- Email: [email protected]
The Free-to-Total Serum Prostate Specific Antigen Ratio for Staging Prostate Carcinoma
0022-5347/97/1572-0544$03.00/0
vol. 157, 544-547, February 1997 Printed rn U.S.A.
THE JOlJTWN. OF UROLOGY
Copyright Q 1997 by AMEFUCAN UROUGICAL. A ~ ~ ~ . I A INC TION,
THE FREE-TO-TOTAL SERUM PROSTATE SPECIFIC ANTIGEN RATIO FOR STAGING PROSTATE CARCINOMA CHRIS H. BANGMA, RIES KRANSE, BERT G. BLIJENBERG AND FRITZ H. SCHRODER From the Departments of Urology and CIinrcal Chemistry, Academic Hospital Rotterdam, Rotterdam, The Netherlands
ABSTRACT
Purpose: We analyzed the relationship between the free-to-total PSA ratio and prostate cancer tumor stage and grade compared to total serum PSA. Materials and Methods: In 123 patients clinical and pathological grade and stage were related to total serum PSA and free-to-total PSA ratio. Results: Total serum PSA paralleled clinical staging of prostate cancer. The distributions of total serum PSA and the free-to-total PSA ratio were significantly different between benign and malignant diseases (any stage), and between any T category and nodal disease. For serum PSA significant differences were noted between the distributions of men with locally confined (stages T1 and T2) and locally extended (stage T3) disease, and between all T categories and systemic metastatic disease. This finding was not noted for the free-to-total PSA ratio. Conclusions: The free-to-total PSA ratio has no additional value in clinical staging of prostate carcinoma compared to serum PSA. The free-to-total PSA ratio may be considered the result of cell differentiation and not a n indicator of tumor load. KEY WORDS:prostate-specific antigen, neoplasm staging, prostatic neoplasms
Box plots were used to illustrate the distribution of various PSA values, including the results of 50% of the participants with the median value plotted inside the box. The middle vertical line represented the 5th and 95th percentiles of all values. The values for total serum PSA and the free-to-total PSA ratio were measured in 1,659 controls without detected carcinoma and with benign prostatic hyperplasia (BPH), all of whom were participants of the randomized study of screening for prostate carcinoma.L2 Continuous parameters were compared with the MannWhitney U test. A 2-sided p = 0.05 level of significance was used. Multivariate analysis (forward inclusion based on Wilks statistics) was performed to analyze any additional benefit of total PSA or free PSA in predicting stage or grade.
Prostate carcinoma is staged clinically by physical examination and additional radiological procedures. Prostate specific antigen (PSA) has been of limited help in the differentiation between clinically important and unimportant carcinomas' and confined versus extended disease: and in prediction of pathological stage.3 New developments in serum PSA assays offer the possibility of distinguishing free circulating PSA molecules from those complexed to serum proteins, of which al-antichymotrypsin is predominant.4.5 In select groups the ratio of free-tocomplexed PSA improves the differentiation between prostate carcinoma and benign conditions.4.6-8 I t is not known whether the free-to-total PSA ratio may improve differentiation between various grades and stages of prostate cancer. We analyzed the relationship between the free-to-total PSA ratio, and prostate cancer tumor stage and grade, and assessed whether the free-to-total PSA ratio is better than total serum PSA in differentiating various disease stages.
RESULTS
PATIENTS AND METHODS
We studied 128 patients with untreated prostate carcinoma, including 67 recruited from a screening population and 61 from an outpatient clinic. Histological status was confirmed by ultrasound guided prostate biopsy. Clinical stage was assessed by digital rectal examination and bone scan according to the 1992 TNM classification.9 Abdominal computerized tomography was performed when indicated, usually for suspicion of positive pelvic lymph nodes due to increased serum PSA. The biopsy specimen was graded according to the Anderson classification.lO For all participants deep frozen serum ( - 70 C) was used to determine retrospectively total serum PSA and the free-tototal serum PSA ratio using a commercial assay, which has been evaluated extensively.11.12 The detection limits are less than 0.01 and less than 0.1 ng./ml. for free and total PSA, respectively. In 49 men with clinically confined prostate carcinoma pelvic lymph node dissection was performed, followed by radical prostatectomy. Pathological stage and grade of disease were analyzed separately from the clinical stage. Accepted for publication September 20, 1996
The table shows the distribution of grade and clinical stage among 123 study participants. Men with lower T categories had lower grades compared to those with higher T categories or metastatic disease. Figure 1,A shows box plots of the total serum PSA for 123 men with various clinical stages of prostate carcinoma according to locally advanced (stages T1 to TSNOMO), lymph node metastatic (stages T1 to T4N+MO) and systemic metastatic (stages T1 to T4NxM1) disease. Also a diagram of the 1,659 men without prostate carcinoma (controls) is shown. Figure 1, B shows box plots of the free-to-total PSA ratios for the same group. Considerable overlap exists between stages regarding total serum PSA and free-to-total PSA ratio. Remarkably, the most explicit difference is appre-
544
Distribution of grade and clinical stage in 128 patients Stage (No. cases)
Grade T1
T2
T3
T4
Nt
M+
1 2 3
16
3 6 5 1 15
2 2 4
-
4 16 10 30
-
Unknown Total No. cases
25 29 9 -1 64
8
2 26
2 1 ~
3
I 9
FREE-TO-TOTAL PROSTATE SPECIFIC ANTIGEN RATIO FOR STAGING PROSTATE CANCER
10000
100
j*
9
10
0.1
0 40
0 30 0 20
0 10 0 00
!
+ I
I
I
I
I
1
pA 6€L14+I BPH (1658)
T i (25)
T3 (191
T2 ( 5 6 )
M (81
N (14)
FIG.1. Total serum PSA (A) and free-to-total PSA (FIT) ratio ( B ) for 123 men with stages T1, T2, T3 tall NOMO), N + (T1 to T4N+MO) and M + (T1 to T4NxM1) prostate carcinoma, and 1,659 men with BPH.
ciated between BPH and malignant disease. Median free-tototal PSA ratio for men with BPH was 0.28. For the distributions of total serum PSA and free-to-total PSA ratio, significant differences were found between BPH and malignant disease (any stage), and between any T category and nodal disease (p <0.05). There was no significant difference between systemic metastatic disease and any of the T categories regarding free-to-total PSA ratio. However, for total serum PSA this finding was true only for category T3 (p = 0.06). Figure 2 shows total serum PSA and free-to-total PSA ratios for 81 patients with clinically confined prostate cancer
0.50
545
(stages T 1 and T2NxMO) and 19 with locally extended disease (stages T3 and T4NxMO). The free-to-total PSA ratio does not add to the differentiation between the 2 groups. Significant differences were found between the distributions of locally confined and locally extended disease for PSA (p = 0.02) but not for free-to-total PSA ratio (p = 0.08). Multivariate analysis was performed on the 101 men with clinically local and extended (nonmetastatic) disease. There was no significant additional benefit of free-to-total PSA ratio in prediction of disease stage. Total PSA showed the best prediction and classified 84.2% of the cases correctly. Figure 3 shows the distribution of total serum PSA and the free-to-total PSA ratio for various grades in 122 men with prostate carcinoma as well as a box plot for 1,659 controls. There was a significant difference between BPH and malignant disease for total serum PSA and free-to-total PSA ratio (p <0.05). Among men with prostate carcinoma no significant difference was found between median total serum PSA or free-to-total PSA ratio for the various grades (p >0.05). Multivariate analysis of these 123 men showed that free PSA was able to improve the predictive value of total PSA for grade significantly. However, the combination of free and total PSA was able to predict only 45.9% of cases correctly among the 3 grades. Figure 4 illustrates the distributions of total serum PSA and free-to-total PSA ratios according to pathological stage for 49 men who underwent radical prostatectomy. Two men had nodal micrometastases in the paraffin embedded sections after a negative perioperative frozen section of the lymph nodes. Clinical under staging was noted in 23 of these 49 patients and over staging was noted in 1. Regarding the distributions of total serum PSA and free-to-total PSA ratio, there was no statistical difference between locally confined (stages T1 and T2) and locally extended (stages T3 and T4) disease or nodal metastases (stage N+, p <0.05). DISCUSSION
In our 123 patients with clinically staged disease the freeto-total PSA ratio was compared to total serum PSA for ability to stage prostate carcinoma. Multivariate analysis confirmed that total serum PSA distinguished better between confined and nonconfined disease compared to free-to-
FIT Ratio (Prostatus)
0.401
n
A
n A
A
a
0.30
A
A.
A
A
O 0.10 ..i
0.00
I
0.1
I
I
I
I
IIIII
1
I
I
l l l ! l lI
I
I
I
I I I I I I
10
100
I
I
I i I l I
1000
Total PSA (Prostatus) [ng I mll
FIG.2. Total semm PSA "emus free-to-total PSA (FIT)ratio in 81 men with clinical stages T1 and T2 (A), and 19 with clinical etages T3 and T4 (0) prostate cancer (all stage NOMO).
546
FREE-TO-TOTAL PROSTATE SPECIFIC ANTIGEN RATIO FOR STAGING PROSTATE CANCER
grades 2 and 3 cancers. In similar groups of selected clinical patients such distributions are not uncommon. The number of patients is too small to analyze for grade among various stages. These distributional effects may be responsible for 1 100 the finding that the free-to-total PSA ratio differed significantly between men with local disease versus nodal metastases, while no significance was noted between men with local disease versus systemic metastases (fig. 1, B ) . Staging by PSA only'" or combined with biopsy grade" has shown previously t h a t a considerable overlap exists between various stages of disease. For clinical purposes often t h e most important differentiation i s that between locally confined BPH (16591 G1 (45) G2 (551 G3 (22) and metastatic disease for stratification of patients for diagnostic lymph node dissection. oo FIT m110 (ProSlslus) Tumor volume is the predominant factor influencing ser u m PSA. In our series i t was not possible to compare PSA to 0 90 tumor volume by measurements of pathological specimens. Therefore, stage probably represents tumor volume best in our study. Median total serum PSA increased with increas0 60 ing stage (fig. 1,A ) and grade (fig. 3, A), as in other studies. 050 Partin e t a1 reported on the correlation between grade and total serum PSA, and noted a n overall positive correlation between pathological Gleason score and serum PSA ( r = 0 20 0.30) in 350 men who underwent radical prostatectomy for 0 1 0 t - E l z clinically localized carcinoma." When corrected for tumor 0001 volume, a negative correlation (r = -0.37) was calculated. BPH (1659) G1 (45) 0 2 (55) G3 (22) Regarding the free-to-total PSA ratio, the median value FIG.3. Total serum PSA ( A )a n d free-to-totalPSA (Fl'13 ratio ( B ) for various grades ( G )in 122 m e n m t h prostate carcinoma a n d 1,659 does not parallel stage or grade as well as PSA (figs. 1and 3). The free-to-total PSA ratio differentiates well between bewith BPH. nign and malignant disease12 but not between local and metastatic cancer or locally confined and locally extended disease (figs. 1, B and 2). In figure 2 the free-to-total PSA ratio does not add to t h e discrimination between confined a n d extended local tumor. Free-to-total PSA ratio is a mathematical ratio, and its application is based on the observation t h a t free and complexed PSA forms appear in different concentrations i n men with versus those without prostate carcinoma. The biochemical pathway of the complexation of free PSA to crl-antichymotrypsin is not clear. However, immunohistochemical studies support the idea t h a t the free-to-total PSA ratio is determined by intracellular processes within the prostatic epithelial cells.14 I
0 1 '
T2 (30)
BPH (1659)
, 090 000
T3 (13)
1
I
T4 (4)
N (21
I
CONCLUSION F / l ra180 (ProSlalus)
--
I
070
060
0 50 0 4 o c
-
_-I
1
REFERENCES
:;:;:g L, T -F-1
I
~
_L
010
000 ' BPH (16591
T 2 130)
The free-to-total PSA ratio is of no additional value in clinical staging of prostate carcinoma compared to serum PSA. The free-to-total PSA ratio may be considered the result of cell differentiation and not a parameter of tumor load.
I
I
T3 (13)
7 4 (41
~
I N 121
FIG 4 Total serum PSA ( A )and free-to-total PSA ( F i T i ratio i B ) for 49 men with pathologxal stages T2, T3 (all NOMO), T4 and N t (T1 to T4N 1 MOI prostate carcinoma, and 1.659 with BPH
total PSA ratio. However, total serum PSA a s well a s freeto-total PSA ratio vaned considerably among stages and grades. Neither total PSA, free-to-total PSA ratio nor their combination could classify the majority of cases correctly according to grade. The interpretation of PSA among various stages in our study is biased by the uneven distribution of grades (table 1I. Almost 80% of grade 1 tumors were confined to the prostate (low volume), compared to 50 and 30%. respectively. of
1. Shinohara, K.. Wolf, J. S.. Jr.. Narayan, P. and Carroll, P. R.: Comparison of prostate specific antigen with prostate specific antigen density for 3 clinical applications. J. Urol., 152: 120, 1994. 2. Bluestein. D. L.. Bostwick. D. G., Bergstralh, E. J. and Ocsterling. .J. E: Eliminating t h e need for hilateral pelvic lyniphadencctomy in select patients with prostate cancer. J. Urnl.. 151: 1315. 1994. 3. ['artin. A. W.. Yoo. J . , Carter, H. B.. Prarson. J. D , Chan. D. W., Epstein. J. I. a n d Walsh, P. C . :Use of prostate specific antigen, clinical stage, a n d Gleason score to predict pathological stage in men with localized prostate cancer. J. Urol., 150 110, 1993. 4 . Stenman, U., Leinonen, J., Alfthan. H , Rannikko, S., Tuhkanen, K. a n d Alfthan, 0.:A complex between prostate-specific antigen and alpha 1-antichymotrypsin is t h e major form of prostate-specific antigen in serum of patients with prostatic cancer: assay of t h e complex improves clinical sensitivity for cancer. Cancer Res., 51: 222, 1991. 5. Liija, H.. Chnstensson, A.. Dahlen, LJ., Matikainen. M.-T., Nilsson. O., Pettersson, K. and Lovgren. T . : Prostatr-specific antigen in serum occurs predominantly i n complex with alpha 1-antichyrnotcpsin. Clin. Chcm.. 37: 1618. 1991.
FREE-TO-TOTAL PROSTATE SPECIFIC ANTIGEN RATIO FOR STAGING PROSTATE CANCER 6. Chnstensson, A., Bjork, T., Nilsson, 0.. Dahlen, U., Matikainen,
M.-T., Cockett, A. T. K., Abrahamsson, P. and Lilja, H.: Serum prostate specific antigen complexed to alpha-l-antichymotrypsin as a n indicator of prostate cancer. J. Urol., 1 5 0 100, 1993. 7. Lilja, H., Bjork, T., Abrahamsson, P., Stenman, U., Shaw, N., Dowell, B., Oesterling, J., Pettersson, T. K., Piironen, T. and Lovgren, T.: Improved separation between normals, benign prostatic hyperplasia (BPH), and carcinoma of the prostate by measuring free, complexed and total concentrations of prostate specific antigen. J. Urol., part 2, 151: 400A abstract 692, 1994. 8. King, C.,Friese, J., Lauren, L., Dowell, B., Shaw, N., Lilja, H., Stenman, U., Piironen, T. and Pettersson, K.: Measurement on IMx of free and total forms of prostate specific antigen for differentiation of patients with benign prostatic hyperplasia and prostate cancer. Clin. Chem., 4 0 1007, 1994. 9. Schroder, T.H.: The TNM classification of prostate carcinoma. Prostate, suppl., 4: 129, 1992. 10. Brawn, P. N.,Ayala, A. G., von Eschenbach, A. H., Hussey, D. H. and Johnson, D. E.: Histologic grading study of prostate carcinoma: the development of a new system and comparison
547
with other methods. A preliminary study. Cancer, 49 525, 1982. 11. Blijenberg, B. G . ,Bangma, C. H., Kranse, R. andSchroder, F. H.: Analytical evaluation of the new DELFIA free and total prostate-specific antigen assays. Read at International European Congress of Clinical Chemistry, Helsinki, Finland, 1995. 12. Bangma, C. H., Kranse, R., Blijenberg, B. G. and Schroder, F. H.: The value of screening tests in the detection of prostate cancer. Part I. Results of a retrospective evaluation of 1726 men. Part 11. A simulation of the role of the fredtotal prostate-specific antigen ratio, age-specific reference ranges, and PSA density. Urology, 4 6 773 and 779, 1995. 13. Bogdanowicz, J. F., Bentvelsen, F. M., Oosterom, R. and Schroeder, F. H.: Evaluation of prostate-specific antigen and prostatic acid phosphatase in untreated prostatic carcinoma and benign prostatic hyperplasia. Scand. J. Urol. Nephrol., suppl., 138 97, 1991. 14. Bjork, T., Bjartell, A,, Abrahamsson, P., Hulkko, S., Di Sant’agnese, A. and Lilja, H.: Alpha-1-antichymotrypypsinproduction in PSA-producing cells is common in prostate cancer but rare in benign prostatic hyperplasia. Urology, 43: 427, 1994.
vol. 157, 544-547, February 1997 Printed rn U.S.A.
THE JOlJTWN. OF UROLOGY
Copyright Q 1997 by AMEFUCAN UROUGICAL. A ~ ~ ~ . I A INC TION,
THE FREE-TO-TOTAL SERUM PROSTATE SPECIFIC ANTIGEN RATIO FOR STAGING PROSTATE CARCINOMA CHRIS H. BANGMA, RIES KRANSE, BERT G. BLIJENBERG AND FRITZ H. SCHRODER From the Departments of Urology and CIinrcal Chemistry, Academic Hospital Rotterdam, Rotterdam, The Netherlands
ABSTRACT
Purpose: We analyzed the relationship between the free-to-total PSA ratio and prostate cancer tumor stage and grade compared to total serum PSA. Materials and Methods: In 123 patients clinical and pathological grade and stage were related to total serum PSA and free-to-total PSA ratio. Results: Total serum PSA paralleled clinical staging of prostate cancer. The distributions of total serum PSA and the free-to-total PSA ratio were significantly different between benign and malignant diseases (any stage), and between any T category and nodal disease. For serum PSA significant differences were noted between the distributions of men with locally confined (stages T1 and T2) and locally extended (stage T3) disease, and between all T categories and systemic metastatic disease. This finding was not noted for the free-to-total PSA ratio. Conclusions: The free-to-total PSA ratio has no additional value in clinical staging of prostate carcinoma compared to serum PSA. The free-to-total PSA ratio may be considered the result of cell differentiation and not a n indicator of tumor load. KEY WORDS:prostate-specific antigen, neoplasm staging, prostatic neoplasms
Box plots were used to illustrate the distribution of various PSA values, including the results of 50% of the participants with the median value plotted inside the box. The middle vertical line represented the 5th and 95th percentiles of all values. The values for total serum PSA and the free-to-total PSA ratio were measured in 1,659 controls without detected carcinoma and with benign prostatic hyperplasia (BPH), all of whom were participants of the randomized study of screening for prostate carcinoma.L2 Continuous parameters were compared with the MannWhitney U test. A 2-sided p = 0.05 level of significance was used. Multivariate analysis (forward inclusion based on Wilks statistics) was performed to analyze any additional benefit of total PSA or free PSA in predicting stage or grade.
Prostate carcinoma is staged clinically by physical examination and additional radiological procedures. Prostate specific antigen (PSA) has been of limited help in the differentiation between clinically important and unimportant carcinomas' and confined versus extended disease: and in prediction of pathological stage.3 New developments in serum PSA assays offer the possibility of distinguishing free circulating PSA molecules from those complexed to serum proteins, of which al-antichymotrypsin is predominant.4.5 In select groups the ratio of free-tocomplexed PSA improves the differentiation between prostate carcinoma and benign conditions.4.6-8 I t is not known whether the free-to-total PSA ratio may improve differentiation between various grades and stages of prostate cancer. We analyzed the relationship between the free-to-total PSA ratio, and prostate cancer tumor stage and grade, and assessed whether the free-to-total PSA ratio is better than total serum PSA in differentiating various disease stages.
RESULTS
PATIENTS AND METHODS
We studied 128 patients with untreated prostate carcinoma, including 67 recruited from a screening population and 61 from an outpatient clinic. Histological status was confirmed by ultrasound guided prostate biopsy. Clinical stage was assessed by digital rectal examination and bone scan according to the 1992 TNM classification.9 Abdominal computerized tomography was performed when indicated, usually for suspicion of positive pelvic lymph nodes due to increased serum PSA. The biopsy specimen was graded according to the Anderson classification.lO For all participants deep frozen serum ( - 70 C) was used to determine retrospectively total serum PSA and the free-tototal serum PSA ratio using a commercial assay, which has been evaluated extensively.11.12 The detection limits are less than 0.01 and less than 0.1 ng./ml. for free and total PSA, respectively. In 49 men with clinically confined prostate carcinoma pelvic lymph node dissection was performed, followed by radical prostatectomy. Pathological stage and grade of disease were analyzed separately from the clinical stage. Accepted for publication September 20, 1996
The table shows the distribution of grade and clinical stage among 123 study participants. Men with lower T categories had lower grades compared to those with higher T categories or metastatic disease. Figure 1,A shows box plots of the total serum PSA for 123 men with various clinical stages of prostate carcinoma according to locally advanced (stages T1 to TSNOMO), lymph node metastatic (stages T1 to T4N+MO) and systemic metastatic (stages T1 to T4NxM1) disease. Also a diagram of the 1,659 men without prostate carcinoma (controls) is shown. Figure 1, B shows box plots of the free-to-total PSA ratios for the same group. Considerable overlap exists between stages regarding total serum PSA and free-to-total PSA ratio. Remarkably, the most explicit difference is appre-
544
Distribution of grade and clinical stage in 128 patients Stage (No. cases)
Grade T1
T2
T3
T4
Nt
M+
1 2 3
16
3 6 5 1 15
2 2 4
-
4 16 10 30
-
Unknown Total No. cases
25 29 9 -1 64
8
2 26
2 1 ~
3
I 9
FREE-TO-TOTAL PROSTATE SPECIFIC ANTIGEN RATIO FOR STAGING PROSTATE CANCER
10000
100
j*
9
10
0.1
0 40
0 30 0 20
0 10 0 00
!
+ I
I
I
I
I
1
pA 6€L14+I BPH (1658)
T i (25)
T3 (191
T2 ( 5 6 )
M (81
N (14)
FIG.1. Total serum PSA (A) and free-to-total PSA (FIT) ratio ( B ) for 123 men with stages T1, T2, T3 tall NOMO), N + (T1 to T4N+MO) and M + (T1 to T4NxM1) prostate carcinoma, and 1,659 men with BPH.
ciated between BPH and malignant disease. Median free-tototal PSA ratio for men with BPH was 0.28. For the distributions of total serum PSA and free-to-total PSA ratio, significant differences were found between BPH and malignant disease (any stage), and between any T category and nodal disease (p <0.05). There was no significant difference between systemic metastatic disease and any of the T categories regarding free-to-total PSA ratio. However, for total serum PSA this finding was true only for category T3 (p = 0.06). Figure 2 shows total serum PSA and free-to-total PSA ratios for 81 patients with clinically confined prostate cancer
0.50
545
(stages T 1 and T2NxMO) and 19 with locally extended disease (stages T3 and T4NxMO). The free-to-total PSA ratio does not add to the differentiation between the 2 groups. Significant differences were found between the distributions of locally confined and locally extended disease for PSA (p = 0.02) but not for free-to-total PSA ratio (p = 0.08). Multivariate analysis was performed on the 101 men with clinically local and extended (nonmetastatic) disease. There was no significant additional benefit of free-to-total PSA ratio in prediction of disease stage. Total PSA showed the best prediction and classified 84.2% of the cases correctly. Figure 3 shows the distribution of total serum PSA and the free-to-total PSA ratio for various grades in 122 men with prostate carcinoma as well as a box plot for 1,659 controls. There was a significant difference between BPH and malignant disease for total serum PSA and free-to-total PSA ratio (p <0.05). Among men with prostate carcinoma no significant difference was found between median total serum PSA or free-to-total PSA ratio for the various grades (p >0.05). Multivariate analysis of these 123 men showed that free PSA was able to improve the predictive value of total PSA for grade significantly. However, the combination of free and total PSA was able to predict only 45.9% of cases correctly among the 3 grades. Figure 4 illustrates the distributions of total serum PSA and free-to-total PSA ratios according to pathological stage for 49 men who underwent radical prostatectomy. Two men had nodal micrometastases in the paraffin embedded sections after a negative perioperative frozen section of the lymph nodes. Clinical under staging was noted in 23 of these 49 patients and over staging was noted in 1. Regarding the distributions of total serum PSA and free-to-total PSA ratio, there was no statistical difference between locally confined (stages T1 and T2) and locally extended (stages T3 and T4) disease or nodal metastases (stage N+, p <0.05). DISCUSSION
In our 123 patients with clinically staged disease the freeto-total PSA ratio was compared to total serum PSA for ability to stage prostate carcinoma. Multivariate analysis confirmed that total serum PSA distinguished better between confined and nonconfined disease compared to free-to-
FIT Ratio (Prostatus)
0.401
n
A
n A
A
a
0.30
A
A.
A
A
O 0.10 ..i
0.00
I
0.1
I
I
I
I
IIIII
1
I
I
l l l ! l lI
I
I
I
I I I I I I
10
100
I
I
I i I l I
1000
Total PSA (Prostatus) [ng I mll
FIG.2. Total semm PSA "emus free-to-total PSA (FIT)ratio in 81 men with clinical stages T1 and T2 (A), and 19 with clinical etages T3 and T4 (0) prostate cancer (all stage NOMO).
546
FREE-TO-TOTAL PROSTATE SPECIFIC ANTIGEN RATIO FOR STAGING PROSTATE CANCER
grades 2 and 3 cancers. In similar groups of selected clinical patients such distributions are not uncommon. The number of patients is too small to analyze for grade among various stages. These distributional effects may be responsible for 1 100 the finding that the free-to-total PSA ratio differed significantly between men with local disease versus nodal metastases, while no significance was noted between men with local disease versus systemic metastases (fig. 1, B ) . Staging by PSA only'" or combined with biopsy grade" has shown previously t h a t a considerable overlap exists between various stages of disease. For clinical purposes often t h e most important differentiation i s that between locally confined BPH (16591 G1 (45) G2 (551 G3 (22) and metastatic disease for stratification of patients for diagnostic lymph node dissection. oo FIT m110 (ProSlslus) Tumor volume is the predominant factor influencing ser u m PSA. In our series i t was not possible to compare PSA to 0 90 tumor volume by measurements of pathological specimens. Therefore, stage probably represents tumor volume best in our study. Median total serum PSA increased with increas0 60 ing stage (fig. 1,A ) and grade (fig. 3, A), as in other studies. 050 Partin e t a1 reported on the correlation between grade and total serum PSA, and noted a n overall positive correlation between pathological Gleason score and serum PSA ( r = 0 20 0.30) in 350 men who underwent radical prostatectomy for 0 1 0 t - E l z clinically localized carcinoma." When corrected for tumor 0001 volume, a negative correlation (r = -0.37) was calculated. BPH (1659) G1 (45) 0 2 (55) G3 (22) Regarding the free-to-total PSA ratio, the median value FIG.3. Total serum PSA ( A )a n d free-to-totalPSA (Fl'13 ratio ( B ) for various grades ( G )in 122 m e n m t h prostate carcinoma a n d 1,659 does not parallel stage or grade as well as PSA (figs. 1and 3). The free-to-total PSA ratio differentiates well between bewith BPH. nign and malignant disease12 but not between local and metastatic cancer or locally confined and locally extended disease (figs. 1, B and 2). In figure 2 the free-to-total PSA ratio does not add to t h e discrimination between confined a n d extended local tumor. Free-to-total PSA ratio is a mathematical ratio, and its application is based on the observation t h a t free and complexed PSA forms appear in different concentrations i n men with versus those without prostate carcinoma. The biochemical pathway of the complexation of free PSA to crl-antichymotrypsin is not clear. However, immunohistochemical studies support the idea t h a t the free-to-total PSA ratio is determined by intracellular processes within the prostatic epithelial cells.14 I
0 1 '
T2 (30)
BPH (1659)
, 090 000
T3 (13)
1
I
T4 (4)
N (21
I
CONCLUSION F / l ra180 (ProSlalus)
--
I
070
060
0 50 0 4 o c
-
_-I
1
REFERENCES
:;:;:g L, T -F-1
I
~
_L
010
000 ' BPH (16591
T 2 130)
The free-to-total PSA ratio is of no additional value in clinical staging of prostate carcinoma compared to serum PSA. The free-to-total PSA ratio may be considered the result of cell differentiation and not a parameter of tumor load.
I
I
T3 (13)
7 4 (41
~
I N 121
FIG 4 Total serum PSA ( A )and free-to-total PSA ( F i T i ratio i B ) for 49 men with pathologxal stages T2, T3 (all NOMO), T4 and N t (T1 to T4N 1 MOI prostate carcinoma, and 1.659 with BPH
total PSA ratio. However, total serum PSA a s well a s freeto-total PSA ratio vaned considerably among stages and grades. Neither total PSA, free-to-total PSA ratio nor their combination could classify the majority of cases correctly according to grade. The interpretation of PSA among various stages in our study is biased by the uneven distribution of grades (table 1I. Almost 80% of grade 1 tumors were confined to the prostate (low volume), compared to 50 and 30%. respectively. of
1. Shinohara, K.. Wolf, J. S.. Jr.. Narayan, P. and Carroll, P. R.: Comparison of prostate specific antigen with prostate specific antigen density for 3 clinical applications. J. Urol., 152: 120, 1994. 2. Bluestein. D. L.. Bostwick. D. G., Bergstralh, E. J. and Ocsterling. .J. E: Eliminating t h e need for hilateral pelvic lyniphadencctomy in select patients with prostate cancer. J. Urnl.. 151: 1315. 1994. 3. ['artin. A. W.. Yoo. J . , Carter, H. B.. Prarson. J. D , Chan. D. W., Epstein. J. I. a n d Walsh, P. C . :Use of prostate specific antigen, clinical stage, a n d Gleason score to predict pathological stage in men with localized prostate cancer. J. Urol., 150 110, 1993. 4 . Stenman, U., Leinonen, J., Alfthan. H , Rannikko, S., Tuhkanen, K. a n d Alfthan, 0.:A complex between prostate-specific antigen and alpha 1-antichymotrypsin is t h e major form of prostate-specific antigen in serum of patients with prostatic cancer: assay of t h e complex improves clinical sensitivity for cancer. Cancer Res., 51: 222, 1991. 5. Liija, H.. Chnstensson, A.. Dahlen, LJ., Matikainen. M.-T., Nilsson. O., Pettersson, K. and Lovgren. T . : Prostatr-specific antigen in serum occurs predominantly i n complex with alpha 1-antichyrnotcpsin. Clin. Chcm.. 37: 1618. 1991.
FREE-TO-TOTAL PROSTATE SPECIFIC ANTIGEN RATIO FOR STAGING PROSTATE CANCER 6. Chnstensson, A., Bjork, T., Nilsson, 0.. Dahlen, U., Matikainen,
M.-T., Cockett, A. T. K., Abrahamsson, P. and Lilja, H.: Serum prostate specific antigen complexed to alpha-l-antichymotrypsin as a n indicator of prostate cancer. J. Urol., 1 5 0 100, 1993. 7. Lilja, H., Bjork, T., Abrahamsson, P., Stenman, U., Shaw, N., Dowell, B., Oesterling, J., Pettersson, T. K., Piironen, T. and Lovgren, T.: Improved separation between normals, benign prostatic hyperplasia (BPH), and carcinoma of the prostate by measuring free, complexed and total concentrations of prostate specific antigen. J. Urol., part 2, 151: 400A abstract 692, 1994. 8. King, C.,Friese, J., Lauren, L., Dowell, B., Shaw, N., Lilja, H., Stenman, U., Piironen, T. and Pettersson, K.: Measurement on IMx of free and total forms of prostate specific antigen for differentiation of patients with benign prostatic hyperplasia and prostate cancer. Clin. Chem., 4 0 1007, 1994. 9. Schroder, T.H.: The TNM classification of prostate carcinoma. Prostate, suppl., 4: 129, 1992. 10. Brawn, P. N.,Ayala, A. G., von Eschenbach, A. H., Hussey, D. H. and Johnson, D. E.: Histologic grading study of prostate carcinoma: the development of a new system and comparison
547
with other methods. A preliminary study. Cancer, 49 525, 1982. 11. Blijenberg, B. G . ,Bangma, C. H., Kranse, R. andSchroder, F. H.: Analytical evaluation of the new DELFIA free and total prostate-specific antigen assays. Read at International European Congress of Clinical Chemistry, Helsinki, Finland, 1995. 12. Bangma, C. H., Kranse, R., Blijenberg, B. G. and Schroder, F. H.: The value of screening tests in the detection of prostate cancer. Part I. Results of a retrospective evaluation of 1726 men. Part 11. A simulation of the role of the fredtotal prostate-specific antigen ratio, age-specific reference ranges, and PSA density. Urology, 4 6 773 and 779, 1995. 13. Bogdanowicz, J. F., Bentvelsen, F. M., Oosterom, R. and Schroeder, F. H.: Evaluation of prostate-specific antigen and prostatic acid phosphatase in untreated prostatic carcinoma and benign prostatic hyperplasia. Scand. J. Urol. Nephrol., suppl., 138 97, 1991. 14. Bjork, T., Bjartell, A,, Abrahamsson, P., Hulkko, S., Di Sant’agnese, A. and Lilja, H.: Alpha-1-antichymotrypypsinproduction in PSA-producing cells is common in prostate cancer but rare in benign prostatic hyperplasia. Urology, 43: 427, 1994.