- Email: [email protected]
mL total PSA
ELSEVIER
FREE TO TOTAL PROSTATE-SPECIFIC ANTIGEN (PSA) RATIO IMPROVES THE DISCRIMINATION BETWEEN PROSTATE CANCER AND BENIGN PROSTATIC HYPERPLASIA (BPH) IN THE DIAGNOSTIC GRAY ZONE OF 1.8 TO 10 ng/mL TOTAL PSA PAUL J. VAN CANGH, BERTRAND TOMBAL,
PHILIPPE DE NAYER, LUC FRANCIS LORGE, FRANCQIS
DE VISCHER, PHILIPPE X. WESE, AND REINIER
SAUVAGE, OPSOMER
ABSTRACT Objectives. Improved discrimination between prostate cancer (PC) and benign prostatic hyperplasia (BPH) is clearly needed. Our aim in this study was to evaluate whether the free to total prostate-specific antigen (PSA) ratio would be useful in the gray zone of 1.8-l 0 ng/mL total PSA range. Methods. In a consecutive series of 435 clinic patients referred for prostate evaluation, 308 had a total PSA ~10 ng/mL (92 had PC and 216 BPH). Free and total PSA were measured, and the free to total PSA ratio calculated. Results. Total PSA values were significantly different between the two groups. For the 200 patients with a total PSA <6 ng/mL, no significant difference in total PSA values were seen (P = 0.41 I), whereas free to total PSA ratios remained statistically different (P < 0.001). Receiver operating characteristic (ROC) curve analysis comparing the performances of total PSA over the ratio of free to total PSA showed a clear advantage for the ratio at all sensitivity levels. Conclusions. These data demonstrate that in a significant number (n = 308) of prostatic patients in the diagnostic gray zone of 1.8-l 0 ng/mL total PSA, the routine use of free to total PSA might be advantageous in discriminating between cancer and benign hyperplasia. This advantage remained for total PSA <4 ng/mL. Further study is warranted to confirm these findings in an unselected population. Copyright 1996by Elsevier Science Inc. UROLOGY 48(6A): 67-70, 1996.
he diagnostic value of serum prostate-specific T antigen (PSA) is limited by its lack of specificity. The proportion of free PSA in the serum is higher in benign prostate hyperplasia (BPH) than in prostate cancer (PC), and measurement of the ratio of free to total PSA has therefore been recommended to improve diagnostic accuracy.3-5 The objective of this study was to evaluate the clinical utility of the free to total PSA ratio in the so-called gray zone of 1.8-10 ngImL total PSA where improved discrimination would be most desirable.
From the University of Louvain Medical School, Cliniques Universitaires Saint Luc, Brussels, Belgium Presented at the 9Ist ALJA annual Meeting, Orlando, Florida,
May 1996 Reprint Hospital,
requests: Paul L. Van Cangh, M.D., St. Luc University 10 Avenue Hippocrate, B-1200 Brussels, Belgium
COPYRIGHT 1996 BY ELSEVIER ALL RIGHTS RESERVED
SCIENCE
INC.
MATERIALS
AND
METHODS
Free PSA was measured by the immunoradiometric assay (IRMA) method provided by Hybritech@ (Hybritech Inc., San Diego, CA) and described previously.6 A first monoclonal antibody directed toward a specific site shared by the free and bound PSA is coated on plastic beads (capture antibody). The second antibody labeled with “‘1 is directed against a distinct antigenic site specific for the free-PSA form. After incubation the bead is washed, and the bound radioactivity is measured. All samples are assayed in duplicate. Total PSA was measured by the Hybritech Tandem@ assay, and the free to total PSA ratio was calculated by dividing free by total PSA measured values. PATIENTS From April 1994 to July 1995 we routinely measured free PSA concomitantly with total PSA in 435 consecutive clinic patients referred for prostate evaluation; 308 had a total PSA
67
TABLE
I.
Descriptive statistics for total PSA and free to total PSA All Patients (n = 308)
Total PSA (1.8-10 rig/W Median Mean (SD) Free to total PSA ratio Median Mean (SW Range
(n 52)
BPH (n = 216)
4.75
5.75
4.4
5.16 (2.29)
5.86(2.45)
4.86 (2.15)
0.13
0.21
0.22 (0.10)
0.23
0.15 (0.08)
0.032-0.417
0.24 (0.09) 0.08&0.5&8
transrectal ultrasound (TRUS) examination of the prostate. Prostate cancer was diagnosed histologically in 92 patients aged 54-85 years (mean 70); radical prostatectomy was performed in 66 cancer patients. Presumed benign prostatic hyperplasia was present in 216 patients aged from 4589 years (mean 68); of the BPH patients, 100 underwent transurethral or open adenomectomy (surgical group: histologically verified), and 116 carried a diagnosis of BPH based on the following clinical criteria: negative DRE, negative transrectal ultrasound, and stable sequential total PSA with a 1- 10 vear follow-un I (clinical group: no biopsy performed). ’
0.60
DATA ANALYSIS The Mann-Whitney test was used to evaluate the significance of the differences between the median values of total PSA and free to total PSA ratio, and between the two groups of BPH patients. Receiver operating characteristic (ROC) curves were constructed by plotting the true positive fraction (sensitivity) versus the true negative fraction (inverse of specificity). Areas under the curves were measured and compared using Medcalc@ software (Broekstraat 52, 9030 Mariakerke, Belgium) statistical analyses were performed with SPSS@ software (SPSS Inc., Chicago, IL)
RESULTS
Table I summarizes the respective value of total and free to total PSA ratio, and the scatter diagram (Fig. 1) illustrates their distribution for the 92 PC and 216 BPH patients. Median values of total PSA were clinically similar in both groups (5.75 and 4.4 for the medians of total PSA for PC and BPH), although statistically different by the Mann-Whitney test; at <6 ng/mL this statistical difference was no longer observed (Table II). The free to total PSA ratio differed significantly between the two groups, and this difference remained statistically significant even for total PSA below 4 ng/mL (Table II; P
-
Cl
BPH (n=216)
.
PC (n=92)
FIGURE 1. Scatter diagram of free to total PSA ratio versus total-PSA. Cancer: n = 92; BPH: n = 2 16. Total PSA concentration between 1.8 and 10.0 nglmL. Horizontal line: free to total PSA cut-off point of 0.25; Vertical line: classical 4.0 nglmL cut-off value for total PSA.
68
UROLOGY 48 @Xl, 19%
TABLE
II.
P values uccording
to
range of PSA PSA Range hzfmL1
n
Free to Total PSA Ratio uJ*1
Total PSA v*1
308 200 123
0.001 0.4 0.4
test
88% of the cancers while eliminating 44% of the BPH; for a sensitivity level of 92%, the cut-off value of free to total PSA ratio is 0.25, with a specificity of 32% (Fig. 1). The ROC curves comparing the relative performances of total PSA and free to total PSA ratio show a clear advantage for the ratio at all sensitivity levels (Fig. 2). The areas under the curve were 0.664 and 0.803 respectively (P < 0.0001). COMMENT
Our study, involving a significant number of patients from a single institution, confirms the value of free to total PSA measurement in the gray zone of total PSA
were found for similar groups of patients.2-5 An important aspect of our study is that 29/92 of PC patients had a total PSA below the classical cutoff value of 4 ng/mL; the left side of Fig. 1 clearly illustrates the persistently superior discriminatory power of free to total PSA ratio in that low total PSA range. One third of our PC population presented with a PSA <4.0 ng/mL; this proportion is larger than in most reported studies; in a recent study of 428 serum samples collected from many diverse clinical practices, only 141 165 or 8% of patients fell into that category.3 We explain this difference by our policy initiated in 1986 of setting the cut-off value of total PSA at 2.4 ng/mL. Nine of those 29 patients underwent radical prostatectomy; that those cancers were significant is demonstrated by the pathology findings of capsular perforation (pT3) in four, including two with positive margins. In this experience, extreme values of free to total PSA ratio carry an evident informative power, but cannot be regarded as diagnostic.3 In 7.5% (7/92) of PC patients, the free to total PSA ratio was >27%. The absolute diagnosis of BPH remains a challenge, and several patients included in that category probably have prostate cancer, in both the surgical and the clinical groups. In the present study, statistical comparison between the two groups of BPH patients reveals no detectable difference (Mann-Whitney); we therefore postulate
l-
0.2
0.4 0.6 True negative fraction (l-specificity)
+
total PSA (rig/ml) --f
f/t PSA (%)
0.8
1
I
FIGURE 2. ROC analysis. Comparison of free to total PSA ratio with total PSA for discrimination between prostate cancer and benign hyperplasia. Cancer: n = 92; BPH: n = 216. Total PSA concentration between 1.8 and 10.0 nglmL. UROLOGY
48 (6A), 1996
69
that the proportion of unrecognized cancer was similar in both groups and analyzed those patients together; exclusion of the clinical group does not modify the conclusions (data not shown). The literature is evolving rapidly on this topic, and selection of cut-off values vary according to each author’s concerns. A trade-off between sensitivity (detection of the largest possible number of cancers) and specificity (avoidance of unnecessary biopsies) has to be accepted. This is best illustrated by the ROC curve (Fig. 2) which helps in selecting the appropriate cut-off point by providing a qualitative comparison of accuracy of the two tests. In conclusion, in our urology clinic population of 308 patients with a total PSA
70
REFERENCES 1. Stenmann UH, Leinonem J, Alfthan H, Ranniko S, Tuhkanen K, and Alfthan 0: A complex between prostatespecific antigen and crt-antichymotrypsin is the major form of prostate-specific antigen in serum of patients with prostatic cancer: assay of the complex improves clinical sensitivity for cancer. Cancer Res 51: 222-226, 1991. 2. Catalona WJ, Smith DS, Wolfert RL, Wang TJ, Rittenhouse HG, Rathff TL, and Nadler RB: Evaluation of free serum prostate-specific antigen to improve specificity of prostate cancer screening. JAMA 274: 1214-1220, 1995. 3. Chen YT, Luderer AA, Thiel RP, Carlson G, Cuny CL, and Soriano TF: Using proportions of free to total prostatespecific antigen, age, and total prostate-specific antigen to predict the probability of prostate cancer. Urology 47: 518-524, 1996. 4. Luderer AA, Chen YT, Soriano TF, Kramp WJ, Carlson G, Cuny C, Sharp T, Smith W, Petteway J, Brawer MK, and Thiel R: Measurement of the proportion 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: 187-194, 1995. 5. Oesterling JE, Jacobsen SJ, Klee GG, Petterson K, Pironen P, Abrahamson PA, Stenman UH, Dowel1 B, Lougren T, and Lilja H: Free, complexed and total serum prostate specific antigen: the establishment of appropriate reference ranges for their concentrations and ratios. J Urol 144: 10901095,1995. 6. Van Cangh PJ, De Nayer P, Sauvage P, Tombal B, Elsen M, Lorge F, Opsomer R, and Wese FX. Free to total PSA ratio is superior to total PSA in differentiating benign prostate hypertrophy from prostate cancer. Prostate 30, 1996 (in press). 7. Campbell MJ and Machin D: Medical Statistics. A Commonsense Approach, (2nd Edition). Chichester, New York, John Wiley & Sons, 1993: 32-43.
UROLOGY
48 (6A), 1996
FREE TO TOTAL PROSTATE-SPECIFIC ANTIGEN (PSA) RATIO IMPROVES THE DISCRIMINATION BETWEEN PROSTATE CANCER AND BENIGN PROSTATIC HYPERPLASIA (BPH) IN THE DIAGNOSTIC GRAY ZONE OF 1.8 TO 10 ng/mL TOTAL PSA PAUL J. VAN CANGH, BERTRAND TOMBAL,
PHILIPPE DE NAYER, LUC FRANCIS LORGE, FRANCQIS
DE VISCHER, PHILIPPE X. WESE, AND REINIER
SAUVAGE, OPSOMER
ABSTRACT Objectives. Improved discrimination between prostate cancer (PC) and benign prostatic hyperplasia (BPH) is clearly needed. Our aim in this study was to evaluate whether the free to total prostate-specific antigen (PSA) ratio would be useful in the gray zone of 1.8-l 0 ng/mL total PSA range. Methods. In a consecutive series of 435 clinic patients referred for prostate evaluation, 308 had a total PSA ~10 ng/mL (92 had PC and 216 BPH). Free and total PSA were measured, and the free to total PSA ratio calculated. Results. Total PSA values were significantly different between the two groups. For the 200 patients with a total PSA <6 ng/mL, no significant difference in total PSA values were seen (P = 0.41 I), whereas free to total PSA ratios remained statistically different (P < 0.001). Receiver operating characteristic (ROC) curve analysis comparing the performances of total PSA over the ratio of free to total PSA showed a clear advantage for the ratio at all sensitivity levels. Conclusions. These data demonstrate that in a significant number (n = 308) of prostatic patients in the diagnostic gray zone of 1.8-l 0 ng/mL total PSA, the routine use of free to total PSA might be advantageous in discriminating between cancer and benign hyperplasia. This advantage remained for total PSA <4 ng/mL. Further study is warranted to confirm these findings in an unselected population. Copyright 1996by Elsevier Science Inc. UROLOGY 48(6A): 67-70, 1996.
he diagnostic value of serum prostate-specific T antigen (PSA) is limited by its lack of specificity. The proportion of free PSA in the serum is higher in benign prostate hyperplasia (BPH) than in prostate cancer (PC), and measurement of the ratio of free to total PSA has therefore been recommended to improve diagnostic accuracy.3-5 The objective of this study was to evaluate the clinical utility of the free to total PSA ratio in the so-called gray zone of 1.8-10 ngImL total PSA where improved discrimination would be most desirable.
From the University of Louvain Medical School, Cliniques Universitaires Saint Luc, Brussels, Belgium Presented at the 9Ist ALJA annual Meeting, Orlando, Florida,
May 1996 Reprint Hospital,
requests: Paul L. Van Cangh, M.D., St. Luc University 10 Avenue Hippocrate, B-1200 Brussels, Belgium
COPYRIGHT 1996 BY ELSEVIER ALL RIGHTS RESERVED
SCIENCE
INC.
MATERIALS
AND
METHODS
Free PSA was measured by the immunoradiometric assay (IRMA) method provided by Hybritech@ (Hybritech Inc., San Diego, CA) and described previously.6 A first monoclonal antibody directed toward a specific site shared by the free and bound PSA is coated on plastic beads (capture antibody). The second antibody labeled with “‘1 is directed against a distinct antigenic site specific for the free-PSA form. After incubation the bead is washed, and the bound radioactivity is measured. All samples are assayed in duplicate. Total PSA was measured by the Hybritech Tandem@ assay, and the free to total PSA ratio was calculated by dividing free by total PSA measured values. PATIENTS From April 1994 to July 1995 we routinely measured free PSA concomitantly with total PSA in 435 consecutive clinic patients referred for prostate evaluation; 308 had a total PSA
67
TABLE
I.
Descriptive statistics for total PSA and free to total PSA All Patients (n = 308)
Total PSA (1.8-10 rig/W Median Mean (SD) Free to total PSA ratio Median Mean (SW Range
(n 52)
BPH (n = 216)
4.75
5.75
4.4
5.16 (2.29)
5.86(2.45)
4.86 (2.15)
0.13
0.21
0.22 (0.10)
0.23
0.15 (0.08)
0.032-0.417
0.24 (0.09) 0.08&0.5&8
transrectal ultrasound (TRUS) examination of the prostate. Prostate cancer was diagnosed histologically in 92 patients aged 54-85 years (mean 70); radical prostatectomy was performed in 66 cancer patients. Presumed benign prostatic hyperplasia was present in 216 patients aged from 4589 years (mean 68); of the BPH patients, 100 underwent transurethral or open adenomectomy (surgical group: histologically verified), and 116 carried a diagnosis of BPH based on the following clinical criteria: negative DRE, negative transrectal ultrasound, and stable sequential total PSA with a 1- 10 vear follow-un I (clinical group: no biopsy performed). ’
0.60
DATA ANALYSIS The Mann-Whitney test was used to evaluate the significance of the differences between the median values of total PSA and free to total PSA ratio, and between the two groups of BPH patients. Receiver operating characteristic (ROC) curves were constructed by plotting the true positive fraction (sensitivity) versus the true negative fraction (inverse of specificity). Areas under the curves were measured and compared using Medcalc@ software (Broekstraat 52, 9030 Mariakerke, Belgium) statistical analyses were performed with SPSS@ software (SPSS Inc., Chicago, IL)
RESULTS
Table I summarizes the respective value of total and free to total PSA ratio, and the scatter diagram (Fig. 1) illustrates their distribution for the 92 PC and 216 BPH patients. Median values of total PSA were clinically similar in both groups (5.75 and 4.4 for the medians of total PSA for PC and BPH), although statistically different by the Mann-Whitney test; at <6 ng/mL this statistical difference was no longer observed (Table II). The free to total PSA ratio differed significantly between the two groups, and this difference remained statistically significant even for total PSA below 4 ng/mL (Table II; P
-
Cl
BPH (n=216)
.
PC (n=92)
FIGURE 1. Scatter diagram of free to total PSA ratio versus total-PSA. Cancer: n = 92; BPH: n = 2 16. Total PSA concentration between 1.8 and 10.0 nglmL. Horizontal line: free to total PSA cut-off point of 0.25; Vertical line: classical 4.0 nglmL cut-off value for total PSA.
68
UROLOGY 48 @Xl, 19%
TABLE
II.
P values uccording
to
range of PSA PSA Range hzfmL1
n
Free to Total PSA Ratio uJ*1
Total PSA v*1
308 200 123
0.001 0.4 0.4
test
88% of the cancers while eliminating 44% of the BPH; for a sensitivity level of 92%, the cut-off value of free to total PSA ratio is 0.25, with a specificity of 32% (Fig. 1). The ROC curves comparing the relative performances of total PSA and free to total PSA ratio show a clear advantage for the ratio at all sensitivity levels (Fig. 2). The areas under the curve were 0.664 and 0.803 respectively (P < 0.0001). COMMENT
Our study, involving a significant number of patients from a single institution, confirms the value of free to total PSA measurement in the gray zone of total PSA
were found for similar groups of patients.2-5 An important aspect of our study is that 29/92 of PC patients had a total PSA below the classical cutoff value of 4 ng/mL; the left side of Fig. 1 clearly illustrates the persistently superior discriminatory power of free to total PSA ratio in that low total PSA range. One third of our PC population presented with a PSA <4.0 ng/mL; this proportion is larger than in most reported studies; in a recent study of 428 serum samples collected from many diverse clinical practices, only 141 165 or 8% of patients fell into that category.3 We explain this difference by our policy initiated in 1986 of setting the cut-off value of total PSA at 2.4 ng/mL. Nine of those 29 patients underwent radical prostatectomy; that those cancers were significant is demonstrated by the pathology findings of capsular perforation (pT3) in four, including two with positive margins. In this experience, extreme values of free to total PSA ratio carry an evident informative power, but cannot be regarded as diagnostic.3 In 7.5% (7/92) of PC patients, the free to total PSA ratio was >27%. The absolute diagnosis of BPH remains a challenge, and several patients included in that category probably have prostate cancer, in both the surgical and the clinical groups. In the present study, statistical comparison between the two groups of BPH patients reveals no detectable difference (Mann-Whitney); we therefore postulate
l-
0.2
0.4 0.6 True negative fraction (l-specificity)
+
total PSA (rig/ml) --f
f/t PSA (%)
0.8
1
I
FIGURE 2. ROC analysis. Comparison of free to total PSA ratio with total PSA for discrimination between prostate cancer and benign hyperplasia. Cancer: n = 92; BPH: n = 216. Total PSA concentration between 1.8 and 10.0 nglmL. UROLOGY
48 (6A), 1996
69
that the proportion of unrecognized cancer was similar in both groups and analyzed those patients together; exclusion of the clinical group does not modify the conclusions (data not shown). The literature is evolving rapidly on this topic, and selection of cut-off values vary according to each author’s concerns. A trade-off between sensitivity (detection of the largest possible number of cancers) and specificity (avoidance of unnecessary biopsies) has to be accepted. This is best illustrated by the ROC curve (Fig. 2) which helps in selecting the appropriate cut-off point by providing a qualitative comparison of accuracy of the two tests. In conclusion, in our urology clinic population of 308 patients with a total PSA
70
REFERENCES 1. Stenmann UH, Leinonem J, Alfthan H, Ranniko S, Tuhkanen K, and Alfthan 0: A complex between prostatespecific antigen and crt-antichymotrypsin is the major form of prostate-specific antigen in serum of patients with prostatic cancer: assay of the complex improves clinical sensitivity for cancer. Cancer Res 51: 222-226, 1991. 2. Catalona WJ, Smith DS, Wolfert RL, Wang TJ, Rittenhouse HG, Rathff TL, and Nadler RB: Evaluation of free serum prostate-specific antigen to improve specificity of prostate cancer screening. JAMA 274: 1214-1220, 1995. 3. Chen YT, Luderer AA, Thiel RP, Carlson G, Cuny CL, and Soriano TF: Using proportions of free to total prostatespecific antigen, age, and total prostate-specific antigen to predict the probability of prostate cancer. Urology 47: 518-524, 1996. 4. Luderer AA, Chen YT, Soriano TF, Kramp WJ, Carlson G, Cuny C, Sharp T, Smith W, Petteway J, Brawer MK, and Thiel R: Measurement of the proportion 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: 187-194, 1995. 5. Oesterling JE, Jacobsen SJ, Klee GG, Petterson K, Pironen P, Abrahamson PA, Stenman UH, Dowel1 B, Lougren T, and Lilja H: Free, complexed and total serum prostate specific antigen: the establishment of appropriate reference ranges for their concentrations and ratios. J Urol 144: 10901095,1995. 6. Van Cangh PJ, De Nayer P, Sauvage P, Tombal B, Elsen M, Lorge F, Opsomer R, and Wese FX. Free to total PSA ratio is superior to total PSA in differentiating benign prostate hypertrophy from prostate cancer. Prostate 30, 1996 (in press). 7. Campbell MJ and Machin D: Medical Statistics. A Commonsense Approach, (2nd Edition). Chichester, New York, John Wiley & Sons, 1993: 32-43.
UROLOGY
48 (6A), 1996