followup of 15 months 8% of our cohort who had RP experienced biochemical failure. On logistic regression analysis only cancer volume at initial upgrade (greater than 33% positive cores or greater than 50% positive on a single core compared to 33% or less positive cores and 50% or less positive on a single core) was significantly associated with adverse pathological outcome at RP (OR 3.33, 95% CI 1.19e9.29, p¼0.02, table 3). None of the patient and tumor characteristics was associated with PSM at RP (table 4). DISCUSSION Overtreatment of clinically insignificant PCa is a concern for early detection protocols. Previous studies have used variables such as clinical stage, Gleason grade, cancer volume, the number of positive biopsies and the extent of cancer in each core to estimate the risk of significant cancer, and thereby select patients with tumors amenable to AS. Of these, Gleason grade may offer the strongest prognostic value and informs the regard for grade progression as a critical event during AS, though a growing clinical experience with intermediate risk patients may question rigid inclusion criteria. 9e11 To evaluate the natural history of untreated grade progression we evaluated a cohort of men who experienced Gleason upgrading during AS. Among those with an initial upgrade pattern of 3þ4 40% were downgraded on biopsy on BX1 and 61% were downgraded on BX2. While for those with 4þ3 or greater at upgrade, 7 were downgraded at BX1 and 3 were upgraded. Of 163 men (74%) who sought treatment after upgrade 125 (77%) received intervention without further biopsy. The extent of Gleason upgrade (4þ3 or greater in comparison to 3þ4), shorter time to upgrade and PSAD at diagnosis were all significantly associated with receiving treatment after the initial upgrade. Gleason grade progression on serial biopsy has been shown to have the strongest association with time to treatment in our study as well as others. 9,10 In

GLEASON GRADE PROGRESSION AND OUTCOMES 89 Table 2. Surgical pathology for 111 men who underwent RP No. Initial Upgrade 3þ4 (%) No. Initial Upgrade 4þ3 or Greater (%) p Value Pathological T-stage: pt2 56 (73) 17 (50) 0.02 pt3 21 (27) 17 (50) Pathological N-stage (pn1): Yes 0 (0) 1 (3) <0.001 No 20 (26) 20 (59) Not evaluated 57 (74) 13 (38) Pos margins: Yes 16 (21) 8 (24) 0.98 No 61 (79) 26 (76) Extracapsular extension: Yes 18 (23) 17 (50) 0.005 No 59 (77) 17 (50) Seminal vesicle involvement: Yes 4 (5) 2 (6) 0.88 No 73 (95) 32 (94) Surgical CAPRA:* Low (0e2) 47 (61) 10 (30) <0.001 Intermediate (3e5) 29 (38) 18 (55) High (6e12) 1 (1) 5 (15) Pathological GS:* Upgrade 7 (10) 0 (0) 0.04 Downgrade 9 (13) 28 (70) Unchanged 54 (77) 12 (30) Adverse outcome at RP (pt3a or greater and/or pn1): Yes 21 (27) 18 (53) 0.009 No 56 (73) 16 (47) * Pathological GS was not identified for 1 patient treated with neoadjuvant androgen deprivation. addition, downgrading on surveillance biopsy was associated with lower risk of treatment. Previous reports have shown that the absence of cancer on the first surveillance biopsy was associated with lower risk of further progression. 12 Nevertheless, change in GS during surveillance may not always represent true disease progression but rather under detection at the time of initial biopsy. This could be due to differences in grading from pathologist assessment or, perhaps more likely, by initial under sampling, where smaller volumes of more aggressive disease may be missed on biopsy. 13 Promise in reducing the risk of under sampling may be held by improved sampling methods including the use of extended Table 3. Logistic regression modeling adverse pathological outcome, defined as greater than pt3a and/or positive lymph nodes in 111 men who underwent RP Covariate OR 95% CI p Value Age at diagnosis per 5-yr increment 1.42 0.95e2.12 0.09 PSAD at diagnosis 0.89 0.27e2.91 0.85 Time to initial upgrade 1.00 0.98e1.03 0.81 Time from initial upgrade to RP 1.00 0.98e1.03 0.98 Initial upgrade category 4þ3 or higher vs 3þ4 2.50 0.91e6.88 0.08 Ca vol at initial upgrade (greater than 33% pos 3.33 1.19e9.29 0.02 cores or greater than 50% pos on single core vs 33% or less pos cores þ 50% or less pos on single core) Prostate vol at diagnosis 0.77 0.15e3.90 0.75 CAPRA risk at diagnosis (intermediate vs low) 1.03 0.2e5.39 0.98 Table 4. Logistic regression modeling positive surgical margins in 111 men who underwent RP Covariate OR 95% CI p Value Age at diagnosis per 5-yr increment 1.55 0.78e3.06 0.21 PSAD at diagnosis 4.39 0.55e35.41 0.16 Time to initial upgrade 0.96 0.90e1.02 0.17 Time from initial upgrade to RP 1.01 0.97e1.05 0.62 Initial upgrade category 4þ3 or higher vs 3þ4 0.38 0.03e4.24 0.43 Ca vol at initial upgrade (greater than 33% pos 0.45 0.09e2.21 0.33 cores or greater than 50% pos on single core vs 33% or less pos cores þ 50% or less pos on single core) CAPRA risk at diagnosis (intermediate vs low) 0.20 0.01e4.82 0.32 Prostate vol at diagnosis 1.81 0.15e22.3 0.64 Patients with pt3a or greater disease and/or positive lymph nodes were excluded from analysis. patterns for surveillance biopsies, use of experienced uropathologists, anterior gland sampling and the use of image targeted biopsy. 14 To date none of the nomograms designed to predict indolent cancers has shown added benefit predicting grade progression at RP, limiting their usefulness in patient counseling and decision making. 15 Most of the men who sought treatment in our study underwent RP. Pathological outcomes did not appear to differ significantly from men with equivalent risk treated immediately. Only 6% were upgraded at surgery, a comparatively low proportion to studies of patients undergoing RP for newly diagnosed low grade disease, a difference which may be explained by the extended sampling of these men on followup biopsy. 16 In our analysis only biopsy tumor volume (greater than 33% of cores or greater than 50% of a single core) was associated with adverse outcome at RP. Despite the wide confidence interval (1.19e9.29) observed in our study, these findings appear to agree with prior reports suggesting that cancer volume may be associated with undetected high grade cancers. 13,17 Our findings are consistent with previous studies reporting no significant difference in the pathological outcome at RP or biochemical progression between low risk patients who had immediate RP vs those who deferred treatment. 18 Of our cohort 22% had PSM, which is similar to our larger RP cohort. We did not find a significant association between patient and disease characteristics and PSM. However, patients with higher upgrade category (4þ3 or greater vs 3þ4) and PSM tended to have multifocal and larger extent of PSM. Patel et al found that for localized PCa, preoperative PSA was the most important predictor of PSM at RP, in addition to higher body mass index and lower prostate weight. 19 On the other hand, biopsy GS did not have a significant impact on univariate or multivariate analyses. In limited followup the observed 8% biochemical failure rate in our study appears consistent with prior longitudinal series,

90 GLEASON GRADE PROGRESSION AND OUTCOMES yet is inadequate to draw any conclusions regarding the oncologic outcomes of these patients. Although this study contributes to the understanding of the natural history of grade progression for men on AS, it has several limitations. The retrospective design of the study and the limited sample size may limit the conclusions drawn from the study. We could not account for all factors that could influence decision making. Adverse outcome was defined in terms of pathological findings at RP rather than clinical end points. While pathological and preoperative factors have previously been shown to predict long-term outcomes, additional followup of this subset of patients will provide us with better estimates of disease specific and recurrence-free survival in the future. The findings of the current study raise the possibility that not all men who experience grade progression on AS need to be treated immediately. The majority of our patients who remained on AS did not have further grade progression and only 6% of our patients had progression on surgical pathology. Among this select group of patients, PSA kinetics and nomograms likely are not reliable enough to differentiate who needs treatment. 20 Multiparametric magnetic resonance imaging, and molecular and gene expression profiling of biopsy tissue appear to be promising tools to improve the differentiation between indolent and aggressive tumors, and may assist in identifying the subset of men who may benefit most from early treatment and those that can be put safely under surveillance. 21,22 CONCLUSIONS In this AS cohort most patients elected local treatment after upgrade. Upgrade to GS 4þ3 or greater, shorter initial time to upgrade and higher PSAD at diagnosis were associated with receiving treatment. While few men who initially deferred treatment remained on AS at the last followup, few experienced additional upgrading. Furthermore, most men who elected RP as treatment had a stable or downgraded GS on surgical pathology. Apart from cancer volume at initial upgrade, none of the patient or tumor characteristics was significantly associated with an adverse outcome at RP. With further development of new tools to determine the lethal potential of PCa, it may be safe to continue to follow men who experience GS upgrade. REFERENCES 1. Bray F, Lortet-Tieulent J, Ferlay J et al: Prostate cancer incidence and mortality trends in 37 European countries: an overview. Eur J Cancer 2010; 46: 3040. 2. Hugosson J, Carlsson S, Aus G et al: Mortality results from the Goteborg randomised population-based prostate-cancer screening trial. Lancet Oncol 2010; 11: 725. 3. Mohler JL, Kantoff PW, Armstrong AJ et al: Prostate cancer, version 2.2014. J Natl Compr Canc Netw 2014; 12: 686. 4. Carter HB, Albertsen PC, Barry MJ et al: Early detection of prostate cancer: AUA guideline. J Urol 2013; 190: 419. 5. Sheridan TB, Carter HB, Wang W et al: Change in prostate cancer grade over time in men followed expectantly for stage T1c disease. J Urol 2008; 179: 901. 6. Klotz L, Zhang L, Lam A et al: Clinical results of long-term follow-up of a large, active surveillance cohort with localized prostate cancer. J Clin Oncol 2010; 28: 126. 7. Klotz L: Active surveillance for favorable-risk prostate cancer: background, patient selection, triggers for intervention, and outcomes. Curr Urol Rep 2012; 13: 153. 8. Carter HB, Walsh PC, Landis P et al: Expectant management of nonpalpable prostate cancer with curative intent: preliminary results. J Urol 2002; 167: 1231. 9. Epstein JI, Walsh PC, Carmichael M et al: Pathologic and clinical findings to predict tumor extent of nonpalpable (stage T1c) prostate cancer. JAMA 1994; 271: 368. 10. Cheng L, Poulos CK, Pan CX et al: Preoperative prediction of small volume cancer (less than 0.5 ml) in radical prostatectomy specimens. J Urol 2005; 174: 898. 11. Cooperberg MR, Cowan JE, Hilton JF et al: Outcomes of active surveillance for men with intermediate-risk prostate cancer. J Clin Oncol 2011; 29: 228. 12. Cary KC, Cowan JE, Sanford M et al: Predictors of pathologic progression on biopsy among men on active surveillance for localized prostate cancer: the value of the pattern of surveillance biopsies. Eur Urol 2014; 66: 337. 13. Corcoran NM, Hovens CM, Hong MK et al: Underestimation of Gleason score at prostate biopsy reflects sampling error in lower volume tumours. BJU Int 2012; 109: 660. 14. Emiliozzi P, Maymone S, Paterno A et al: Increased accuracy of biopsy Gleason score obtained by extended needle biopsy. J Urol 2004; 172: 2224. 15. Wang SY, Cowan JE, Cary KC et al: Limited ability of existing nomograms to predict outcomes in men undergoing active surveillance for prostate cancer. BJU Int 2014; 114: E18. 16. Freedland SJ, Kane CJ, Amling CL et al: Upgrading and downgrading of prostate needle biopsy specimens: risk factors and clinical implications. Urology 2007; 69: 495. 17. Dong F, Jones JS, Stephenson AJ et al: Prostate cancer volume at biopsy predicts clinically significant upgrading. J Urol 2008; 179: 896. 18. van den Bergh RC, Steyerberg EW, Khatami A et al: Is delayed radical prostatectomy in men with low-risk screen-detected prostate cancer associated with a higher risk of unfavorable outcomes? Cancer 2010; 116: 1281. 19. Patel VR, Coelho RF, Rocco B et al: Positive surgical margins after robotic assisted radical prostatectomy: a multi-institutional study. J Urol 2011; 186: 511. 20. Iremashvili V, Burdick-Will J and Soloway MS: Improving risk stratification in patients with prostate cancer managed by active surveillance: a nomogram predicting the risk of biopsy progression. BJU Int 2013; 112: 39. 21. van den Bergh RC, Ahmed HU, Bangma CH et al: Novel tools to improve patient selection and monitoring on active surveillance for low-risk prostate cancer: a systematic review. Eur Urol 2014; 65: 1023. 22. Haffner MC, De Marzo AM, Yegnasubramanian S et al: Diagnostic challenges of clonal heterogeneity in prostate cancer. J Clin Oncol 2015; 33: e38.