- Email: [email protected]
THE VALUE OF THE RATIO OF FREE-TO-TOTAL PROSTATE SPECIFIC ANTIGEN FOR STAGING PURPOSES IN PREVIOUSLY UNTREATED PROSTATE CANCER
0022-5347/98/1596-2~03.00/0
Vol. 159,2004-2008, June 1998 Printed in U.S.A.
THE JOURNAL OF UROLOGY
Copyright 8 1998 by AMERICAN UROLOCICAL ASSOCIATION, INC
THE VALUE OF THE RATIO OF FREE-TO-TOTAL PROSTATE SPECIFIC ANTIGEN FOR STAGING PURPOSES IN PREVIOUSLY UNTREATED PROSTATE CANCER JOACHIM NOLDUS,* MARKUS GRAEFEN, EDITH HULAND, CHRISTOPHER BUSCH, PETER HAMMERER AND HARTWIG HULAND From the Department of Urology, University Hospital Eppendorf, Hamburg, Germany
ABSTRACT
Purpose: We analyzed the use of the ratio of free-to-total prostate specific antigen (PSA), also termed percentage of free PSA, for predicting tumor stage, volume and grade in patients with clinically localized prostate cancer. Materials and Methods: A total of 515 consecutive patients underwent further prostate evaluation due to elevated PSA (greater than 4.0 ng./ml.) or abnormal digital rectal examination. Prostate cancer was diagnosed in 307 patients (59.6%), including 170 (55.4%)who underwent radical retropubic prostatectomy. Data on pathological stage, Gleason grade, and total and Gleason grade 4 cancer volume were available in all patients. In the remaining 208 men (40.4%) benign prostate hyperplasia was diagnosed. Total and free PSA was measured in preoperative serum. Results: Total PSA was significantly higher (p <0.0001) in the 71 men with stage pT3 tumors than in the 91 with pT2 disease. Eight patients had stage pT4 tumors. Cancer volume correlated well with advancing pathological stage (p <0.0001)and total PSA (p <0.0001).The free-to-total PSA ratio was not significantly different (p = 0.93) in stages pT2 and pT3 tumors, and it did not correlate with total (p = 0.71) or pure Gleason grade 4 (p = 0.94) cancer volume. However, the ratio of free-to-total PSA tended to decrease (p = 0.07) in tumors of increasing Gleason grade. Conclusions: The ratio of free-to-total PSA does not help in the preoperative prediction of final tumor stage and volume. However, disease grading may alter the free-to-total PSA ratio. KEY WORDS: prostate, prostatic neoplasms, prostate-specific antigen Prostate specific antigen (PSA), the most sensitive serum marker in men with prostatic diseases, is produced and released almost exclusively by the prostatic epithelial cells. Because even benign conditions may increase PSA, it is not the ideal tumor marker. On the contrary, approximately 20% of all men with prostate cancer present at diagnosis with serum PSA concentrations in the normal range. Others have demonstrated that serum PSA correlates well with advancing clinical or pathological stage, although PSA varied remarkably at different disease stages.15 In 1990 the independent studies of Lilja‘ and Stenman7et al indicated that PSA is predominantly complexed to a l antichymotrypsin, and the complex, PSA-a1-antichymotrypsin, is the major form of PSA that may be measured by immunoassay in the serum of patients with benign prostatic hyperplasia (BPH)? Later Christensson et al reported that the noncomplexed form of PSA constitutes a minor fraction that is much smaller in untreated prostatic cancer than in BPH.s Stamey et al showed that pooled serum from patients with prostatic cancer who have different tumor volumes a t different stages and after failed radiation or hormonal therapy did not have significantly different fractions of PSA-al-antichymotrypsin? It is well known that the use of the free-to-total PSA ratio, also termed percentage of free PSA, enhances the specificity of PSA and is clinically valuable for distinguishing BPH from prostatic However, only a few attempts have
been made to use the ratio of free-to-total PSA for staging prostatic cancer.” lo,13-’‘ We investigated the usefulness of the ratio of free-to-total PSA for preoperative staging, and to determine cancer volume and tumor differentiation in patients with previously untreated prostatic cancer. MATERIALS AND METHODS
Our study included 5 15 consecutive patients thought to have prostate cancer due to elevated PSA (greater than 4 ng./ml.) or abnormal digital rectal examination who were referred by private practitioners for further prostatic evaluation. All patients underwent digital rectal examination and systematic sextant biopsies of the prostate via transrectal ultrasonography. Before prostatic manipulation serum was obtained for PSA testing. Total and free PSA was determined by the monoclonal immunoassay, Immulite DPC.? The ratio of free-to-total PSA was calculated using the formula, (free PSNtotal PSA) x 100. Patients with clinical localized prostatic cancer underwent radical retropubic prostatectomy when estimated life expectancy was at least 10 years and frozen sections of the obturator lymph nodes were negative. Patients with positive nodes a t the final evaluation and those who received hormonal therapy before radical retropubic prostatectomy were excluded from study. All radical retropubic prostatectomy specimens were prospectively processed according to the Stanford protocol. After surface marking and fixation the prostates were sectioned Accepted for publication December 26, 1997. * Requests for re rints Urologische Universitatsklinik, Martin- at 3 mm. intervals and paraffin embedded. Slides ( 5 pm.) istr. 52, 20246 Harnturg, ‘Germany. were cut and stained with hematoxylin and eosin. Total and Su ported by the Dr. -Karl Wilder Foundation. Gleason grade 4 cancer volume was determined after the Egtor’s Note: This article is the third of 5 published in this issue for which category 1 CME credits can be earned. In- area of cancer was marked with computer planimetry and structions for obtaining credits are given with the questions on pages 2144 and 2145. 2004
tDiagnostic Products Corporation, Los Angeles, California.
FREE-TO-TOTAL PROSTATE SPECIFIC ANTIGEN FOR STAGING PROSTATE CANCER
multiplied by section thickness (3 mrn.).l7 For pathological staging we used the second revision of the fourth edition of the TNM classification and for histological grading we used the Gleason system." The Mann-Whitney U and KruskalWallis tests were performed to compare medians and the Pearson correlation coefficient (r) was determined for regression analysis. The Bonferoni-Holm method was used for multiple testing procedures with statistical significance considered at p c0.05.
2005
TABLE2. Clinical stage in 307 patients with prostatic cancer and 170 who underwent radical retropubic prostatectomy 4E Clinical Stage
Prostate Ca Radical retropubic prostatectomy
Normal (Tlc)
Firm
17.9 29.4
32.6 20.0
B
C
38.8 50.6
10.7
-
ity of PSA testing. Stamey et a1 initially calculated a quotient of 3.5 ng./ml./gm. from preoperative serum PSA (polyclonal Yang assay) and cancer volume on the basis of a high correlation (r = 0.70)." Many studies have shown that serum PSA correlates well with advancing clinical and pathological stage, and staging accuracy may be enhanced by combining the results of serum PSA testing with preoperative variables, such as prostate biopsy Gleason grade, clinical disease stage and transrectal ultrasonography.'-*. 20,21 Most recently, we have shown that preoperative PSA only is unable to predict precisely cancer volume in peripheral and transition zone cancer^.^ The ratio of free-to-total PSA has proved useful for enhancing the specificity of PSA to distinguish between malignant and benign prostatic disease.6*7v10-'2 Therefore, it seems logical to use this ratio for staging prostatic cancer. To our knowledge information on the usefulness of free PSA in staging prostatic cancer is based on a small number of patients.9.10.13-16 pannek et a1 reported the largest series of 301 cases.16 Using pooled serum of patients with prostatic cancer at various stages Stamey et al showed that the fraction of serum PSA complexed to al-antichymotrypsin remains relatively constant even when hormonal or radiation treatment fails.g Catalona et a1 described the relationship of 50 radical retropubic prostatectomy specimens of organ confined and advanced prostatic cancer with the free-to-total PSA ratio." In contrast to our study, advanced disease included not only pathological stage pT3, but also lymph node positive disease. All patients in our study had negative obturator nodes, were thoroughly evaluated according to the Stanford protocol and, therefore, were clearly divided according to pathological stage. Most recently, Lerner13 and Bangma15 et a1 investigated the molecular forms of PSA and their relationship to stage, grade and deoxyribonucleic acid ploidy in prostatic cancer. There was no statistically significant difference among pathological tumor stages and grades, and the free-to-total PSA ratio in either study. However, Lerner et a1 described statistically significant differences in free PSA values among pathological substages as well as borderline significance (p = 0.05) for the ratio of complexed-to-total PSA.13 To our knowledge the clinical importance of the latter observation has yet to be proved. Preliminary data of Graefen et a1 from our institution showed no statistically significant differences in the ratio of free-to-total PSA among tumor stages, volumes and differentiation. l4 DISCUSSION We currently report a large series of patients in whom the The preoperative estimation of tumor volume in patients ratio of free-to-total PSA was assessed for staging purposes with prostate cancer has been a challenging problem for in those who underwent radical retropubic prostatectomy. many years and it remains unresolved despite the availabil- The current data of Graefen et a1 confirm our preliminary results that the free-to-total PSA ratio does not differ significantly among pathological tumor stages and volurne~.'~ However, subdividing stage pT3 disease into tumors with TABLE1. Data on 515 patients diagnosed with BPH or microscopic extracapsular extension (pT3a + b) and seminal prostatic cancer vesicle invasion (pT3c) reveals borderline significance withRadical out clinical relevance (fig. 1, B ) . There seemed to be no trend BPH b t a t i c Ca Retmpubic toward a decreasing or increasing ratio of free-to-total PSA No. Pts. (208 pts.) (307 pts.) h t a w m y with advancing tumor stage. However, in contrast to our (170 pts.)' preliminary data, a larger number of patients reveal a trend 64.5(44-86) 61.9 (44-79) 62.2 (31-88) Mean age (range) of a decreasing free-to-total PSA ratio among the clinically 24.8 (9.3) 14.1(8.0) Mean ngM. (median) 8.0 (6.4) significant and predominant Gleason scores (fig. 2, B ) . The 0'7 the PSA (median) 16.7 (14.6) 10.6 (8.8) 10.0(7.8) data of McNeal et a1 support our suggestion of differenti* * RESULTS
Of the 515 patients 208 (40.4%) had BPH and 307 (59.6%) had prostatic cancer. Of the 307 men with prostatic cancer 170 (55.4%) underwent radical retropubic prostatectomy. Tables 1 and 2 show basic data on the 3 groups (BPH, prostatic cancer and radical retropubic prostatectomy) and prostatic cancer clinical stages, respectively. At presentation preoperative PSA was in the normal range in 10% of the patients who underwent radical retropubic prostatectomy, in the diagnostic gray zone of 4 to 10 ng./ml. in 54.7% and greater than 10 ng./ml. in 35.3%. Pathological examination of the radical retropubic prostatectomy specimens revealed stage pT2 tumors in 91 patients (53.5%),pT3 in 71 (41.8%) and pT4 in the remaining 8 (4.7%). Serum PSA was significantly higher in stage pT3 than in pT2 disease (p <0.0001), and tumor volume correlated well with advancing pathological stage (p <0.0001). No statistical significance (p = 0.93) was noted in the free-to-total PSA ratio in stages pT2 and pT3 disease (fig. 1, A). To distinguish between patients who had extracapsular extension without (stage pT3a + b) and with (stage pT3c) seminal vesicle invasion those who underwent radical retropubic prostatectomy were further divided into 3 groups (fig. 1,B ) . The Kruskal-Wallis test revealed borderline significance (p = 0.052), which the multiple BonferroniHolm test indicated was due to a statistical difference between the median free-to-total PSA ratio values of stages pT3a + b and pT3c cancer (p = 0.015 < p = 0.052/2). Total and Gleason grade 4 cancer volume was available in all cases. PSA correlated well with total cancer volume (p <0.0001) but the free-to-total PSA ratio did not correlate with total (p = 0.71) or Gleason grade 4 (p = 0.94) cancer volume. Gleason scores were grouped according to clinical significancel7 as 3 or less + 3 , 3 + 4 , 4 + 3 and 4 or greater + 4. Of the 170 radical retropubic prostatectomy specimens Gleason grade was 4 in 116 (68.2%). Mean serum PSA was 17.2 ng./ml. (median 8.5) in these 116 patients and 8.5 ng./ml. (median 7.5) in those with a pure grade 3 or less Gleason Pattern (p = 0.03). There was a statistically significant difference (p = 0.03) in the ratio of free-to-total PSA among the 4 Gleason score groups (fig. 2, A). Excluding the 8 cases of pure Gleason grade 4 cancer, no significance was noted (P = 0.07) but there was a trend toward a decreasing ratio of free-to-total PSA (fig. 2, B ) .
2006
FREE-TO-TOTAL PROSTATE SPECIFIC ANTIGEN FOR STAGING PROSTATE CANCER
FIG.1. Box plot of percentage of free PSA (“SA%). A, stages pT2 and pT3 prostate cancer (P = 0.93). B7stages PT2, PT3a+b and PT3c prostate cancer (p = 0.052).
FIG.2. Box plot of percentage of free PSA WSA%)in radical prostatectomy specimens.A, Gleason score groups (p = 0.03). B, Gleason score groups excluding pure Gleason grade 4 (p = 0.07).
REFERENCES ating Gleason grades 3 + 4 and 4 + 3 disease, since an increasing Gleason grade 4 composition indicates an in1. Stamey, T. A., Yang, N., Hay, A. R., McNeal, J. E., Freiha, F. S. creased probability of more advanced disease.17 However, and Redwine, E.: Prostate-specific antigen as a serum marker for adenocarcinoma of the prostate. New Engl. J. Med., 317: to interpret figure 2, A, it is necessary to know the statis909, 1987. tical power of the Kruskal-Wallis test. Since it determined 2. Oesterling, J. E., Chan, D. W., Epstein, J. I., Kimball, A. W., Jr., any significance among the 4 Gleason score groups, the Bruzek, D. J., Ruck, R. C., Brendler, C. B. and Walsh, P. C.: value of p = 0.03 is questionable. In figure 2, B p = 0.07 is Prostate specific antigen in the preoperative and postoperative presumably more powerful, because it also shows a trend evaluation of localized prostate cancer treated with radical toward a lower median free-to-total PSA ratio with cancer prostatectomy. J. Urol., 139 766, 1988. dedifferentiation. 3. Stamey, T. A. and Kabalin, J. N.: Prostate specific antigen in the The underlying background of these studiess~10~13-16 is diagnosis and treatment of adenocarcinoma of the prostate. I. that cancer of greater volume or more dedifferentiation may Untreated patients. J. Urol., 141: 1070, 1989. have a smaller free-to-total PSA ratio than well differenti4. Partin, A. W., Carter, H. B., Chan, D. W., Epstein, J. I., Oesterling, J. E., Rock, R. C., Weber, J. P. and Walsh, P. C.: ated or small tumors. However, the mechanism of complexProstate specific antigen in the staging of localized prostate ing PSA to a1-antichymotrypsin in BPH and prostatic cancer cancer: influence of tumor differentiation, tumor volume and remains unclear. Therefore, Bjartell et a1 performed immubenign hyperplasia. J. Urol., 143: 747, 1990. nocytochemical studies for al-antichymotrypsin immuno5. Noldus, J. and Stamey, T. A.: Limitations of serum prostate staining of neuroendocrine cells, basal epithelium and secrespecific antigen in predicting peripheral and transition zone tory epithelium of the prostate.22Neither the basal epithelial cancer volumes as measured by correlation coefficients. nor neuroendocrine cells were positive for al-antichymotrypsin J. Urol., 156 232, 1996. immunostaining. Moreover, Bjork et al showed the lack of al6.Lilja, H., Christensson, A,, Dahlen, U., Matikainen, M.-T., antichymotrypsin immunostaining in BPH tissue that conNilsson, O., Pettersson, K. and Uvgren, T.: Prostate-specific antigen in human serum occurs predominantly in complex tained PSA, whereas prostatic cancer tissue with a low Gleason with a-1-antichymotrypsin. Clin. Chem., 37: 1618, 1991. score stained positive for FSA and al-anti~hymotrypsin.~~ Those series presupposed that the condition for a1- 7. Stenman, U. H., Leinonen, J., AlRhan, H., Rannikko, S., Tuhkanen, K. and Alfthan, 0.: A complex between prostateantichymotrypsin complexing to PSA is less than optimal in specific antigen and alpha-1-antichymotrypsinis the major BPH tissue, and that al-antichymotrypsin and PSA in prostatic form of prostate-specific antigen in serum of patients with cancer tissue enhance the complex formation between PSA and prostatic cancer: assay of the complex improves clinical sensial-antichymotrypsin. tivity for cancer. Cancer Res., 51: 222, 1991. CONCLUSIONS
Use of the ratio of free-to-total PSA enhances the specificity of PSA for distinguishing benign from malignant prostatic lesions. However, that ratio provides no additional diagnostic information with respect to pathological tumor stage, volume or grade than total PSA only.
8. Christensson, A., Bjork, T., Nilsson, O., Dahlen, U., Matikainen, M.-T., Cockett, A. T. K, Abrahamsson, P. and Lilja, H.: Serum prostate specific antigen complexed to a-1-antichymotrypsin as an indicator of prostate cancer. J. Urol., 150 100, 1993. 9. Stamey, T. A., Chen, 2. and F’restigiacomo, A,: Serum prostate specific antigen binding al-antichymotrypsin: influence of cancer volume, location and therapeutic selection of resistant clones. J. Urol., 152 1510, 1994.
FREE-TO-TOTAL PROSTATE SPECIFIC ANTIGEN FOR STAGING PROSTATE CANCER 10. Catalona, W. J., Smith, D. S., Wolfert, R. L., Wang, T. J., Rittenhouse, H. G., Ratliff T. L. and Nadler, R. B.: Evaluation of percentage of free serum prostate-specific antigen to improve specificity of prostate cancer screening. J.A.M.A., 274 1214, 1995. 11. Luderer, A. A., Chen, Y.-T., Soriano, T. F., Kramp, W. J., Carlson, G., Cuny, C., Sharp, T., Smith, W., Pettewav. J.. Brawer, M. K. and Thiel, R.: Measurement of the proportl'on of free to total prostate-specific antigen improves diagnostic performance of prostate-specific antigen in the diagnostic gray zone of total prostate-specific antigen. Urology, 46 187, 1995. 12. Prestigiacomo, A. F., Lilja, H., Pettersson, K., Wolfert, R. L. and Stamey, T. A,: A comparison of the free fraction of serum prostate specific antigen in men with benign and cancerous prostates: the best case scenario. J . Urol., 156 350 1996. 13. Lerner, S. E., Jacobsen, S. J., Lilja, H., Bergstralh, E. J., Ransom, J., Klee, G. G., Piironen, T., Blute, M. L., Lieber, M. M., Zincke, H., Pettersson, K., Peterson, D. and Oesterling, J. E.: Free, complexed, and total serum prostate-specific antigen concentrations and their proportions in predicting stage, grade, and deoxyribonucleic acid ploidy in patients with adenocarcinoma of the prostate. Urology, 48 240, 1996. 14. Graefen, M., Hammerer, P., Henke, P., Hilz, H., Huland, E. and Huland, H.: Percentage of free PSA does not correlate with pathological outcome. J. Urol., part 2, 155 A370, abstact 237, 1996. 15. Bangma, C. H., Kranse, R., Blijenberg, B. G. and Schroder, F. H.: The free-to-total serum prostate specific antigen ratio for staging prostate carcinoma. J . Urol., 157:544, 1997. 16 Pannek, J., Subong, E. N., Jones, K. A,, Marschke, P. L., Epstein, J. I., Chan, D. W., Carter, H. B., Luderer, A. A. and Partin, A. W.: The role of fredtotal prostate-specific antigen ratio in the prediction of final pathologic stage for men with clinically localized prostate cancer. Urology, 48.51, 1996. 17 McNeal, J . E., Villers, A. A,, Redwine, E. A., Freiha, F. S. and Stamey, T. A,: Histologic differentiation, cancer volume, and pelvic lymph node metastasis in adenocarcinoma of the prostate. Cancer, 66:1225. 1990. 18. Gleason, D. F.: Histolo&c grading and clinical staging of prostatic carcinoma. In: Urologic Pathology: The Prostate. Edited by M. Tannenbaum. Philadelphia: Lea & Febiger, chapt. 9, pp. 171-197, 1977. 19. Stamey, T. A,, Kabalin, J. N., McNeal, J . E., Johnstone, I. M., Freiha, R., Redwine, E. A. and Yang, N.: Prostate specific antigen in the diagnosis and treatment of adenocarcinoma of the prostate. 11. Radical prostatectomv treated patients. J. Uiol., 141: 1076, 1989. 20. Partin, A. W.. Yoo. J.. Carter, H. B.. Pearson, J. D.. Chan. D. W., Epstein, J.' I. and Walsh, P. C.:'The use of prostate specific antigen, clinical stage and Gleason score to predict pathological stage in men with localized prostate cancer. J . Urol., 150 110, 1993. 21. Blackwell, K.L., Bostwick, D. G., Myers, R. P., Zincke, H. and Oesterling, J. E.: Combining prostate specific antigen with cancer and gland volume to predict more reliably pathological stage: the influence of prostate specific antigen cancer density. J. Urol., 151: 1565, 1994. 22. Bjartell, A,, Bjork, T., Matikainen, M. T., Abrahamsson, P. A., di Sant'Agnese, A. and Lilja, H.: Production of alpha-lantichymotrypsin by PSA-containing cells of the human prostate epithelium. Urology, 42: 502, 1993. 23. Bjork, T., Bjartell, A,, Abrahamsson, P. A., Hulkko, S., di proSant'Agnese, A. and Lilja, H.: Alpha-1-antichymotrypsin duction in PSA-producing cells is common in prostate cancer but rare in benign prostatic hyperplasia. Urology, 43: 427. 1994
EDITORIAL COMMENT Most investigators now agree that percent free PSA performs better than total pSA alone in discriminating between men with Prostate cancer and those without, when the total PSA is slightly elevated within a "diagnostic gray zone." That zone has varied between 2.0 and 20 nglml. depending on the study. Outside of this zone percent free PSA has been of less value since the total PSA discnmlnates well between prostate cancer and benign disease. Because the Percent free PSA increases with increasing age and prostate volume
2007
in a11 men (with and without prostate cancer) and decreases with increasing total PSA, the reported performance of percent free PSA, including the sensitivity, specificity, and cut point values, has varied as study populations have differed with regard to these critical parameters. Therefore, it is not surprising that various investigators studying the correlation between percent free PSA and final pathological stage in men with clinically localized prostate cancer have come to differing conclusions about the value of percent free PSA in predicting final pathological stage after radical prostatectomy. Again the distribution of patient age, prostate volume and total PSA in the study populations has the potential to alter the results significantly. Other study design differences, such as the inclusion or exclusion of patients with a positive digital rectal examination and the choice of PSA assay, have an unknown impact. In this study the authors failed to identify a n association between percent free PSA and final pathological stage or tumor volume in men undergoing radical prostatectomy. However, they did identify a trend toward lower percent free PSA and final Gleason grade. Their study population was comprised of men referred for evaluation because of a PSA greater than 4.0 ngJml. or a n abnormal digital rectal examination. Of the 515 consecutive patients referred for prostate biopsy 307 (59.6%) were diagnosed with prostate cancer. The fact that most studies have found only a 25 to 30%,cancer detection rate using a similar screening algorithm suggests the possibility of the existence of other selection biases. For example, a more aggressive biopsy strategy in men known to have a lower percent free PSA, or inclusion of a significant number of older men or men with relatively high total PSA might have significantly biased the study population. The mean PSA in the study group with prostate cancer was 24.8 ng./ml. and the mean PSA of men undergoing radical prostatectomy was 14.1 ng./ml., suggesting that this study population was comprised of men with relatively high total PSA. In addition, only 17.9% of the patients with prostate cancer had a normal digital rectal examination (cTlc) and only 10.0% had a PSA in the normal range, indicating that the majority of prostate cancer patients had an elevated PSA and an abnormal digital rectal examination. Of the patients undergoing radical prostatectomy only 53.5%)had pT2 disease and 41.8% had pT3 disease. Thus, this study population was comprised of men with relatively high stage prostate cancer and with high total PSA. Interestingly, when the authors subdivided pT3 cases into those with or without seminal vesicle invasion, they detected a borderline statistically significant association with percent free PSA. While the authors stated that this distribution was clinically irrelevant, the higher recurrence rate in patients with seminal vesicle invasion compared to those with simply extracapsular extension makes this distinction clinically important. Finally, exclusion of patients with lymph node involvement a t the time of surgery may have obscured an association between percent free PSA and final pathological stage. Other studies including men with more limited disease have revealed a correlation between final pathological stage and percent free PSA. Along with other institutions, we recently participated in a multicenter clinical trial evaluating the ability of percent free PSA to distinguish men with prostate cancer from those without prostate cancer.' The study population included men with a normal digital rectal examination and PSA between 4.0 and 10.0 ng./ml. In the men with prostate cancer who underwent radical prostatectomy lower percent free PSA correlated significantly with more advanced final pathological stage and higher final Gleason score. Only 1 of 12 patients (8%)with seminal vesicle invasion or lymph node involvement after radical prostatectomy had a percent free PSA above a cut point of 156. Using a definition of favorable pathology as final Gleason score less than 7, pT1-pT2, negative nodes and tumor volume (less than lo%), only 345%of our study population with a percent free PSA of 15% or less had favorable pathology, while 754 with a percent free PSA of greater than 15% were categorized as favorable. In this select group of patients with moderately elevated PSA the percent free PSA performed similarly to PSA density in separating prostate cancer patients from those with benign disease. Multivariate logistic regression analyses have been reported from institutions attempting to develop algorithms to predict the extent of prostate cancer in men undergoing radical prostatectomy. Both analyses revealed that a PSA density of less than 0.1 to 0.15 was 1 of several variables that together were predictive of insignificant prostate cancer.2.3 In our multicenter study a percent free PSA cut point of 15% had a sensitivity and specificity similar to a PSA density cut point of 0.15 on receiver operator characteristics curve analysis. While they
2008
FREE-TO-TOTAL PROSTATE SPECIFIC ANTIGEN FOR STAGING PROSTATE CANCER
may be uaefd for the prediction of finalpathological stage, a percent free PSA cut point of 25% or a PSA density cut point of 0.07 was necessary to maintain 95% sensitivity in detecting prostate cancer.
Kevin M . Slawin Scott Department of Urology The Baylor Prostate Center Houston, Texas 1. Catalona, W. J., Partin, A. W., Slawin, K. M.,Brawer, M. K., Patel, A., Flanigan, R. C., Richie, J. P., deKernion, J. B., Walsh, P. C., Scardino, P. T., Lange, P. H., Herschman, J . D., Subong, E. N. P., Petteway, J. C., Parson, R. E., Loveland,
K G., Gasior, G. H., and Southwick, P. C.: A multicenter clinical trial evaluation of free PSA in the differentiation of D r o s t a t e cancer from benim - disease. J. Urol., part 2,167:111. abstract 434, 1997. 2. Goto. Y..Ohori. M.. Arakawa. A.. Kattan, M. W., Wheeler. T. W. and Scardino, P.’T.: Distin’&shing clinically important from unimportant prostate cancers before treatment: value of systematic biopsies. J. Urol., 156: 1059, 1996. 3. Epstein, J. I., Walsh, P. C., Carmichael, M. and Brendler, C. B.: Pathologic and clinical findings to predict tumor extent of nonpalpable (stage Tlc) prostate cancer. J.A.M.A., 271: 368, 1994. ~
Vol. 159,2004-2008, June 1998 Printed in U.S.A.
THE JOURNAL OF UROLOGY
Copyright 8 1998 by AMERICAN UROLOCICAL ASSOCIATION, INC
THE VALUE OF THE RATIO OF FREE-TO-TOTAL PROSTATE SPECIFIC ANTIGEN FOR STAGING PURPOSES IN PREVIOUSLY UNTREATED PROSTATE CANCER JOACHIM NOLDUS,* MARKUS GRAEFEN, EDITH HULAND, CHRISTOPHER BUSCH, PETER HAMMERER AND HARTWIG HULAND From the Department of Urology, University Hospital Eppendorf, Hamburg, Germany
ABSTRACT
Purpose: We analyzed the use of the ratio of free-to-total prostate specific antigen (PSA), also termed percentage of free PSA, for predicting tumor stage, volume and grade in patients with clinically localized prostate cancer. Materials and Methods: A total of 515 consecutive patients underwent further prostate evaluation due to elevated PSA (greater than 4.0 ng./ml.) or abnormal digital rectal examination. Prostate cancer was diagnosed in 307 patients (59.6%), including 170 (55.4%)who underwent radical retropubic prostatectomy. Data on pathological stage, Gleason grade, and total and Gleason grade 4 cancer volume were available in all patients. In the remaining 208 men (40.4%) benign prostate hyperplasia was diagnosed. Total and free PSA was measured in preoperative serum. Results: Total PSA was significantly higher (p <0.0001) in the 71 men with stage pT3 tumors than in the 91 with pT2 disease. Eight patients had stage pT4 tumors. Cancer volume correlated well with advancing pathological stage (p <0.0001)and total PSA (p <0.0001).The free-to-total PSA ratio was not significantly different (p = 0.93) in stages pT2 and pT3 tumors, and it did not correlate with total (p = 0.71) or pure Gleason grade 4 (p = 0.94) cancer volume. However, the ratio of free-to-total PSA tended to decrease (p = 0.07) in tumors of increasing Gleason grade. Conclusions: The ratio of free-to-total PSA does not help in the preoperative prediction of final tumor stage and volume. However, disease grading may alter the free-to-total PSA ratio. KEY WORDS: prostate, prostatic neoplasms, prostate-specific antigen Prostate specific antigen (PSA), the most sensitive serum marker in men with prostatic diseases, is produced and released almost exclusively by the prostatic epithelial cells. Because even benign conditions may increase PSA, it is not the ideal tumor marker. On the contrary, approximately 20% of all men with prostate cancer present at diagnosis with serum PSA concentrations in the normal range. Others have demonstrated that serum PSA correlates well with advancing clinical or pathological stage, although PSA varied remarkably at different disease stages.15 In 1990 the independent studies of Lilja‘ and Stenman7et al indicated that PSA is predominantly complexed to a l antichymotrypsin, and the complex, PSA-a1-antichymotrypsin, is the major form of PSA that may be measured by immunoassay in the serum of patients with benign prostatic hyperplasia (BPH)? Later Christensson et al reported that the noncomplexed form of PSA constitutes a minor fraction that is much smaller in untreated prostatic cancer than in BPH.s Stamey et al showed that pooled serum from patients with prostatic cancer who have different tumor volumes a t different stages and after failed radiation or hormonal therapy did not have significantly different fractions of PSA-al-antichymotrypsin? It is well known that the use of the free-to-total PSA ratio, also termed percentage of free PSA, enhances the specificity of PSA and is clinically valuable for distinguishing BPH from prostatic However, only a few attempts have
been made to use the ratio of free-to-total PSA for staging prostatic cancer.” lo,13-’‘ We investigated the usefulness of the ratio of free-to-total PSA for preoperative staging, and to determine cancer volume and tumor differentiation in patients with previously untreated prostatic cancer. MATERIALS AND METHODS
Our study included 5 15 consecutive patients thought to have prostate cancer due to elevated PSA (greater than 4 ng./ml.) or abnormal digital rectal examination who were referred by private practitioners for further prostatic evaluation. All patients underwent digital rectal examination and systematic sextant biopsies of the prostate via transrectal ultrasonography. Before prostatic manipulation serum was obtained for PSA testing. Total and free PSA was determined by the monoclonal immunoassay, Immulite DPC.? The ratio of free-to-total PSA was calculated using the formula, (free PSNtotal PSA) x 100. Patients with clinical localized prostatic cancer underwent radical retropubic prostatectomy when estimated life expectancy was at least 10 years and frozen sections of the obturator lymph nodes were negative. Patients with positive nodes a t the final evaluation and those who received hormonal therapy before radical retropubic prostatectomy were excluded from study. All radical retropubic prostatectomy specimens were prospectively processed according to the Stanford protocol. After surface marking and fixation the prostates were sectioned Accepted for publication December 26, 1997. * Requests for re rints Urologische Universitatsklinik, Martin- at 3 mm. intervals and paraffin embedded. Slides ( 5 pm.) istr. 52, 20246 Harnturg, ‘Germany. were cut and stained with hematoxylin and eosin. Total and Su ported by the Dr. -Karl Wilder Foundation. Gleason grade 4 cancer volume was determined after the Egtor’s Note: This article is the third of 5 published in this issue for which category 1 CME credits can be earned. In- area of cancer was marked with computer planimetry and structions for obtaining credits are given with the questions on pages 2144 and 2145. 2004
tDiagnostic Products Corporation, Los Angeles, California.
FREE-TO-TOTAL PROSTATE SPECIFIC ANTIGEN FOR STAGING PROSTATE CANCER
multiplied by section thickness (3 mrn.).l7 For pathological staging we used the second revision of the fourth edition of the TNM classification and for histological grading we used the Gleason system." The Mann-Whitney U and KruskalWallis tests were performed to compare medians and the Pearson correlation coefficient (r) was determined for regression analysis. The Bonferoni-Holm method was used for multiple testing procedures with statistical significance considered at p c0.05.
2005
TABLE2. Clinical stage in 307 patients with prostatic cancer and 170 who underwent radical retropubic prostatectomy 4E Clinical Stage
Prostate Ca Radical retropubic prostatectomy
Normal (Tlc)
Firm
17.9 29.4
32.6 20.0
B
C
38.8 50.6
10.7
-
ity of PSA testing. Stamey et a1 initially calculated a quotient of 3.5 ng./ml./gm. from preoperative serum PSA (polyclonal Yang assay) and cancer volume on the basis of a high correlation (r = 0.70)." Many studies have shown that serum PSA correlates well with advancing clinical and pathological stage, and staging accuracy may be enhanced by combining the results of serum PSA testing with preoperative variables, such as prostate biopsy Gleason grade, clinical disease stage and transrectal ultrasonography.'-*. 20,21 Most recently, we have shown that preoperative PSA only is unable to predict precisely cancer volume in peripheral and transition zone cancer^.^ The ratio of free-to-total PSA has proved useful for enhancing the specificity of PSA to distinguish between malignant and benign prostatic disease.6*7v10-'2 Therefore, it seems logical to use this ratio for staging prostatic cancer. To our knowledge information on the usefulness of free PSA in staging prostatic cancer is based on a small number of patients.9.10.13-16 pannek et a1 reported the largest series of 301 cases.16 Using pooled serum of patients with prostatic cancer at various stages Stamey et al showed that the fraction of serum PSA complexed to al-antichymotrypsin remains relatively constant even when hormonal or radiation treatment fails.g Catalona et a1 described the relationship of 50 radical retropubic prostatectomy specimens of organ confined and advanced prostatic cancer with the free-to-total PSA ratio." In contrast to our study, advanced disease included not only pathological stage pT3, but also lymph node positive disease. All patients in our study had negative obturator nodes, were thoroughly evaluated according to the Stanford protocol and, therefore, were clearly divided according to pathological stage. Most recently, Lerner13 and Bangma15 et a1 investigated the molecular forms of PSA and their relationship to stage, grade and deoxyribonucleic acid ploidy in prostatic cancer. There was no statistically significant difference among pathological tumor stages and grades, and the free-to-total PSA ratio in either study. However, Lerner et a1 described statistically significant differences in free PSA values among pathological substages as well as borderline significance (p = 0.05) for the ratio of complexed-to-total PSA.13 To our knowledge the clinical importance of the latter observation has yet to be proved. Preliminary data of Graefen et a1 from our institution showed no statistically significant differences in the ratio of free-to-total PSA among tumor stages, volumes and differentiation. l4 DISCUSSION We currently report a large series of patients in whom the The preoperative estimation of tumor volume in patients ratio of free-to-total PSA was assessed for staging purposes with prostate cancer has been a challenging problem for in those who underwent radical retropubic prostatectomy. many years and it remains unresolved despite the availabil- The current data of Graefen et a1 confirm our preliminary results that the free-to-total PSA ratio does not differ significantly among pathological tumor stages and volurne~.'~ However, subdividing stage pT3 disease into tumors with TABLE1. Data on 515 patients diagnosed with BPH or microscopic extracapsular extension (pT3a + b) and seminal prostatic cancer vesicle invasion (pT3c) reveals borderline significance withRadical out clinical relevance (fig. 1, B ) . There seemed to be no trend BPH b t a t i c Ca Retmpubic toward a decreasing or increasing ratio of free-to-total PSA No. Pts. (208 pts.) (307 pts.) h t a w m y with advancing tumor stage. However, in contrast to our (170 pts.)' preliminary data, a larger number of patients reveal a trend 64.5(44-86) 61.9 (44-79) 62.2 (31-88) Mean age (range) of a decreasing free-to-total PSA ratio among the clinically 24.8 (9.3) 14.1(8.0) Mean ngM. (median) 8.0 (6.4) significant and predominant Gleason scores (fig. 2, B ) . The 0'7 the PSA (median) 16.7 (14.6) 10.6 (8.8) 10.0(7.8) data of McNeal et a1 support our suggestion of differenti* * RESULTS
Of the 515 patients 208 (40.4%) had BPH and 307 (59.6%) had prostatic cancer. Of the 307 men with prostatic cancer 170 (55.4%) underwent radical retropubic prostatectomy. Tables 1 and 2 show basic data on the 3 groups (BPH, prostatic cancer and radical retropubic prostatectomy) and prostatic cancer clinical stages, respectively. At presentation preoperative PSA was in the normal range in 10% of the patients who underwent radical retropubic prostatectomy, in the diagnostic gray zone of 4 to 10 ng./ml. in 54.7% and greater than 10 ng./ml. in 35.3%. Pathological examination of the radical retropubic prostatectomy specimens revealed stage pT2 tumors in 91 patients (53.5%),pT3 in 71 (41.8%) and pT4 in the remaining 8 (4.7%). Serum PSA was significantly higher in stage pT3 than in pT2 disease (p <0.0001), and tumor volume correlated well with advancing pathological stage (p <0.0001). No statistical significance (p = 0.93) was noted in the free-to-total PSA ratio in stages pT2 and pT3 disease (fig. 1, A). To distinguish between patients who had extracapsular extension without (stage pT3a + b) and with (stage pT3c) seminal vesicle invasion those who underwent radical retropubic prostatectomy were further divided into 3 groups (fig. 1,B ) . The Kruskal-Wallis test revealed borderline significance (p = 0.052), which the multiple BonferroniHolm test indicated was due to a statistical difference between the median free-to-total PSA ratio values of stages pT3a + b and pT3c cancer (p = 0.015 < p = 0.052/2). Total and Gleason grade 4 cancer volume was available in all cases. PSA correlated well with total cancer volume (p <0.0001) but the free-to-total PSA ratio did not correlate with total (p = 0.71) or Gleason grade 4 (p = 0.94) cancer volume. Gleason scores were grouped according to clinical significancel7 as 3 or less + 3 , 3 + 4 , 4 + 3 and 4 or greater + 4. Of the 170 radical retropubic prostatectomy specimens Gleason grade was 4 in 116 (68.2%). Mean serum PSA was 17.2 ng./ml. (median 8.5) in these 116 patients and 8.5 ng./ml. (median 7.5) in those with a pure grade 3 or less Gleason Pattern (p = 0.03). There was a statistically significant difference (p = 0.03) in the ratio of free-to-total PSA among the 4 Gleason score groups (fig. 2, A). Excluding the 8 cases of pure Gleason grade 4 cancer, no significance was noted (P = 0.07) but there was a trend toward a decreasing ratio of free-to-total PSA (fig. 2, B ) .
2006
FREE-TO-TOTAL PROSTATE SPECIFIC ANTIGEN FOR STAGING PROSTATE CANCER
FIG.1. Box plot of percentage of free PSA (“SA%). A, stages pT2 and pT3 prostate cancer (P = 0.93). B7stages PT2, PT3a+b and PT3c prostate cancer (p = 0.052).
FIG.2. Box plot of percentage of free PSA WSA%)in radical prostatectomy specimens.A, Gleason score groups (p = 0.03). B, Gleason score groups excluding pure Gleason grade 4 (p = 0.07).
REFERENCES ating Gleason grades 3 + 4 and 4 + 3 disease, since an increasing Gleason grade 4 composition indicates an in1. Stamey, T. A., Yang, N., Hay, A. R., McNeal, J. E., Freiha, F. S. creased probability of more advanced disease.17 However, and Redwine, E.: Prostate-specific antigen as a serum marker for adenocarcinoma of the prostate. New Engl. J. Med., 317: to interpret figure 2, A, it is necessary to know the statis909, 1987. tical power of the Kruskal-Wallis test. Since it determined 2. Oesterling, J. E., Chan, D. W., Epstein, J. I., Kimball, A. W., Jr., any significance among the 4 Gleason score groups, the Bruzek, D. J., Ruck, R. C., Brendler, C. B. and Walsh, P. C.: value of p = 0.03 is questionable. In figure 2, B p = 0.07 is Prostate specific antigen in the preoperative and postoperative presumably more powerful, because it also shows a trend evaluation of localized prostate cancer treated with radical toward a lower median free-to-total PSA ratio with cancer prostatectomy. J. Urol., 139 766, 1988. dedifferentiation. 3. Stamey, T. A. and Kabalin, J. N.: Prostate specific antigen in the The underlying background of these studiess~10~13-16 is diagnosis and treatment of adenocarcinoma of the prostate. I. that cancer of greater volume or more dedifferentiation may Untreated patients. J. Urol., 141: 1070, 1989. have a smaller free-to-total PSA ratio than well differenti4. Partin, A. W., Carter, H. B., Chan, D. W., Epstein, J. I., Oesterling, J. E., Rock, R. C., Weber, J. P. and Walsh, P. C.: ated or small tumors. However, the mechanism of complexProstate specific antigen in the staging of localized prostate ing PSA to a1-antichymotrypsin in BPH and prostatic cancer cancer: influence of tumor differentiation, tumor volume and remains unclear. Therefore, Bjartell et a1 performed immubenign hyperplasia. J. Urol., 143: 747, 1990. nocytochemical studies for al-antichymotrypsin immuno5. Noldus, J. and Stamey, T. A.: Limitations of serum prostate staining of neuroendocrine cells, basal epithelium and secrespecific antigen in predicting peripheral and transition zone tory epithelium of the prostate.22Neither the basal epithelial cancer volumes as measured by correlation coefficients. nor neuroendocrine cells were positive for al-antichymotrypsin J. Urol., 156 232, 1996. immunostaining. Moreover, Bjork et al showed the lack of al6.Lilja, H., Christensson, A,, Dahlen, U., Matikainen, M.-T., antichymotrypsin immunostaining in BPH tissue that conNilsson, O., Pettersson, K. and Uvgren, T.: Prostate-specific antigen in human serum occurs predominantly in complex tained PSA, whereas prostatic cancer tissue with a low Gleason with a-1-antichymotrypsin. Clin. Chem., 37: 1618, 1991. score stained positive for FSA and al-anti~hymotrypsin.~~ Those series presupposed that the condition for a1- 7. Stenman, U. H., Leinonen, J., AlRhan, H., Rannikko, S., Tuhkanen, K. and Alfthan, 0.: A complex between prostateantichymotrypsin complexing to PSA is less than optimal in specific antigen and alpha-1-antichymotrypsinis the major BPH tissue, and that al-antichymotrypsin and PSA in prostatic form of prostate-specific antigen in serum of patients with cancer tissue enhance the complex formation between PSA and prostatic cancer: assay of the complex improves clinical sensial-antichymotrypsin. tivity for cancer. Cancer Res., 51: 222, 1991. CONCLUSIONS
Use of the ratio of free-to-total PSA enhances the specificity of PSA for distinguishing benign from malignant prostatic lesions. However, that ratio provides no additional diagnostic information with respect to pathological tumor stage, volume or grade than total PSA only.
8. Christensson, A., Bjork, T., Nilsson, O., Dahlen, U., Matikainen, M.-T., Cockett, A. T. K, Abrahamsson, P. and Lilja, H.: Serum prostate specific antigen complexed to a-1-antichymotrypsin as an indicator of prostate cancer. J. Urol., 150 100, 1993. 9. Stamey, T. A., Chen, 2. and F’restigiacomo, A,: Serum prostate specific antigen binding al-antichymotrypsin: influence of cancer volume, location and therapeutic selection of resistant clones. J. Urol., 152 1510, 1994.
FREE-TO-TOTAL PROSTATE SPECIFIC ANTIGEN FOR STAGING PROSTATE CANCER 10. Catalona, W. J., Smith, D. S., Wolfert, R. L., Wang, T. J., Rittenhouse, H. G., Ratliff T. L. and Nadler, R. B.: Evaluation of percentage of free serum prostate-specific antigen to improve specificity of prostate cancer screening. J.A.M.A., 274 1214, 1995. 11. Luderer, A. A., Chen, Y.-T., Soriano, T. F., Kramp, W. J., Carlson, G., Cuny, C., Sharp, T., Smith, W., Pettewav. J.. Brawer, M. K. and Thiel, R.: Measurement of the proportl'on of free to total prostate-specific antigen improves diagnostic performance of prostate-specific antigen in the diagnostic gray zone of total prostate-specific antigen. Urology, 46 187, 1995. 12. Prestigiacomo, A. F., Lilja, H., Pettersson, K., Wolfert, R. L. and Stamey, T. A,: A comparison of the free fraction of serum prostate specific antigen in men with benign and cancerous prostates: the best case scenario. J . Urol., 156 350 1996. 13. Lerner, S. E., Jacobsen, S. J., Lilja, H., Bergstralh, E. J., Ransom, J., Klee, G. G., Piironen, T., Blute, M. L., Lieber, M. M., Zincke, H., Pettersson, K., Peterson, D. and Oesterling, J. E.: Free, complexed, and total serum prostate-specific antigen concentrations and their proportions in predicting stage, grade, and deoxyribonucleic acid ploidy in patients with adenocarcinoma of the prostate. Urology, 48 240, 1996. 14. Graefen, M., Hammerer, P., Henke, P., Hilz, H., Huland, E. and Huland, H.: Percentage of free PSA does not correlate with pathological outcome. J. Urol., part 2, 155 A370, abstact 237, 1996. 15. Bangma, C. H., Kranse, R., Blijenberg, B. G. and Schroder, F. H.: The free-to-total serum prostate specific antigen ratio for staging prostate carcinoma. J . Urol., 157:544, 1997. 16 Pannek, J., Subong, E. N., Jones, K. A,, Marschke, P. L., Epstein, J. I., Chan, D. W., Carter, H. B., Luderer, A. A. and Partin, A. W.: The role of fredtotal prostate-specific antigen ratio in the prediction of final pathologic stage for men with clinically localized prostate cancer. Urology, 48.51, 1996. 17 McNeal, J . E., Villers, A. A,, Redwine, E. A., Freiha, F. S. and Stamey, T. A,: Histologic differentiation, cancer volume, and pelvic lymph node metastasis in adenocarcinoma of the prostate. Cancer, 66:1225. 1990. 18. Gleason, D. F.: Histolo&c grading and clinical staging of prostatic carcinoma. In: Urologic Pathology: The Prostate. Edited by M. Tannenbaum. Philadelphia: Lea & Febiger, chapt. 9, pp. 171-197, 1977. 19. Stamey, T. A,, Kabalin, J. N., McNeal, J . E., Johnstone, I. M., Freiha, R., Redwine, E. A. and Yang, N.: Prostate specific antigen in the diagnosis and treatment of adenocarcinoma of the prostate. 11. Radical prostatectomv treated patients. J. Uiol., 141: 1076, 1989. 20. Partin, A. W.. Yoo. J.. Carter, H. B.. Pearson, J. D.. Chan. D. W., Epstein, J.' I. and Walsh, P. C.:'The use of prostate specific antigen, clinical stage and Gleason score to predict pathological stage in men with localized prostate cancer. J . Urol., 150 110, 1993. 21. Blackwell, K.L., Bostwick, D. G., Myers, R. P., Zincke, H. and Oesterling, J. E.: Combining prostate specific antigen with cancer and gland volume to predict more reliably pathological stage: the influence of prostate specific antigen cancer density. J. Urol., 151: 1565, 1994. 22. Bjartell, A,, Bjork, T., Matikainen, M. T., Abrahamsson, P. A., di Sant'Agnese, A. and Lilja, H.: Production of alpha-lantichymotrypsin by PSA-containing cells of the human prostate epithelium. Urology, 42: 502, 1993. 23. Bjork, T., Bjartell, A,, Abrahamsson, P. A., Hulkko, S., di proSant'Agnese, A. and Lilja, H.: Alpha-1-antichymotrypsin duction in PSA-producing cells is common in prostate cancer but rare in benign prostatic hyperplasia. Urology, 43: 427. 1994
EDITORIAL COMMENT Most investigators now agree that percent free PSA performs better than total pSA alone in discriminating between men with Prostate cancer and those without, when the total PSA is slightly elevated within a "diagnostic gray zone." That zone has varied between 2.0 and 20 nglml. depending on the study. Outside of this zone percent free PSA has been of less value since the total PSA discnmlnates well between prostate cancer and benign disease. Because the Percent free PSA increases with increasing age and prostate volume
2007
in a11 men (with and without prostate cancer) and decreases with increasing total PSA, the reported performance of percent free PSA, including the sensitivity, specificity, and cut point values, has varied as study populations have differed with regard to these critical parameters. Therefore, it is not surprising that various investigators studying the correlation between percent free PSA and final pathological stage in men with clinically localized prostate cancer have come to differing conclusions about the value of percent free PSA in predicting final pathological stage after radical prostatectomy. Again the distribution of patient age, prostate volume and total PSA in the study populations has the potential to alter the results significantly. Other study design differences, such as the inclusion or exclusion of patients with a positive digital rectal examination and the choice of PSA assay, have an unknown impact. In this study the authors failed to identify a n association between percent free PSA and final pathological stage or tumor volume in men undergoing radical prostatectomy. However, they did identify a trend toward lower percent free PSA and final Gleason grade. Their study population was comprised of men referred for evaluation because of a PSA greater than 4.0 ngJml. or a n abnormal digital rectal examination. Of the 515 consecutive patients referred for prostate biopsy 307 (59.6%) were diagnosed with prostate cancer. The fact that most studies have found only a 25 to 30%,cancer detection rate using a similar screening algorithm suggests the possibility of the existence of other selection biases. For example, a more aggressive biopsy strategy in men known to have a lower percent free PSA, or inclusion of a significant number of older men or men with relatively high total PSA might have significantly biased the study population. The mean PSA in the study group with prostate cancer was 24.8 ng./ml. and the mean PSA of men undergoing radical prostatectomy was 14.1 ng./ml., suggesting that this study population was comprised of men with relatively high total PSA. In addition, only 17.9% of the patients with prostate cancer had a normal digital rectal examination (cTlc) and only 10.0% had a PSA in the normal range, indicating that the majority of prostate cancer patients had an elevated PSA and an abnormal digital rectal examination. Of the patients undergoing radical prostatectomy only 53.5%)had pT2 disease and 41.8% had pT3 disease. Thus, this study population was comprised of men with relatively high stage prostate cancer and with high total PSA. Interestingly, when the authors subdivided pT3 cases into those with or without seminal vesicle invasion, they detected a borderline statistically significant association with percent free PSA. While the authors stated that this distribution was clinically irrelevant, the higher recurrence rate in patients with seminal vesicle invasion compared to those with simply extracapsular extension makes this distinction clinically important. Finally, exclusion of patients with lymph node involvement a t the time of surgery may have obscured an association between percent free PSA and final pathological stage. Other studies including men with more limited disease have revealed a correlation between final pathological stage and percent free PSA. Along with other institutions, we recently participated in a multicenter clinical trial evaluating the ability of percent free PSA to distinguish men with prostate cancer from those without prostate cancer.' The study population included men with a normal digital rectal examination and PSA between 4.0 and 10.0 ng./ml. In the men with prostate cancer who underwent radical prostatectomy lower percent free PSA correlated significantly with more advanced final pathological stage and higher final Gleason score. Only 1 of 12 patients (8%)with seminal vesicle invasion or lymph node involvement after radical prostatectomy had a percent free PSA above a cut point of 156. Using a definition of favorable pathology as final Gleason score less than 7, pT1-pT2, negative nodes and tumor volume (less than lo%), only 345%of our study population with a percent free PSA of 15% or less had favorable pathology, while 754 with a percent free PSA of greater than 15% were categorized as favorable. In this select group of patients with moderately elevated PSA the percent free PSA performed similarly to PSA density in separating prostate cancer patients from those with benign disease. Multivariate logistic regression analyses have been reported from institutions attempting to develop algorithms to predict the extent of prostate cancer in men undergoing radical prostatectomy. Both analyses revealed that a PSA density of less than 0.1 to 0.15 was 1 of several variables that together were predictive of insignificant prostate cancer.2.3 In our multicenter study a percent free PSA cut point of 15% had a sensitivity and specificity similar to a PSA density cut point of 0.15 on receiver operator characteristics curve analysis. While they
2008
FREE-TO-TOTAL PROSTATE SPECIFIC ANTIGEN FOR STAGING PROSTATE CANCER
may be uaefd for the prediction of finalpathological stage, a percent free PSA cut point of 25% or a PSA density cut point of 0.07 was necessary to maintain 95% sensitivity in detecting prostate cancer.
Kevin M . Slawin Scott Department of Urology The Baylor Prostate Center Houston, Texas 1. Catalona, W. J., Partin, A. W., Slawin, K. M.,Brawer, M. K., Patel, A., Flanigan, R. C., Richie, J. P., deKernion, J. B., Walsh, P. C., Scardino, P. T., Lange, P. H., Herschman, J . D., Subong, E. N. P., Petteway, J. C., Parson, R. E., Loveland,
K G., Gasior, G. H., and Southwick, P. C.: A multicenter clinical trial evaluation of free PSA in the differentiation of D r o s t a t e cancer from benim - disease. J. Urol., part 2,167:111. abstract 434, 1997. 2. Goto. Y..Ohori. M.. Arakawa. A.. Kattan, M. W., Wheeler. T. W. and Scardino, P.’T.: Distin’&shing clinically important from unimportant prostate cancers before treatment: value of systematic biopsies. J. Urol., 156: 1059, 1996. 3. Epstein, J. I., Walsh, P. C., Carmichael, M. and Brendler, C. B.: Pathologic and clinical findings to predict tumor extent of nonpalpable (stage Tlc) prostate cancer. J.A.M.A., 271: 368, 1994. ~