Benign Prostate Specific Antigen Distribution and Associations With Urological Outcomes in Community Dwelling Black and White Men

Oncology: Prostate/Testis/Penis/Urethra

Benign Prostate Specific Antigen Distribution and Associations With Urological Outcomes in Community Dwelling Black and White Men Thomas Rhodes,* Debra J. Jacobson, Michaela E. McGree, Jennifer L. St. Sauver,† Aruna V. Sarma, Cynthia J. Girman,* Michael M. Lieber, George G. Klee,‡ Kitaw Demissie and Steven J. Jacobsen§ From the Department of Epidemiology, University of Medicine and Dentistry of New Jersey (TR, KD), Piscataway, New Jersey, Departments of Health Sciences Research (DJJ, MEM, JLS), Urology (MML) and Laboratory Medicine and Pathology (GGK), Mayo Clinic College of Medicine, Rochester, Minnesota, Departments of Epidemiology and Urology, University of Michigan (AVS), Ann Arbor, Michigan, Epidemiology Department, Merck Research Laboratories (TR, CJG), North Wales, Pennsylvania, and Department of Research and Evaluation, Kaiser Permanente-Southern California (SJJ), Pasadena, California

Purpose: We describe cross-sectional associations of benign prostate specific antigen with clinical urological measures and examined the risk of future urological outcomes in 2 population based cohorts of black and white men, respectively. Materials and Methods: Two population based cohort studies were established to characterize the natural history of and risk factors for prostate disease progression in white and black male residents of Olmsted County, Minnesota, and Genesee County, Michigan, respectively. Results: The benign prostate specific antigen distribution was similar in black men at a median of 32.9 pg/ml (25th, 75th percentiles 17.3, 68.0) and white men at a median of 32.2 pg/ml (25th, 75th percentiles 16.6, 68.9, respectively). However, it was much lower than in previous reports. For Olmsted County men in the upper quartile of benign prostate specific antigen there was a fifteenfold increased risk of prostate cancer (HR 14.6, 95% CI 3.1– 68.6) and a twofold higher risk of treatment for benign prostatic hyperplasia (HR 2.2, 95% CI 1.2– 4.2) after adjusting for age. After additional adjustment for baseline prostate specific antigen the association between benign prostate specific antigen and prostate cancer risk was attenuated but remained almost ninefold higher for men in the upper quartile of benign prostate specific antigen (HR 8.7, 95% CI 1.8 – 42.4). The twofold higher risk of treatment for benign prostatic hyperplasia also remained after adjusting for baseline prostate specific antigen for men in the upper benign prostate specific antigen quartile (HR 1.9, 95% CI 0.9 – 4.0). Conclusions: Results suggest that increased benign prostate specific antigen may help identify men with prostate cancer and those at risk for benign prostatic hyperplasia treatment. Key Words: prostate, prostatic hyperplasia, prostate-specific antigen, prostatic neoplasms, continental population groups PROSTATE specific antigen is a widely used serum marker for CaP early detection. PSA measurement is recommended in men older than 50 years

and those at high risk for CaP by National Cancer Institute1 and AUA2 guidelines. However, elevated PSA is not specific to CaP since PSA can also

0022-5347/12/1871-0087/0 THE JOURNAL OF UROLOGY® © 2012 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION

Vol. 187, 87-91, January 2012 Printed in U.S.A. DOI:10.1016/j.juro.2011.09.061

AND

RESEARCH, INC.

Abbreviations and Acronyms AUA ⫽ American Urological Association AUASI ⫽ AUA symptom index AUR ⫽ acute urinary retention BPH ⫽ benign prostatic hyperplasia BPSA ⫽ benign PSA CaP ⫽ prostate cancer FMHS ⫽ Flint Men’s Health Study LUTS ⫽ lower urinary tract symptoms OCS ⫽ Olmsted County Study of Urinary Symptoms and Health Status among Men PSA ⫽ prostate specific antigen TRUS ⫽ transrectal ultrasound Submitted for publication April 8, 2011. Study received approval from the Mayo Clinic, Olmsted Medical Center and University of Michigan Medical School institutional review boards. Supported by Grants DK58859, AR30582 and 1UL1 RR024150-01 from the Public Health Service, National Institutes of Health, National Cancer Institute Grant P50CA69568 and Merck Research Laboratories. Supplementary material for this article can be obtained at http://www.jurology.com. * Financial interest and/or other relationship with Merck. † Correspondence: Division of Epidemiology, Mayo Clinic, 200 First St. Southwest, Rochester, Minnesota 55905 (telephone: 507-538-6916; FAX: 507284-1516; e-mail: [email protected]). ‡ Financial interest and/or other relationship with Beckman-Coulter. § Financial interest and/or other relationship with Beckman-Coulter and Merck.

For another article on a related topic see page 322.

www.jurology.com

87

88

BENIGN PROSTATE SPECIFIC ANTIGEN AND UROLOGICAL OUTCOMES IN BLACK AND WHITE MEN

be increased in men with benign conditions such as BPH or prostatitis. BPSA is an inactive form of PSA that has been cleaved at lysine residues 145 and 182.3,4 This form of free PSA is increased in nodular BPH tissue and also correlates with transition zone volume and prostatic enlargement.3,5 BPSA was significantly higher in men with increased PSA and percent free PSA less than 15% who were also biopsy negative for CaP.6 Together these results suggest that BPSA may be associated with benign disease. BPSA characteristics were described in clinical and convenience samples.3,5–7 These studies provide initial support for the usefulness of BPSA as a marker for benign prostate disease but they may not reflect the full spectrum of men with and without benign prostate disease. We provide normative data on a population based, biracial sample of men and describe the cross-sectional associations between baseline BPSA and common clinical urological measures. We also examined associations between baseline BPSA and the risk of clinically relevant urological outcomes such as AUR, CaP and BPH treatment.

METHODS Details on subject selection for OCS and FMHS were published previously.8,9 Briefly, OCS and FMHS are population based, cohort studies established to characterize the natural history of and risk factors for prostate disease progression in white and black male residents of Olmsted County, Minnesota, and Genesee County, Michigan, respectively. In OCS 2,115 of 3,874 eligible white men (55%) who were 40 to 79 years old in 1990 and had no history of CaP, surgery or other conditions known to interfere with voiding completed the self-administered AUASI.10 A detailed urological examination, including uroflowmetry, digital rectal examination, TRUS and serum PSA measurement, was done in a 25% random subsample comprising 476 of the 537 participants sampled (89%).8,11 The cohort was actively followed biennially for 17 years using a protocol similar to that of the initial examination. By applying criteria and procedures similar to those of OCS 730 of 943 eligible black men (77%) were recruited in 1996 to FMHS to complete an interview administered questionnaire.9,12 Of the men 369 (51%) who were free of CaP completed a comprehensive urological examination, including uroflowmetry, TRUS, serum PSA measurement and the self-administered AUASI,9 as in OCS. Selective participation in clinical examination and the potential resulting selection bias were addressed previously.13,14 Since FMHS urological measurements were collected in 1996, BPSA measures were obtained from serum samples collected in 1996 for each study and the corresponding 1996 OCS measurements served as initial/baseline values. BPSA was measured in 420 men in OCS and in 329 in FMHS. This study received approval from the Mayo Clinic, Olmsted Medical Center and University of Michigan Medical School institutional review boards.

Urinary Flow Rate The urinary flow rate was measured using a Model 1000 uroflowmeter (Dantec, Royal Portbury, United Kingdom) calibrated by trained study assistants.15 Maximum and average urinary flow rates, and voided volume were measured electronically. Measurements were repeated if voided volume was less than 150 ml. The highest rate with a voided volume of 150 ml or greater was used for analysis when multiple measurements existed.

Prostate Volume Total prostate volume for OCS and FMHS participants was measured by TRUS.9,16 In addition to assessing the echogenic pattern of the prostate gland, 3 measurements were made to calculate total prostate volume, assuming a prolate ellipsoid shape.17

CaP and AUR Information on medical and surgical LUTS/BPH treatments was obtained by self-report and passive followup of the community medical records of men in OCS. Dates of biopsy confirmed CaP and AUR were abstracted from community medical records. AUR was defined as medical record documentation of the inability to urinate and of catheterization to empty the bladder. Catheterizations done perioperatively were excluded from study but episodes were included in which a catheter was needed after surgery secondary to AUR. Data on CaP, AUR and treatment were not available on men in FMHS. Thus, analysis of associations between BPSA and these outcomes was confined to the OCS cohort.

Assays For each study serum samples were obtained before prostatic manipulations, including digital rectal examination and TRUS. Samples were frozen at ⫺70C for latter assays. Stored blood samples from the 1996 OCS fourth biennial followup in 420 men and from the 1996 FMHS baseline visit in 329 were used to measure BPSA. Serum PSA for men in OCS was assayed using the Tandem-E PSA assay (Hybritech®). Serum PSA for men in FMHS was assayed using the AxSYM® polyclonal-monoclonal immunoassay with a lower limit of detection of 0.1 ng/ml.18 We previously reported that PSA determinations were consistent across different assays and laboratories.19,20 For each study BPSA was assessed using an automated, sequential, 2-step immunoenzymatic sandwich assay developed for the Access® instrument. BPSA measurements were run on an Access 2 Immunoassay analyzer using research use only, 2-site immunoenzymatic reagents. In our hands the intra-assay variation range was 4.3% to 8.1% and the inter-assay variation range was 5.1% to 5.2%.

Statistical Analysis Descriptive statistics and cumulative distribution function plots were used to describe the BPSA distribution. The Spearman correlation (rS) and logistic regression were used to investigate relationships between BPSA and other baseline urological measures. Multivariate logistic regression was used to describe the association of BPSA with clinically meaningful diagnostic measures, which are shown as the OR and 95% CI. Moderate to severe LUTS was defined as an AUASI score of greater than 7 while a

BENIGN PROSTATE SPECIFIC ANTIGEN AND UROLOGICAL OUTCOMES IN BLACK AND WHITE MEN

Table 1. Baseline characteristics of OCS and FMHS participants

Age: 40–49 50–59 60–69 70⫹ AUASI score greater than 7 Max urinary flow rate less than 12 ml/sec Prostate vol greater than 30 cc PSA greater than 1.4 ng/ml

89

Table 2. Spearman correlations of baseline BPSA and clinical urological measures (p ⬍0.05)

No. OCS (%)*

No. FMHS (%)*

p Value†

112 (26.7) 143 (34.0) 92 (21.9) 73 (17.4) 183 (43.6) 67 (18.9)

91 (27.7) 104 (31.6) 83 (25.2) 51 (15.5) 128 (39.6) 28 (14.4)

0.64

153 (40.7) 146 (34.8)

111 (35.8) 109 (33.1)

0.19 0.64

0.28 0.18

* Percents based on nonmissing observations. † Rank sum test for continuous and chi-square test for dichotomous variables.

decreased urinary flow rate was defined as a maximum urinary flow rate of less than 12 ml per second.21 Prostatic enlargement was defined as prostate volume greater than 30 cc.21 Due to availability of the followup data Cox proportional hazards regression was used in the OCS cohort only to investigate the association of baseline BPSA with the risk of a subsequent diagnosis of AUR, biopsy confirmed CaP and BPH treatment, which are shown as the HR and 95% CI.

RESULTS Table 1 lists the baseline characteristics of participants in each cohort. Urological characteristics were similar for FMHS and OCS men. The 2 cohorts were also similar in BPSA distribution. BPSA percentile levels increased across age decades almost uniformly for each percentile shown (see figure). The 5th, 10th, 25th, 50th, 75th, 90th and 95th percentiles of the combined BPSA distributions were 7.3, 10.1, 16.9, 32.5, 68.8, 157.0 and 267.9 pg/ml, respectively. Median BPSA was 32.2 pg/ml (25th, 75th

Cumulative distribution function plots of BPSA by age decade in combined OCS and FMHS sample.

BPSA rS

Age Max flow rate Prostate vol AUASI score PSA

OCS

FMHS

0.58 ⫺0.18 0.52 0.28 0.82

0.40 ⫺0.21 0.50 0.11 0.80

percentiles 16.6, 68.9) in OCS and 32.9 pg/ml (25th, 75th percentiles 17.3, 68.0, respectively) in FMHS. BPSA significantly correlated with age and all urological measures examined (Spearman rS ⫽ 0.11 to 0.82, each p ⬍0.05, table 2). In OCS and FMHS BPSA most strongly correlated with PSA (rS ⫽ 0.82 and 0.80, p ⬍0.0001) as well as with age (rS ⫽ ⫺0.58 and 0.40) and prostate volume (rS ⫽ 0.52 and 0.50, respectively). In multivariate models for each cohort men in the BPSA upper quartile were more likely to have moderate to severe LUTS, a decreased urinary flow rate and an enlarged prostate than men in the lower quartile. The magnitude of the association between baseline BPSA and urological measures increased with increasing BPSA quartile but it was most pronounced for an enlarged prostate (test for trend each p ⬍0.05). Overall 46 men (11.3%) in OCS had AUR, 111 (27.8%) were treated for BPH and 41 (9.8%) were diagnosed with CaP after baseline BPSA measurement. Men with higher baseline BPSA (upper quartile) were more likely to have a subsequent AUR event (age adjusted HR 2.1, 95% CI 0.7– 6.5), biopsy confirmed CaP (age adjusted HR 14.6, 95% CI 3.1– 68.6) and BPH treatment (age adjusted HR 2.2, 95% CI 1.2– 4.2) than men in the lowest BPSA quartile (table 3). The BPSA range was minimum to 16.9 pg/ml for quartile 1, 17.0 to 32.5 for quartile 2, 32.6 to 68.8 for quartile 3 and 68.9 to maximum for quartile 4 (table 3). For the 3 outcomes AUR, CaP and BPH treatment the magnitude of the association increased with increasing BPSA quartile (age adjusted test for trend p ⫽ 0.17, ⬍0.0001 and 0.003, respectively). Most tumors were low grade, including 78% with a Gleason score of 6 or less. After further adjusting for baseline PSA the associations between BPSA and all outcomes were attenuated (table 3). However, the association between BPSA and CaP remained almost ninefold higher for men in the upper quartile than in men in the lowest quartile (HR 8.7, 95% CI 1.8 – 42.4). Further adjustment for prostate volume did not change the findings (results not shown). Also, the association between the highest BPSA quartile and a twofold increased risk of BPH treatment remained after ad-

90

BENIGN PROSTATE SPECIFIC ANTIGEN AND UROLOGICAL OUTCOMES IN BLACK AND WHITE MEN

Table 3. Crude and adjusted associations of baseline BPSA with urological outcomes in OCS HR (95% CI) BPSA Quartile* AUR: 2 3 4 p Trend Biopsy confirmed CaP: 2 3 4 p Trend BPH treatment: 2 3 4 p Trend

Crude

Age Adjusted

Age ⫹ PSA Adjusted

1.7 (0.5–5.1) 2.6 (0.9–7.3) 4.6 (1.7–12.2) 0.0004

1.5 (0.5–4.5) 1.8 (0.6–5.3) 2.1 (0.7–6.5) 0.17

1.4 (0.5–4.4) 1.7 (0.6–4.9) 1.5 (0.4–5.4) 0.48

3.7 (0.8–17.9) 5.5 (1.2–25.2) 12.9 (3.0–54.8) ⬍0.0001

3.8 (0.8–18.4) 5.8 (1.3–27.0) 14.6 (3.1–68.6) ⬍0.0001

3.7 (0.8–17.6) 5.2 (1.1–24.1) 8.7 (1.8–42.4) 0.004

0.8 (0.4–1.5) 1.6 (0.9–2.9) 2.9 (1.7–4.9) ⬍0.0001

0.7 (0.4–1.4) 1.5 (0.8–2.6) 2.2 (1.2–4.2) 0.003

0.7 (0.4–1.4) 1.4 (0.8–2.5) 1.9 (0.9–4.0) 0.04

* Quartile 1 served as referent.

justing for baseline PSA (HR 1.9, 95% CI 0.9 – 4.0, table 3). However, it was no longer statistically significant.

DISCUSSION Results suggest that black and white men have similar BPSA distributions. Men with higher BPSA were more likely to have a decreased urinary flow rate and an enlarged prostate. In OCS men with higher baseline BPSA were also more likely to receive subsequent BPH treatment, ie medical therapy, a procedure or surgery, and more likely to be subsequently diagnosed with biopsy confirmed CaP even after adjusting for age and baseline PSA. Median BPSA was 32.2 pg/ml (25th, 75th percentiles 16.6, 68.9) in the OCS cohort and 32.9 pg/ml (25th, 75th percentiles 17.3, 68.0) in the FMHS cohort. These levels were much lower than previously reported in patients who underwent biopsy or were diagnosed with CaP. In a study of 91 consecutive patients without CaP who underwent 10-core biopsy Canto et al reported a median BPSA of 220 pg/ml.5 This observed value is at approximately the 93rd percentile in the combined FMHS and OCS distribution. In another study de Vries et al reported data on 61 men participating in the screening arm of a large cancer screening study in which those with less aggressive CaP had a median BPSA of 172 pg/ml and those with more aggressive CaP had a median BPSA of 264 pg/ml.7 Again, these values were in approximately the 91st and 95th percentiles, respectively, of the BPSA distribution observed for the combined cohorts.

Linton et al reported a median BPSA of 220 and 170 pg/ml in 79 men with BPH, defined as elevated PSA with negative biopsy, and in 91 with CaP, respectively.22 These median BPSA values were above the 90th percentile of the BPSA distribution in our study. The higher levels observed in these other studies could have been due to differences in the underlying populations or to laboratory measurement/calibration. BPSA is associated with clinically significant prostatic enlargement in the transition zone. Canto et al referred to this result as evidence of the potential of BPSA as a novel predictor of outcomes or response to therapy in patients with BPH.5 Results from phase II and III studies of the Medical Therapy of Prostatic Symptoms Prostatic Samples Analysis Consortium suggested that BPSA predicts BPH progression after accounting for prostate volume.23 The current study shows that BPSA correlated with prostate volume and PSA in black and white men. BPSA also modestly correlated with AUASI score and the maximum urinary flow rate in each cohort. Men with higher baseline BPSA were also more likely to receive subsequent BPH treatment in OCS. These data suggest that BPSA may be useful for predicting LUTS severity and prostate enlargement. Based on an artificial neural network model Stephan et al also found that BPSA may improve CaP detection.24 In that study higher BPSA was associated with the risk of biopsy confirmed CaP after adjusting for age and baseline PSA. This result seems to support the potential usefulness of this marker for further distinguishing men with vs without CaP in the presence of a PSA test. BPSA may improve the sensitivity and specificity of detecting CaP in men with PSA in the 4 to 10 ng/ml range or a percent free PSA of less than 15%.6 Due to the restriction in the range of values in the benign prostate disease and the CaP groups the study samples of these previous reports may not necessarily have been the most appropriate groups in which to investigate assay performance for differentiating benign prostate disease from CaP. In the previous studies the benign prostate disease and CaP groups had high BPSA compared to the general population. Thus, the previous reports may represent a conservative estimate of the usefulness of distinguishing diseased from healthy men if the higher levels in those series were due to the underlying study populations and the true differences between diseased and nondiseased men are greater.25 The generalizability of the results to other races or ethnicities may be limited. However, few differences were noted between black and white men on cross-sectional analysis. Furthermore, the wide CIs for some parameter estimates may be a result of

BENIGN PROSTATE SPECIFIC ANTIGEN AND UROLOGICAL OUTCOMES IN BLACK AND WHITE MEN

91

higher variability due to the small number of cases. Also, while to our knowledge the stability of BPSA in frozen samples is currently unknown, free and total PSA have appeared to be stable in serum samples frozen for 2 years at ⫺70C.26

for white and black men. Elevated levels suggest some benign or malignant prostate disease. Further study is needed to develop our understanding of how this might be used for diagnostic evaluation.

CONCLUSIONS

ACKNOWLEDGMENTS

These population based data provide useful reference ranges to researchers for future studies of the usefulness of BPSA with similar BPSA distributions

Kristie Shorter assisted with the manuscript. Beckman Coulter provided BPSA test kits and immunoenzymatic reagents.

REFERENCES 1. Prostate Cancer Screening: Prostate Specific Antigen (PSA) Test 2011. National Cancer Institute. Available at http://www.cancer.gov/cancertopics/ factsheet/Detection/PSA. Accessed January 4, 2011. 2. Clinical Guidelines: Prostate Specific Antigen 2011. American Urological Association. Available at http://www.auanet.org/content/media/psa09. pdf. Accessed January 4, 2011. 3. Mikolajczyk SD, Millar LS, Wang TJ et al: “BPSA,” a specific molecular form of free prostate-specific antigen, is found predominantly in the transition zone of patients with nodular benign prostatic hyperplasia. Urology 2000; 55: 41. 4. Mikolajczyk SD, Marks LS, Partin AW et al: Free prostate-specific antigen in serum is becoming more complex. Urology 2002; 59: 797. 5. Canto EI, Singh H, Shariat SF et al: Serum BPSA outperforms both total PSA and free PSA as a predictor of prostatic enlargement in men without prostate cancer. Urology 2004; 63: 905.

prostate size, flow rate and bothersomeness. J Urol 2001; 165: 1521. 10. Barry MJ, Fowler FJ Jr, O’Leary MP et al: The American Urological Association symptom index for benign prostatic hyperplasia. The Measurement Committee of the American Urological Association. J Urol 1992; 148: 1549. 11. Jacobsen SJ, Guess HA, Panser L et al: A population-based study of health care-seeking behavior for treatment of urinary symptoms. The Olmsted County Study of Urinary Symptoms and Health Status Among Men. Arch Fam Med 1993; 2: 729. 12. Sarma AV, Wei JT, Jacobson DJ et al: Comparison of lower urinary tract symptom severity and associated bother between community-dwelling black and white men: the Olmsted County Study of Urinary Symptoms and Health Status and the Flint Men’s Health Study. Urology 2003; 61: 1086. 13. Panser LA, Chute CG, Guess HA et al: The natural history of prostatism: the effects of non-response bias. Int J Epidemiol 1994; 23: 1198.

18. Cooney KA, Strawderman MS, Wojno KJ et al: Age-specific distribution of serum prostate-specific antigen in a community-based study of African-American men. Urology 2001; 57: 91. 19. Jacobsen SJ, Klee GG, Lilja H et al: Stability of serum prostate-specific antigen determination across laboratory, assay, and storage time. Urology 1995; 45: 447. 20. Oesterling JE, Moyad MA, Wright GL Jr et al: An analytical comparison of the three most commonly used prostate-specific antigen assays: Tandem-R, Tandem-E, and IMx. Urology 1995; 46: 524. 21. Jacobsen SJ, Jacobson DJ, Girman CJ et al: Natural history of prostatism: risk factors for acute urinary retention. J Urol 1997; 158: 481. 22. Linton HJ, Marks LS, Millar LS et al: Benign prostate-specific antigen (BPSA) in serum is increased in benign prostate disease. Clin Chem 2003; 49: 253. 23. Mullins C, Lucia MS, Hayward SW et al: A comprehensive approach toward novel serum biomarkers for benign prostatic hyperplasia: the MPSA Consortium. J Urol 2008; 179: 1243.

6. Khan MA, Sokoll LJ, Chan DW et al: Clinical utility of proPSA and “benign” PSA when percent free PSA is less than 15%. Urology 2004; 64: 1160.

14. Heeringa SG, Alcser KH, Doerr K et al: Potential selection bias in a community-based study of PSA levels in African-American men. J Clin Epidemiol 2001; 54: 142.

7. de Vries SH, Raaijmakers R, Blijenberg BG et al: Additional use of [⫺2] precursor prostate-specific antigen and “benign” PSA at diagnosis in screendetected prostate cancer. Urology 2005; 65: 926.

15. Roberts RO, Jacobsen SJ, Jacobson DJ et al: Longitudinal changes in peak urinary flow rates in a community based cohort. J Urol 2000; 163: 107.

24. Stephan C, Cammann H, Deger S et al: Benign prostatic hyperplasia-associated free prostatespecific antigen improves detection of prostate cancer in an artificial neural network. Urology 2009; 74: 873.

8. Chute CG, Panser LA, Girman CJ et al: The prevalence of prostatism: a population-based survey of urinary symptoms. J Urol 1993; 150: 85.

16. Rhodes T, Girman CJ, Jacobsen SJ et al: Longitudinal prostate growth rates during 5 years in randomly selected community men 40 to 79 years old. J Urol 1999; 161: 1174.

25. Jacobsen SJ, Bergstralh EJ, Guess HA et al: Predictive properties of serum-prostate-specific antigen testing in a community-based setting. Arch Intern Med 1996; 156: 2462.

9. Wei JT, Schottenfeld D, Cooper K et al: The natural history of lower urinary tract symptoms in black American men: relationships with aging,

17. Terris MK and Stamey TA: Determination of prostate volume by transrectal ultrasound. J Urol 1991; 145: 984.

26. Woodrum D and York L: Two-year stability of free and total PSA in frozen serum samples. Urology 1998; 52: 247.