Oncology Long-Term Follow-up of Men With Isolated High-grade Prostatic Intra-epithelial Neoplasia Followed by Serial Delayed Interval Biopsy

Guilherme Godoy, George J. Huang, Trushar Patel, and Samir S. Taneja

OBJECTIVES To analyze the outcomes of serial delayed interval biopsy (DIBx) in men with isolated high-grade prostatic intraepithelial neoplasia (HGPIN). The natural history of isolated HGPIN is poorly dened. Since January 2000, we have monitored men with isolated HGPIN using DIBx every 3 years, regardless of the change in prostate-specic antigen (PSA) level. The institutional biopsy records from 1996 onward were reviewed to identify the men with isolated HGPIN found on 12-core needle biopsy specimens who had undergone a minimum of 1 DIBx in our follow-up strategy. Patient age, biopsy and prostatectomy pathologic outcomes, and longitudinal PSA measurements were recorded. A total of 112 men had undergone a rst DIBx and 47 had undergone a second DIBx at the last follow-up examination at a mean of 34.4 and 66.2 months after the HGPIN diagnosis, respectively. Prostate cancer was found in 25 (22.3%) of 112 men and 11 (23.4%) of 47 men at DIBx-1 and DIBx-2, respectively. The PSA velocity was not predictive of cancer during short-term follow-up. Of the men diagnosed with cancer, 63.6% had a Gleason score of 7, and 9 (81.8%) of 11 men had clinically signicant disease (Gleason score of 7 and/or 5% cancer volume) at surgery. All cancers were organ conned at and surgery. Men with isolated HGPIN have a continued risk of developing prostate cancer during long-term follow-up, regardless of the changes in the serum PSA level. Collectively, the relatively high likelihood of organ connement and clinically signicant cancer suggest empiric DIBx every 2-3 years could be a valuable tool in the follow-up of men with isolated HGPIN found by extended core biopsy. UROLOGY xx: xxx, xxxx. 2010 Elsevier Inc.

METHODS

RESULTS

CONCLUSIONS

igh-grade prostatic intraepithelial neoplasia (HGPIN) has been dened by architecturally benign ducts and acini lined by epithelial cells with nuclear changes similar to that of prostate cancer.1 It will be found in 4.4%-25% of men who undergo transrectal needle biopsy for elevated prostate-specic antigen (PSA) levels or abnormal digital rectal examination (DRE) ndings.2 Recently, the mean incidence of HGPIN has been reported in the lower limits of this range, approaching 5%.3 Substantial evidence has suggested a link between HGPIN and prostate cancer. In 85% of radical prostatectomy specimens, HGPIN has been found to coexist multifocally with prostate cancer.4 Furthermore, studies have identied similar molecular alterations in HGPIN and prostate cancer, as evidenced
G. Godoy received a Bruce and Cynthia Sherman Fellowship in Urologic Oncology. From the Division of Urologic Oncology, Department of Urology, New York University School of Medicine, New York, New York Reprint requests: Samir S. Taneja, M.D., New York University Urology Associates, 150 East 32nd Street, 2nd Floor, New York, NY 10016. E-mail: samir.taneja@nyumc.org Submitted: March 23, 2010; accepted (with revisions): July 2, 2010

by numerous studies that analyzed the immunoreactivity prole of basal cells and secretory cells using immunohistochemistry markers, such as CK903, p63, and -methylacyl co-enzyme A-racemase, with ndings that supported the role of HGPIN as a premalignant lesion.5-8 The nding of isolated HGPIN on needle biopsy cores requires that the managing physician both rule out coexistent prostate cancer at presentation and monitor the patient for progression to prostate cancer in the long term. In a previous era, when sextant biopsies, instead of extended biopsies ( 10 cores) were routinely performed, the likelihood of nding cancer on immediate repeat biopsy was reported to be 27%-100%.9-13 These early ndings likely reected inadequate sampling using the sextant protocol, because sextant biopsies have been noted to miss signicant percentages of cancer in the lateral peripheral zone.14 We have more recently demonstrated that when HGPIN is diagnosed using 12-core sampling, the likelihood of nding cancer on a repeat 12-core biopsy within 1 year was only 2.3%.15 Because of this observation and our subsequent clinical experience,
0090-4295/xx/$34.00 doi:10.1016/j.urology.2010.07.519 1

2010 Elsevier Inc. All Rights Reserved

we have not routinely recommended immediate repeat biopsy for men with isolated HGPIN diagnosed using 12-core biopsy.16-18 The ideal long-term follow-up strategy for isolated HGPIN is not clear, because its natural history has been poorly described. We previously reported, in a small series, that empiric biopsy of men with isolated HGPIN after a xed delayed interval of 3 years demonstrated a 25.8% risk of cancer, independent of changes in the serum PSA level.18 Because of these observations, since January 2000, we have monitored all men diagnosed with isolated HGPIN in our practice using serial delayed interval biopsy (DIBx) every 2-3 years. In the present study, we have reported the outcomes of our management strategy and the nature of the cancers identied during longterm follow-up.

Table 1. Baseline characteristics Characteristic Mean age (y) Age range (y) PSA (ng/mL) Mean Median Range Mean TRUS volume (cm3) Value 64.57 41-78 6.25 3.40 5.70 0.22-20.30 39.27 19.39

PSA, prostate-specic antigen; TRUS, transrectal ultrasonography.

performed using the Statistical Package for Social Sciences, version 13.0 (SPSS, Chicago, IL).

RESULTS
The baseline characteristics of the 112 men analyzed are summarized in Table 1. The mean age of the cohort at the HGPIN diagnosis was 64.6 years (range 41-78). Of these 112 patients, 91 had initially undergone biopsy because of an elevated PSA level, 9 because of abnormal DRE ndings, 9 because of both an elevated PSA level and abnormal DRE examination, and 3 for other reasons. The median interval from HGPIN diagnosis (baseline) to DIBx-1 was 34.4 months (range 12.2-71.1). The results from DIBx-1 are summarized in Table 2. At DIBx-1, 25 men were diagnosed with cancer (22.3%), 34 with isolated HGPIN (30.3%), and 53 with benign tissue only (47.3%). No signicant differences were found in the mean age (P .353) between those diagnosed with cancer and those without. A comparison of the mean PSA at DIBx-1 (P .469), mean PSA slope from baseline to DIBx-1 (P .695), and mean PSAV (P .665) during this interval revealed no signicant differences between those with and without cancer found on DIBx-1. A PSAV greater than the traditional cutoff value of 0.75 ng/mL/y20 was not associated with an increased likelihood of cancer diagnosis at DIBx-1 (P .821). Of the 87 patients with negative ndings on biopsy at DIBx-1, 47 had achieved adequate follow-up to undergo DIBx-2 at analysis. For these 47 men, DIBx-2 was performed at a median interval of 66.2 and 32.7 months from baseline and DIBx-1, respectively. The results at DIBx-2 are also summarized in Table 2. Of the 47 patients, 11 (23.4%) had prostate cancer, 13 (27.6%) had isolated HGPIN, and 23 (48.9%) had benign tissue only. Of the men who underwent DIBx-2, 6 (30.0%) of 20 with HGPIN found on DIBx-1 had prostate cancer compared with 5 (18.5%) of 27 with benign tissue only on DIBx-1 (P .489). Overall, at a median follow-up of 46.6 months (mean 49.9, range 13.9-109.6), 36 (32.1%) of the total 112 patients had been found to have cancer. We performed a subset analysis, stratifying the patients by biopsy location and found that the cumulative rate of HGPIN and cancer and rate at both DIBx-1 and DIBx-2 were similar.
UROLOGY xx (x), xxxx

MATERIAL AND METHODS

After approval by the institutional review board, the records from a university faculty practice and afliated Veterans Affairs urology clinic were reviewed to identify men with isolated HGPIN found on needle biopsy who agreed to surveillance using DIBx. A total of 112 men (66 from the private practice and 46 from the Veterans Affairs clinic) were identied as having completed 1 DIBx (DIBx-1), dened as repeat biopsy 12 months after the rst diagnosis of isolated HGPIN. Of the 87 men without cancer on DIBx-1, 47 had undergone a second DIBx (DIBx-2) at analysis. The men who had undergone DIBx within 1 year of the HGPIN diagnosis were excluded from the analysis. The 40 men who had not undergone DIBx-2 had been lost to follow-up, had refused, were deemed of inappropriate age to undergo repeat biopsy, or had not reached an adequate follow-up interval to undergo DIBx-2. The charts of the 112 men meeting the inclusion criteria were reviewed to record the age, date, PSA level, DRE, and pathologic ndings at the HGPIN diagnosis and all subsequent biopsies. The interim PSA data between biopsies were also collected. All biopsies were performed using a previously described technique of 12-core transrectal ultrasound-guided biopsy with sampling of the far lateral and parasagittal peripheral zones.19 A total of 36 men underwent transition zone sampling with 6-8 cores directed at the anterior parasagittal, periurethral region on either side.19 All biopsy cores were examined by 1 of 2 academic pathology services (at the university hospital or the Veterans Affairs clinic). When subsequent radical prostatectomy was performed on-site, the surgical pathology report was reviewed to record the tumor stage, Gleason score, and laterality/tumor side. The PSA slope between 2 different points was calculated by subtracting the PSA of the earlier point from the PSA of the later point. The PSA velocity (PSAV) for a given interval was estimated by the slope divided by the interval, using PSA values 6 months apart. The standard PSAV cutoff values were used to assess the predictive value of the PSAV in cancer detection.20 Students t test was used to compare the mean values of continuous data, such as age and PSA values between the cancer and noncancer patient groups. The Fisher exact test was used to evaluate the categorical data. All statistical analyses were 2-sided, with the signicance level set at P .05, and were 2

Table 2. Summary of results at DIBx-1 and DIBx-2 Parameter DIBx-1 Patients (n) Mean age (y) Mean baseline PSA (ng/mL) Mean DIBx-1 PSA (ng/mL) Mean PSA slope (ng/mL) Mean PSAV (ng/mL/y) PSA cutoff of 0.75 ng/mL/y DIBx-2 Patients (n) Mean age (y) Mean baseline PSA (ng/mL) Mean DIBx-1 PSA (ng/mL) Mean DIBx-2 PSA (ng/mL) From baseline Mean PSA slope (ng/mL) Mean PSAV (ng/mL/y) From DIBx-1 Mean PSA slope (ng/mL) Mean PSAV (ng/mL/y) PSA cutoff of 0.75 ng/mL/y From baseline From DIBx-1 No Cancer 87 (77.6) 64.93 7.75 6.12 3.59 7.72 5.94 1.62 4.24 0.75 1.75 36 (76.59) 63.75 7.94 6.23 3.74 7.61 5.01 7.16 4.67 0.93 0.18 0.74 0.28 4.65 0.98 5.08 1.68 Cancer 25 (22.3) 63.32 7.06 6.69 2.66 8.70 5.87 2.01 4.52 0.92 1.92 11 (23.40) 64.36 6.71 7.41 4.55 11.27 11.18 12.83 16.54 6.68 0.95 5.07 1.33 15.00 1.78 11.86 2.23 P Value* .353 .466 .469 .695 .665 .821 .818 .386 .339 .310 .260 .078 .183 .022 .031 .021

DIBx, delayed interval biopsy; PSA, prostate-specic antigen; PSAV, PSA velocity. * Two-sided t test.

Table 3. Analysis of sensitivity, specicity and predictive values of PSA velocity using traditional cutoff of 0.75 ng/mL/y Analysis Sensitivity (%) Specicity (%) PPV (%) NPV (%) P value (Fishers exact test) DIBx-1 40.0 67.5 27.0 78.9 0.63 DIBx-2 From Baseline 60.0 80.5 46.1 87.9 0.02 DIBx-2 From DIBx-1 66.7 76.5 42.8 89.6 0.02

DIBx, delayed interval biopsy; PPV, positive predictive value; NPV, negative predictive value.

The PSA level at baseline and DIBx-2 did not differ signicantly between those with cancer diagnosed and those not (P .386 and P .310, respectively). When evaluating the PSA increase from DIBx-1 to DIBx-2, only the PSAV (P .022) was increased in the men with prostate cancer compared with those without. When evaluating the PSA increase from baseline to DIBx-2, both the PSA slope (P .260) and PSAV (P .078), were not greater among those with cancer. Stratication by the traditional cutoff value of 0.75 ng/mL/y predicted a greater risk of cancer from both baseline (P .031) and DIBx-1 (P .021). When using a PSAV cutoff of 0.75 ng/mL/y, the sensitivity and specicity for cancer detection was 66.7% and 76.5%, respectively (Table 3). No histopathologic differences were seen between the men diagnosed with prostate cancer at DIBx-1 and DIBx-2, with regard to Gleason score, disease volume, or nal surgical pathologic ndings. Complete biopsy pathologic data were available for 34 of 36 patients diagnosed
UROLOGY xx (x), xxxx

with cancer during follow-up; 3 were diagnosed with Gleason score 5 disease, 16 with Gleason score 6, 11 with Gleason score 7, 2 with Gleason score 8, and 2 with Gleason score 9. None of the patients developed new abnormalities palpable by DRE during follow-up. The transition zone was sampled in 37 patients during follow-up (36 with transition zone-directed needle biopsies and 1 with transurethral resection) at the discretion of the managing urologist. Cancer was noted in the transition zone in 3 (8.1%) of 37 patients. In all 3 cases, the cancer was limited to the transition zone. At prostate cancer diagnosis, all 36 patients had clinically localized disease. All 11 of the 36 patients who subsequently underwent radical retropubic prostatectomy at our institution had organ-conned disease (pT2) on nal pathologic examination. Seven patients had Stage pT2b-T2c disease, 8 had a signicant cancer volume ( 5%), and 7 (63.6%) had Gleason score 7. Only 2 men had clinically insignicant prostate cancer, dened by a Gleason score 7 and cancer volume 5% (Table 4).

COMMENT
Our institution has previously reported that although only 2.4% of men with isolated HGPIN at the baseline 12-core biopsy were diagnosed with cancer when undergoing repeat biopsy within 1 year, 25.8% of men were found to have cancer if the rst repeat biopsy was performed 3 years after the HGPIN diagnosis.15,18 To our knowledge, the present series represents the longest biopsy follow-up of men with isolated HGPIN reported to date. Most published data have reported on the outcomes
3

Table 4. Prostate cancer characteristics Biopsy Pathologic Findings (n 19) 12 4 3 13 13 3 3 16 11 2 2 2 18 6 34 2 0 0 Prostatectomy Pathologic Findings (n 11) 3 0 4 6 (all 3 1 0 2 3 5 2 (T2) 2 2 5

Characteristic Cores (n) 1 2 3 Tumor location (n 29) Sextant Far lateral Transition zone Gleason score (n 34) 5 6 7 8 9 Disease volume* (n 26) 5% 5%-10% 10% Stage (n 36) T1c T2a T2b T2c

4)

* Prostatectomy pathology data unavailable for 1 patient.

of biopsies performed within the immediate follow-up period after the initial diagnosis. In the present study, all men underwent repeat biopsy at a median of 34.4 months. At the last follow-up, 47 men had undergone a second follow-up biopsy at a median follow-up of 66.2 months from the initial HGPIN diagnosis. The remaining patients will undergo DIBx-2 on reaching the appropriate follow-up interval. Izawa et al21 recently published their experience with a similar population. They noted a much lower long-term cancer detection rate, with cancer noted in only 1 of 8 patients who had undergone a third prostate biopsy.21 It is unclear, however, when the third biopsies were performed, relative to the baseline biopsy. In addition, their biopsy strategy involved sampling an average of only 7 cores (range 2-8). The likelihood of a delayed cancer diagnosis in men with isolated HGPIN was reported to be 19% (4 of 21) at repeat biopsy at 36 months of follow-up by Abdel-Khalek et al22 in 2004. All patients in that series, however, had undergone repeat biopsy because of either concerning changes found on DRE or an increased PSAV, unlike our series, in which men underwent repeat biopsy regardless of changes in the DRE ndings or PSA level. The results of our study have suggested that the likelihood of cancer on delayed interval biopsy is not predicted by PSA change, occurs at DIBx-2 independent of pathologic demonstration of HGPIN on DIBx-1, and does not appear to be greater among men undergoing transition zone sampling on DIBx. Additionally, the chosen interval for biopsy appears validated by the
4

demonstration of clinically and/or pathologically organ-conned disease in all men diagnosed with cancer during long-term follow-up and the detection of clinically signicant disease in most men. Although earlier biopsy or more aggressive sampling might result in increased cancer detection, neither appears warranted according to the demonstrated pathologic outcomes in our study. It has been reported that HGPIN by itself does not elevate the serum PSA level signicantly.23 We had previously concluded that the change in PSA alone was not useful for the detection of prostate cancer in men with a history of HGPIN, because only 25% of patients diagnosed with cancer at repeat biopsy had a change in the PSA level 1 ng/mL during a 3-year follow-up interval.18 In the present study, we noted that 13 (52%) of 25 men found to have cancer on DIBx-1 had a change in PSA 1 ng/mL. Although we again noted no relationship between cancer detection and PSAV in the 3-year follow-up period, the PSAV became a signicant predictor of cancer when evaluating men diagnosed with cancer on DIBx-2. This observation suggests that the PSAV might become more predictive of cancer over time in men with isolated HGPIN, potentially suggesting that a critical volume of cancer is necessary before the trends in PSA become meaningful. A critical question that could not be answered by the present study was whether men with isolated HGPIN progress to cancer over time or have pre-existing cancers identied on repeat sampling. Clinically, this might not be a meaningful distinction, given the propensity of our patients to progress to clinically signicant disease; 41.6% and 63.6% of men with a cancer diagnosis on follow-up had a Gleason score of 7 on biopsy and the nal radical prostatectomy specimen, respectively. Although all cancers were clinically or pathologically organ conned at diagnosis, the patients who underwent radical prostatectomy generally demonstrated a clinically significant disease volume (81.8%). The similar percentage of men found to have cancer at DIBx-1 and DIBx-2 (22.3% and 23.4%, respectively) suggests either de novo cancer development or progression/growth of previously microscopic foci. If the detection of cancer on repeat biopsies is attributable solely to detection of a pre-existing lesion missed previously, one would expect the cancer detection rate on DIBx-2 to be signicantly lower than that of DIBx-1 and to serially decline with additional sampling. As such, we believe that isolated HGPIN found on extended core biopsy represents a surrogate for very earlystage prostate cancer and would advocate serial sampling as a part of a surveillance protocol, not unlike that for early-stage prostate cancer. A clear limitation of our evaluation was the absence of a control arm, in which men with benign ndings on the initial biopsy undergo serially biopsy sampling at similar 2-3-year intervals. Without this, it is difcult to rmly conclude that men with isolated HGPIN are at greater
UROLOGY xx (x), xxxx

long-term risk of cancer than men with an elevated PSA level and benign biopsy ndings alone. The most comprehensive data to date on long-term prostate cancer risk in an otherwise normal population have come from the control arm of the Prostate Cancer Prevention Trial. In that study, 24.4% of participants who underwent biopsy in the placebo arm were found to have cancer during a 7-year follow-up period at the rst sampling, rather than at repeat biopsy.24 Presumably, a second biopsy in these men, 2-3 years later would likely have resulted in a lower cancer detection rate than at the rst round of biopsies. In the Rotterdam screening cohort, men underwent 2 rounds of screening biopsies separated by 4 years. Among men with negative baseline screening ndings and a follow-up PSA level 4.0 ng/mL, the rate of cancer detection at the second biopsy was 8.3%.25 In that series, the baseline and repeat biopsies were sextant biopsies, suggesting that with baseline extended core biopsy, the rate of cancer detection at the second screening would likely have been lower. The most comparable cohort of men with benign biopsy ndings undergoing serial follow-up sampling is that of the Reduction by Dutasteride of Prostate Cancer Events trial, in which men with negative biopsy ndings and a PSA level of 4-10 ng/dL underwent repeat biopsy at 2 and 4 years of follow-up. In that study, men receiving placebo were demonstrated to have a 17.2% and 11.8% rate of cancer detection at 2 and 4 years, respectively.26 Despite these observations, in the absence of a control arm, we cannot denitively state that men with isolated HGPIN have a greater risk of cancer during follow-up from the ndings of the present study alone. Our study had several additional limitations. Because we were only able to include patients who agreed to DIBx, a selection bias might have resulted in identication of cancer among those who sought more rigorous follow-up, relative to those who did not. Additionally, the cancer detection rates over time were heavily inuenced by the timing of the prostate biopsy. More frequent biopsy, or more biopsy cores, might have resulted in a much greater risk of cancer detection by 5 years of follow-up than that predicted by our data. It would appear from the disease volume and disease stage detected with our follow-up biopsies that the proposed interval is reective of the clinically relevant natural history of the entity; however, that might have underestimated the true histologic incidence of prostate cancer.

clinically signicant cancer, and the lack of reliable indicators of cancer during follow-up suggest that active surveillance using empiric DIBx would be a valuable tool in the follow-up of men with isolated HGPIN found by extended core biopsy.

References
1. Bostwick DG, Montironi R. Prostatic intraepithelial neoplasia and the origins of prostatic carcinoma. Pathol Res Pract. 1995;191:828832. 2. Bostwick DG. Prostatic intraepithelial neoplasia is a risk factor for cancer. Semin Urol Oncol. 1999;17:187-198. 3. Epstein JI, Herawi M. Prostate needle biopsies containing prostatic intraepithelial neoplasia or atypical foci suspicious for carcinoma: implications for patient care. J Urol. 2006;175:820-834. 4. Qian J, Wollan P, Bostwick DG. The extent and multicentricity of high-grade prostatic intraepithelial neoplasia in clinically localized prostatic adenocarcinoma. Hum Pathol. 1997;28:143-148. 5. Pu YS, Luo W, Lu HH, et al. Differential expression of C-CAM cell adhesion molecule in prostate carcinogenesis in a transgenic mouse model. J Urol. 1999;162:892-896. 6. Rubin MA, Zhou M, Dhanasekaran SM, et al. Alpha-methylacyl coenzyme A racemase as a tissue biomarker for prostate cancer. JAMA. 2002;287:1662-1670. 7. Dhanasekaran SM, Barrette TR, Ghosh D, et al. Delineation of prognostic biomarkers in prostate cancer. Nature. 2001;412:822826. 8. Ayala AG, Ro JY. Prostatic intraepithelial neoplasia: recent advances. Arch Pathol Lab Med. 2007;131:1257-1266. 9. Aboseif S, Shinohara K, Weidner N, et al. The signicance of prostatic intra-epithelial neoplasia. Br J Urol. 1995;76:355-359. 10. Brawer MK, Bigler SA, Sohlberg OE, et al. Signicance of prostatic intraepithelial neoplasia on prostate needle biopsy. Urology. 1991; 38:103-107. 11. Keetch DW, Humphrey P, Stahl D, et al. Morphometric analysis and clinical followup of isolated prostatic intraepithelial neoplasia in needle biopsy of the prostate. J Urol. 1995;154:347-351. 12. Langer JE, Rovner ES, Coleman BG, et al. Strategy for repeat biopsy of patients with prostatic intraepithelial neoplasia detected by prostate needle biopsy. J Urol. 1996;155:228-231. 13. Park S, Shinohara K, Grossfeld GD, Carroll PR. Prostate cancer detection in men with prior high grade prostatic intraepithelial neoplasia or atypical prostate biopsy. J Urol. 2001;165:1409-1414. 14. Gore JL, Shariat SF, Miles BJ, et al. Optimal combinations of systematic sextant and laterally directed biopsies for the detection of prostate cancer. J Urol. 2001;165:1554-1559. 15. Lefkowitz GK, Sidhu GS, Torre P, et al. Is repeat prostate biopsy for high-grade prostatic intraepithelial neoplasia necessary after routine 12-core sampling? Urology. 2001;58:999-1003. 16. Mian BM, Naya Y, Okihara K, et al. Predictors of cancer in repeat extended multisite prostate biopsy in men with previous negative extended multisite biopsy. Urology. 2002;60:836-840. 17. Moore CK, Karikehalli S, Nazeer T, et al. Prognostic signicance of high grade prostatic intraepithelial neoplasia and atypical small acinar proliferation in the contemporary era. J Urol. 2005;173:7072. 18. Lefkowitz GK, Taneja SS, Brown J, et al. Followup interval prostate biopsy 3 years after diagnosis of high grade prostatic intraepithelial neoplasia is associated with high likelihood of prostate cancer, independent of change in prostate specic antigen levels. J Urol. 2002;168:1415-1418. 19. Taneja SS. Optimizing prostate biopsy strategies for the diagnosis of prostate cancer. Rev Urol. 2003;5:149-155. 20. Carter HB, Pearson JD, Metter EJ, et al. Longitudinal evaluation of prostate-specic antigen levels in men with and without prostate disease. JAMA. 1992;267:2215-2220.

CONCLUSIONS
During follow-up of isolated HGPIN using empiric DIBx every 2-3 years, prostate cancer was diagnosed in 32.1% of men at a median follow-up of 46.6 months. The stability of the PSA level or the absence of HGPIN on repeat biopsy did not lower the risk of cancer detection on DIBx in our experience. Collectively, the sustained risk of cancer on follow-up biopsy, the high likelihood of
UROLOGY xx (x), xxxx

21. Izawa JI, Lega I, Downey D, et al. Do all patients with high-grade prostatic intraepithelial neoplasia on initial prostatic biopsy eventually progress to clinical prostate cancer? BJU Int. 2005;96:320-323. 22. Abdel-Khalek M, El-Baz M, El-Houssieny I. Predictors of prostate cancer on extended biopsy in patients with high-grade prostatic intraepithelial neoplasia: a multivariate analysis model. BJU Int. 2004;94:528-533. 23. Alexander EE, Qian J, Wollan PC, et al. Prostatic intraepithelial neoplasia does not appear to raise serum prostate-specic antigen concentration. Urology. 1996;47:693-698.

24. Thompson IM, Goodman PJ, Tangen CM, et al. The inuence of nasteride on the development of prostate cancer. N Engl J Med. 2003;349:215-224. 25. Roobol MJ, van der Cruijsen IW, Schroder FH. No reason for immediate repeat sextant biopsy after negative initial sextant biopsy in men with PSA level of 4.0 ng/mL or greater (ERSPC, Rotterdam). Urology. 2004;63:892-899. 26. Andriole GL, Bostwick DG, Brawley OW, et al. REDUCE Study Group. Effect of dutasteride on the risk of prostate cancer. N Engl J Med. 2010;362:1192-1202.

UROLOGY xx (x), xxxx