ACCURACY OF PROSTATE WEIGHT ESTIMATION BY DIGITAL RECTAL EXAMINATION VERSUS TRANSRECTAL ULTRASONOGRAPHY

0022-5347/05/1731-0063/0 THE JOURNAL OF UROLOGY® Copyright © 2005 by AMERICAN UROLOGICAL ASSOCIATION

Vol. 173, 63– 65, January 2005 Printed in U.S.A.

DOI: 10.1097/01.ju.0000145883.01068.5f

Oncology: Prostate/Testis/Penis/Urethra ACCURACY OF PROSTATE WEIGHT ESTIMATION BY DIGITAL RECTAL EXAMINATION VERSUS TRANSRECTAL ULTRASONOGRAPHY STACY LOEB, MISOP HAN,* KIMBERLY A. ROEHL, JO ANN V. ANTENOR AND WILLIAM J. CATALONA From the Department of Urology, Northwestern University Feinberg School of Medicine (SL, MH, WJC), Chicago, Illinois, and Departments of Psychiatry (KAR), Neurology (JAVA) and Urologic Surgery (WJC), Washington University School of Medicine, St. Louis, Missouri

ABSTRACT

Purpose: The ability to estimate prostate weight is useful. Two commonly used methods for estimating prostate weight are digital rectal examination (DRE) and transrectal ultrasonography (TRUS). We evaluated the relative accuracy of these weight estimates by comparing them to prostate weight following radical retropubic prostatectomy (RRP). Materials and Methods: Between 1989 and 2001 more than 36,000 community men participated in a large prostate cancer screening study. Of these men 2,238 underwent RRP. In this subset we examined the correlation between documented preoperative DRE and TRUS estimates of prostate weight with actual gland weight. Results: DRE estimates of prostate weight by multiple examiners correlated poorly with RRP specimen weight (r ⫽ 0.2743). However, TRUS estimates correlated moderately well (r ⫽ 0.6493). TRUS provided more accurate estimates of prostate weight for smaller glands, although it generally underestimated gland weight compared to the weight of the surgical specimen. Conclusions: In a large, community based prostate cancer screening study prostate weight estimated by DRE was shown to correlate poorly with actual prostate weight. Compared with DRE, TRUS provides a better estimate of prostate weight. In addition, TRUS measurements were more accurate in smaller prostate glands. KEY WORDS: prostate; mass screening; ultrasound, high-intensity focused, transrectal; prostatic neoplasms; physical examination

means to differentiate PSA increases due to BPH from those due to cancer.3, 4 One would not expect the same PSA level in a smaller gland as in a larger gland unless the smaller gland contained prostate cancer and the larger one did not.2, 5 For guiding diagnosis and treatment PSAD must be an accurate measurement. It has previously been shown that TRUS estimate of prostate weight has high interobserver reliability.6 There have also been studies comparing DRE estimates of prostate weight to TRUS estimates7 but to our knowledge none comparing these 2 estimates to the measured weight of surgical specimens of the gland exclusively in men with prostate cancer. Thus, in a large study of men with prostate cancer we evaluated the accuracy of prostate weight estimates using TRUS and DRE by comparing them to gold standard radical retropubic prostatectomy (RRP) specimen weights.

Prostate cancer is the leading noncutaneous cancer and second leading cause of cancer death in American men.1 Prostate cancer screening is routinely performed using digital rectal examination (DRE) and serum prostate specific antigen (PSA) measurements. While PSA is the only tumor marker used for mass screening at this time, there are many limitations in this measurement. PSA increases are seen not only in cancer of the prostate, but also in benign prostatic hyperplasia (BPH) and prostatitis. Various manipulations of the prostate, such as cystoscopy, prostatic massage and biopsy, also increase serum PSA. Nevertheless, it has been demonstrated that serum PSA correlates with tumor volume in men with prostate cancer.2 Different methods of assessing prostate cancer risk using serum PSA have been evaluated to increase the specificity of PSA testing. They include stratifications of PSA for age and race, PSA velocity or doubling time, and the percent of free and complexed PSA. Another potentially useful measurement is PSA density (PSAD), which is defined as total serum PSA divided by prostate gland weight, as estimated by transrectal ultrasonography (TRUS). PSAD was proposed as a possible

MATERIALS AND METHODS

More than 36,000 men were evaluated as part of community based prostate cancer screening studies performed at Washington University, as described previously.8 –11 Subjects underwent PSA measurement and DRE at 6 or 12-month intervals depending on PSA to determine the need for biopsy. Men with prostatitis or a previous diagnosis of prostate cancer were excluded. The study protocols were approved by the Human Studies Committee of Washington University. Informed consent was obtained from all subjects. From the prostate cancer screening program 2,238 men

Submitted for publication April 13, 2004. Study received human studies committee, Washington University approval. Supported by an award from Beckman Coulter, Inc., Fullerton, California and a National Cancer Institute Prostate Cancer SPORE Career Development Award. * Correspondence: Northwestern Medical Faculty Foundation, 675 North Saint Clair St., Suite 20-150, Chicago, Illinois 60611 (telephone: 312-695-8146; FAX: 312-695-7030; e-mail: [email protected]). 63

64

ACCURACY OF PROSTATE WEIGHT ESTIMATES

underwent RRP (mean age 64.6 years, range 40 to 80). In 2,190 of the 2,238 who underwent RRP the weight of the surgical specimen, including the prostate gland and seminal vesicles, was documented. In these 2,190 men mean serum PSA ⫾ SD was 6.14 ⫾ 6.19 ng/ml (range 0.2 to 82.0). Multiple examiners performed DRE and TRUS measurements. Blood samples for PSA were obtained before or at least 1 week after DRE. Calculations of TRUS prostate weight were based on a computer algorithm using the elliptical formula, (length ⫻ width ⫻ height) ⫻ ⌸/6).12 The TRUS estimate of prostate weight and RRP weight were available for 1,844 of the 2,190 men. The DRE estimate of prostate weight and RRP weight were available for 1,566 of the 2,190 men. The prostate has a specific gravity of approximately 1.0.12 Therefore, we compared volume estimates with specimen weight by equating 1 cc to 1 gm. Statistical analyses were performed using STATA SE 8.2 (Stata Corp., College Station, Texas). RESULTS

In our study population of 2,190 men mean age was 64.6 years and mean preoperative PSA was 6.14 ⫾ 6.19 ng/ml. Mean prostate gland weight of the RRP specimen was 49.9 ⫾ 23.0 gm. The mean total gland weight as estimated by DRE was 30.9 ⫾ 9.7 gm, in contrast to the mean TRUS estimated volume of 39.7 ⫾ 24.7 gm. Figure 1 shows the prostate weight estimate by DRE compared to RRP specimen weight in 1,566 subjects. Figure 2 shows the prostate weight estimate by TRUS compared to RRP specimen weight in 1,844 subjects. Overall DRE estimated prostate weight correlated poorly with actual prostate weight in the RRP specimen (Spearman’s r ⫽ 0.2743). Estimation by TRUS correlated more closely with prostate weight in the RRP specimen (Spearman’s r ⫽ 0.6493). Subjects were arbitrarily stratified into 3 groups based on RRP prostate weight (less than 30, 30 to 50 and greater than 50 gm) in an attempt to assess the accuracy of TRUS estimation in each of these subgroups (fig. 3). TRUS estimations correlated better with actual size for smaller prostates. This was consistent with the previous observation that the elliptical formula is a better approximation of prostate shape for smaller prostates.12

FIG. 2. Relationship between prostate weight estimates by TRUS and RRP specimen weight.

DISCUSSION

Serum PSA is widely used for prostate cancer screening based on the fact that serum PSA is significantly associated

FIG. 1. Relationship between prostate weight estimates by DRE and RRP specimen weight.

FIG. 3. Difference in prostate weight estimation by TRUS and actual RRP weight for prostates less than 30, 30 to 50 and greater than 50 gm.

with tumor volume.2 This relationship is not linear because higher grade tumors may actually produce less PSA per unit of weight.5 There are other potential confounding variables when using total PSA measurements as a marker for prostate cancer. The prostatic epithelium produces PSA but the ratio of epithelial to stromal cells varies up to 3-fold among patients.5 Also, there is a variable amount of BPH tissue that contributes to serum PSA. This is the basis for using PSAD measurement as a means of differentiating PSA increases due to BPH from those due to prostate cancer. In 1992 Benson et al pioneered the use of PSAD to select patients for biopsy.4 In that study RRP specimens were collected from 41 men. Actual PSAD was calculated based on prostatectomy specimen weight. The mean PSAD thus calculated was 0.581. As a comparison group, they evaluated 20 patients with BPH but no known prostate cancer. Using magnetic resonance imaging estimations of prostate weight they found a mean PSAD of 0.044 in the BPH group. A similar comparison of PSAD in patients with and without diagnosed prostate cancer was made in 1990 by Veneziano et al.13 They evaluated the PSA/volume index and found a mean of 1.73 in men diagnosed with prostate cancer, 0.107 in men with prostatitis, and 0.0946 in patients with BPH. In a subsequent study of Benson et al total PSA and PSAD were

65

ACCURACY OF PROSTATE WEIGHT ESTIMATES

compared in 533 men who underwent prostate biopsy.3 There was a statistically significant difference in PSAD between men found to have prostate cancer and those who did not (p ⫽ 0.297 vs 0.188). Further support for the importance of gland volume in assessing the PSA measurement came from a study of Babaian et al in 1990 involving subjects with PSA in the range of 4 to 20 ng/ml.14 They demonstrated that PSA greater than 4.0 ng/ml in a gland less than 25 cm3 on TRUS was suspicious for cancer, whereas a volume of greater than 25 cm3 in this PSA range indicated a high probability of BPH. Since there is an evidence suggesting that prostate weight is an important variable for assessing the risk of prostate cancer,14 we examined the accuracy of current methods used clinically to estimate the volume. We compared DRE and TRUS estimates of prostate weight to the actual weight measured from radical prostatectomy specimens. In our study the prostate weight from TRUS measurements was calculated using the elliptical formula. Terris and Stamey used this method to compare 15 methods of weight estimation using TRUS with actual prostatic weight measured in 150 patients who underwent radical prostatectomy or cystoprostatectomy.12 Similar to the groupings in the current study, they separated their specimens into subgroups of less than 40, 40 to 80 and greater than 80 gm. Of all of the different formulas tested they found that the prolate spheroid and elliptical formula was the most accurate at estimating prostate weight in the less than 40 and 40 to 80 gm groups, whereas the spherical formula [⌸ ⫻ (transverse diameter)3/6] was found to be the best approximation for prostates greater than 80 gm. They attributed this to the different configurations assumed by the prostate gland at different weights, such that 1 formula may be a better approximation than another depending on prostate weight. Their study differed from ours in that not all of their subjects had prostate cancer. However, their finding of better performance of the elliptical formula in smaller prostates could account in part for our finding of a better correlation between TRUS estimates and RRP weight in the smaller prostate groups. It is possible that the correlations would have been better in the larger prostate group had we instead used the spherical formula in our calculations. However, we elected to maintain consistency by using the same formula for all of our calculations. The ability to estimate prostate weight is important beyond calculating PSAD. An accurate preoperative estimation of prostate weight is helpful for surgical planning prior to brachytherapy, radical prostatectomy or transurethral resection of the prostate.15 Prostate weight is also important for BPH management since it influences the risk of complications such as urinary retention, the response to 5-␣ reductase inhibitors and the need for appropriate surgical treatment.16 Our finding of a poor correlation between DRE estimates and prostate weight suggests that DRE should be used primarily for the detection of abnormalities of prostate consistency and symmetry but not as a means of accurate volume estimation. For PSAD estimation TRUS estimates of prostate weight are more accurate. However, they are still not optimal compared to actual gland weight, thus, decreasing somewhat the usefulness of PSAD calculations based on them. Several limitations of our study deserve mention. While some other studies showed a higher correlation between TRUS and prostate volume, multiple examiners, including urology residents, performed prostate weight estimations using DRE and TRUS in our study.12 It is possible that with experience fully trained urologists may be able to estimate better prostate weight preoperatively using these modalities. Furthermore, TRUS estimations may be more accurate if performed by an experienced examiner in a standardized manner. Also, there is no perfect algorithm for prostate

weight estimation. However, the literature supports the generalized use of the prolate spheroid formula.12 Finally, RRP specimens in our study were weighed with the seminal vesicles attached. Error or bias may have been introduced since seminal vesicle size varies among individuals. CONCLUSIONS

In our prostate screening study population DRE estimations by multiple examiners correlated poorly with the actual prostate weight measured in the surgical specimen. However, TRUS estimates by multiple examiners correlated better with RRP specimen weight. Compared to DRE, TRUS provides more accurate measurement of prostate size. In addition, TRUS measurements correlated better with actual gland weight for smaller prostates. REFERENCES

1. Palken, M., Cobb, O. E., Warren, B. H. and Hoak, D. C.: Prostate cancer: correlation of digital rectal examination, transrectal ultrasound and prostate specific antigen levels with tumor volumes in radical prostatectomy specimens. J Urol, 143: 1155, 1990 2. Stamey, T. A., Yang, N., Hay, A. R., McNeal, J. E., Freiha, F. S. and Redwine, E.: Prostate-specific antigen as a serum marker for adenocarcinoma of the prostate. N Engl J Med, 317: 909, 1987 3. Benson, M. C., Whang, I. S., Olsson, C. A., McMahon, D. J. and Cooner, W. H.: The use of prostate specific antigen density to enhance the predictive value of intermediate levels of serum prostate specific antigen. J Urol, 147: 817, 1992 4. Benson, M. C., Whang, I. S., Pantuck, A., Ring, K., Kaplan, S. A., Olsson, C. A. et al: Prostate specific antigen density: a means of distinguishing benign prostatic hypertrophy and prostate cancer. J Urol, 147: 815, 1992 5. Partin, A. W., Carter, H. B., Chan, D. W., Epstein, J. I., Oesterling, J. E., Rock, R. C. et al: Prostate specific antigen in the staging of localized prostate cancer: influence of tumor differentiation, tumor volume and benign hyperplasia. J Urol, 143: 747, 1990 6. Sech, S., Montoya, J., Girman, C. J., Rhodes, T. and Roehrborn, C. G.: Interexaminer reliability of transrectal ultrasound for estimating prostate volume. J Urol, 166: 125, 2001 7. Roehrborn, C. G., Girman, C. J., Rhodes, T., Hanson, K. A., Collins, G. N., Sech, S. M. et al: Correlation between prostate size estimated by digital rectal examination and measured by transrectal ultrasound. Urology, 49: 548, 1997 8. Smith, D. S. and Catalona, W. J.: The nature of prostate cancer detected through prostate specific antigen based screening. J Urol, 152: 1732, 1994 9. Catalona, W. J., Smith, D. S. and Ornstein, D. K.: Prostate cancer detection in men with serum PSA concentrations of 2.6 to 4.0 ng/mL and benign prostate examination. Enhancement of specificity with free PSA measurements. JAMA, 277: 1452, 1997 10. Roehl, K. A., Antenor, J. A. V. and Catalona, W. J.: Robustness of free prostate specific antigen measurements to reduce unnecessary biopsies in the 2.6 to 4.0 ng/ml range. J Urol, 168: 922, 2002 11. Smith, D. S., Carvalhal, G. F., Mager, D. E., Bullock, A. D. and Catalona, W. J.: Use of lower prostate specific antigen cutoffs for prostate cancer screening in black and white men. J Urol, 160: 1734, 1998 12. Terris, M. K. and Stamey, T. A.: Determination of prostate volume by transrectal ultrasound. J Urol, 145: 984, 1991 13. Veneziano, S., Pavlica, P., Querze, R., Nanni, G., Lalanne, M. G. and Vecchi, F.: Correlation between prostate-specific antigen and prostate volume, evaluated by transrectal ultrasonography: usefulness in diagnosis of prostate cancer. Eur Urol, 18: 112, 1990 14. Babaian, R. J., Fritsche, H. A. and Evans, R. B.: Prostate-specific antigen and prostate gland volume: correlation and clinical application. J Clin Lab Anal, 4: 135, 1990 15. Roehrborn, C. G., Chinn, H. K. W., Fulgham, P. F., Simpkins, K. L. and Peters, P. C.: The role of transabdominal ultrasound in the preoperative evaluation of patients with benign prostatic hypertrophy. J Urol, 135: 1190, 1986 16. Roehrborn, C. G., McConnell, J., Bonilla, J., Rosenblatt, S., Hudson, P. B., Malek, G. H. et al: Serum prostate specific antigen is a strong predictor of future prostate growth in men with benign prostatic hyperplasia. J Urol, 163: 13, 2000