- Email: [email protected]
EFFECTS OF SYSTEMATIC 12-CORE BIOPSY ON THE PERFORMANCE OF PERCENT FREE PROSTATE SPECIFIC ANTIGEN FOR PROSTATE CANCER DETECTION
0022-5347/04/1723-0900/0 THE JOURNAL OF UROLOGY® Copyright © 2004 by AMERICAN UROLOGICAL ASSOCIATION
Vol. 172, 900 –904, September 2004 Printed in U.S.A.
DOI: 10.1097/01.ju.0000134619.72675.8d
EFFECTS OF SYSTEMATIC 12-CORE BIOPSY ON THE PERFORMANCE OF PERCENT FREE PROSTATE SPECIFIC ANTIGEN FOR PROSTATE CANCER DETECTION EDUARDO I. CANTO, HERB SINGH, SHAHROKH F. SHARIAT, DOV KADMON, BRIAN J. MILES, THOMAS M. WHEELER* AND KEVIN M. SLAWIN†,‡ From the Baylor Prostate Center, Scott Department of Urology (EIC, HS, DK, BJM, KMS) and Department of Pathology (TMW), Baylor College of Medicine and Methodist Hospital (TMW), Houston and Department of Urology, University of Texas Southwestern (SFS), Dallas, Texas
ABSTRACT
Purpose: The performance characteristics of percent free (f) prostate specific antigen (PSA) for differentiating between benign prostatic hyperplasia and prostate cancer were originally established using primarily sextant biopsy. We determined whether the addition of 6 laterally directed cores to the traditional sextant prostate biopsy affects the performance of percent fPSA. Materials and Methods: We retrospectively evaluated a cohort of 350 consecutive biopsies in men with negative digital rectal examinations and PSA between 4 and 10 ng/ml who underwent systematic 12 core biopsy (S12C) biopsy at Scott Department of Urology between March 1999 and January 2003. The effects of 6 additional, laterally directed biopsies on the sensitivity, specificity and area under the ROC curve for percent fPSA was evaluated in the 277 men in whom percent fPSA was measured. Results: Cancers detected exclusively in the 6 laterally directed cores were associated with percent fPSA values similar to those in patients with a benign S12C biopsy. This resulted in a modest and yet predictable decrease in the sensitivity of percent fPSA at each biopsy threshold value without affecting specificity. There was a nonstatistically significant decrease in the area under the ROC curve with the addition of 6 laterally directed cores to sextant biopsy (medial sextant cores 0.66 vs S12C 0.60). Conclusions: The 12 core biopsy strategies have a higher cancer detection rate than sextant biopsies and they are gaining widespread acceptance. The addition of 6 laterally directed cores to traditional sextant biopsy may result in a modest decrease in the sensitivity of percent fPSA at each selected biopsy threshold without affecting specificity. KEY WORDS: prostate, biopsy, prostatic neoplasms, prostate-specific antigen
The fraction of total prostate specific antigen (PSA) comprised of free (f) PSA expressed as a percent is Food and Drug Administration approved for use in the enhancement of PSA for differentiating prostate cancer from benign prostatic hyperplasia when PSA is in the range of 4 to 10 ng/ml. The area under the ROC curve (AUC) of percent fPSA as a screening tool has been reported to be about 0.70 to 0.74.1, 2 Large multicenter studies have shown that the use of percent fPSA can decrease the number of unnecessary biopsies by about 20%, while maintaining 95% sensitivity.1–3 However, the validated biopsy threshold for percent fPSA is based on cancer detection using primarily sextant biopsy.2 Systematic 12 core (S12C) biopsy that includes the standard sextant biopsy cores and an additional 6 laterally directed biopsy cores has a significantly higher rate of prostate cancer detection than traditional sextant biopsy.4, 5 We reported that the addition of 6 laterally directed cores to traditional sextant biopsy increases the prostate cancer detection rate from 31% to 43%,4 a rate that approaches the overall cancer detection rate seen in serial sextant biopsy studies.6 Recent reports suggest that S12C biopsy detects the Accepted for publication April 23, 2004. * Financial interest and/or other relationship with Pintex. † Correspondence: Scott Department of Urology, Baylor College of Medicine, Baylor Prostate Center, 6560 Fannin St., Suite 2100, Houston, Texas 77030 (telephone: 713-798-8670; FAX: 713-798-8030; e-mail: [email protected]). ‡ Financial interest and/or other relationship with GlaxoSmithKline and National Institutes of Health.
highest number of significant and insignificant cancers.7, 8 In an analysis of 176 patients who underwent S12C biopsy followed by radical retropubic prostatectomies at Baylor College of Medicine and Methodist Hospital, Houston, Texas 15 of the 86 cancers (17%) with tumor volume greater than 0.5 cc were detected in the laterally directed cores only and, therefore, they would have been missed with a traditional sextant biopsy strategy.9 As more men undergo yearly PSA based prostate cancer screening, a larger proportion of those who are candidates for prostate biopsy have PSA in the 4 to 10 ng/ml range. Many of these men undergo percent fPSA testing as an adjunct to further stratify the risk of prostate cancer. Although S12C biopsy detects a greater number of clinically significant cancers, the cancers detected exclusively with sextant or laterally directed cores of S12C biopsy are on average smaller and the associated serum percent fPSA is higher than that of those detected with sextant plus laterally directed cores.9 Since many urologists now routinely perform at least 10 biopsy cores during prostate biopsy, it is important to understand how the differences in the nature of cancers detected through the additional cores may affect percent fPSA performance. Because the Food and Drug Administration approved percent fPSA biopsy threshold value was established using sextant biopsy and additional cancers detected by the lateral cores of S12C biopsy have lower associated serum percent fPSA values,9 we hypothesized that the percent fPSA thresh-
900
901
12-CORE BIOPSY AND PERCENT FREE PROSTATE SPECIFIC ANTIGEN
old for S12C biopsy may need to be lowered to maintain the same sensitivity as that based on sextant biopsy. MATERIALS AND METHODS
Patient selection. We retrospectively evaluated a cohort of 350 men with a negative digital rectal examination (DRE) and PSA between 4 and 10 ng/ml who underwent screening S12C biopsy at the Scott Department of Urology between March 1999 and January 2003. Subjects who had undergone an ablative procedure (ie transurethral resection, transurethral microwave ablation, etc) within the preceding 10 years, had received any antiandrogen therapy or had undergone biopsy within the prior 2 years were excluded. Percent fPSA was available for 277 of the 350 patients. Prostate biopsy. Total prostate and transition zone volumes were determined by transrectal ultrasound using the prolate ellipsoid formula, [(maximal transverse diameter) ⫻ (anteroposterior diameter) ⫻ (longitudinal diameter in mid sagittal plane) ⫻ (/6)]. All patients received a Fleets enema and oral antibiotics prior to biopsy. Transrectal biopsies were performed under ultrasound guidance with an end fire transducer. Six medial cores (M6Cs) were obtained from the mid lobar parasagittal plane and 6 lateral cores were obtained just medial to the lateral edge of the prostate, as previously reported.10 Additional lesion directed cores were obtained at attending urologist discretion. Laboratory and pathological analyses. Biopsy cores were individually labeled and scored for Gleason grade and percent cancer. Lesion directed biopsy cores were excluded from these analyses. PSA and percent fPSA tests were performed using Hybritech Tandem-MP assays (Hybritech, Inc., San Diego, California). All PSA measurements were performed within 6 months of the biopsy date. No serum sample was collected within the first 6 weeks after biopsy. Statistical analyses. ROC curve analyses and t tests for equality of means were done with SPSS 10.0 (SPSS, Inc., Chicago, Illinois) and Medcalc (MedCalc Software, Mariakerke, Belgium) software. RESULTS
Patient characteristics. A cohort of 350 consecutive men with negative DRE, PSA between 4 and 10 ng/ml, no biopsy within the preceding 2 years and no history of prostate ablation or antiandrogen therapy underwent S12C biopsy at Scott Department of Urology between March 1999 and January 2003. Mean patient age, PSA and prostate volume were 63 years, 6.1 ng/ml and 63 cc, respectively. Percent fPSA was measured in 277 of the 350 men (79%). There was no statistically significant difference in mean age, PSA, PSA density (PSAD), PSA transition zone (TZ) density, prostate volume or transition zone volume between patients who did and did not undergo percent fPSA testing (table 1). The distribution of percent fPSA values in the study population was comparable to that reported in large multicenter studies with an overall positive sextant biopsy, and negative median sextant biopsy percent fPSA values of 14.9% (mean 16.1%), 11.3% (mean
14.0%) and 15.7% (mean 16.9%), respectively (figs. 1 and 2).1–3 Characteristics of cancers detected by each biopsy strategy. The cancer detection rate for all 350 patients with total PSA between 4 and 10 ng/ml was 30% (95% CI 25 to 35) and 41% (95% CI 36 to 46) for the M6C and S12C biopsy sets, respectively, a statistically significant difference (p ⬍0.05). A comparison of pathological and serological data on additional cancers detected on S12C biopsy (ie not detected by sextant biopsy) to all others (ie those detected on sextant biopsy irrespective of whether they also had cancer detected on laterally directed cores) revealed that cancers detected exclusively on laterally directed cores were smaller, and had lower Gleason scores and lower PSAD. Cancers detected exclusively on laterally directed cores were similar in pathological characteristics to cancers identified exclusively on medially directed cores (table 2). The addition of 6 laterally directed cores (L6Cs) resulted in the detection of an additional 38 cancers, of which 16 (42%) were clinically significant, as defined by biopsy findings of Gleason grade 4 or greater and tumor greater than 3 mm if only 1 core was positive.11 Percent fPSA distribution of cancers detected on lateral cores only. To compare percent fPSA values of the additional prostate cancer cases identified by S12C biopsy and missed by sextant biopsy we plotted the distribution of percent fPSA values in patients with positive biopsies in laterally directed cores only (fig. 2). In patients with cancer detected exclusively on laterally directed cores the distribution of percent fPSA was skewed toward higher percent fPSA. In fact, the percent fPSA distribution for cancers detected exclusively on laterally directed cores was almost identical to that in patients without cancer. The mean ⫾ SD and median percent fPSA of cancers detected exclusively with laterally directed cores (17.2 ⫾ 6.29 and 15.7) were similar (p ⫽ 0.774) to those in biopsy negative patients (16.9 ⫾ 7.24 and 15.7) and higher (p ⫽ 0.024) than those in patients with cancers detected by the traditional sextant biopsy scheme (14.0 ⫾ 7.64 and 11.3, respectively). As expected, the mean percent fPSA of cancers detected on M6C was lower than that in biopsy negative patients (p ⫽ 0.005). Percent fPSA performance by biopsy strategy. To determine the effects of an S12C biopsy strategy on the performance of percent fPSA as an adjunct to total PSA for prostate cancer screening we performed ROC analysis. Although it was not statistically significant, there was a trend toward a poorer overall percent fPSA performance with a S12C biopsy scheme, as demonstrated by a lower AUC than with sextant biopsy (0.60, 95% CI 0.54 to 0.67 vs 0.66, 95% CI 0.58 to 0.73). As expected for a relatively small study, the ROC curves overlapped at the extremes (fig. 3). To evaluate the effects of S12C biopsy on the sensitivity and specificity of percent fPSA we tabulated the sensitivity and specificity of percent fPSA for various biopsy threshold values (table 3). At each biopsy threshold value percent fPSA specificity was not affected by the biopsy strategy, whereas
TABLE 1. Patient clinical characteristics All Pts Mean No. pts Age 63 PSA (ng/ml) 6.1 % fPSA 16.0 PSAD (ng/ml/cc) 0.12 PSATZ vol (ng/ml/cc) 0.32 Prostate vol (cc) 63 TZ vol (cc) 36 * Equal variances not assumed.
Median
Range
% fPSA Available ⫾ SD
% fPSA Not Available ⫾ SD
2-Tailed Significance (independent sample t test)*
63 5.8 14.9 0.11 0.20 55 29
34–83 4.0–10.0 0.8–45.0 0.02–0.44 0.03–8.4 18–318 1–223
277 63 ⫾ 8.3 6.1 ⫾ 1.6 16.0 ⫾ 7.2 0.12 ⫾ 0.07 0.34 ⫾ 0.8 61 ⫾ 34.8 35 ⫾ 26.8
73 63 ⫾ 7.9 5.9 ⫾ 1.6
0.88 0.34
0.12 ⫾ 0.07 0.26 ⫾ 0.2 67 ⫾ 41.4 39 ⫾ 32.8
0.29 0.15 0.28 0.31
902
12-CORE BIOPSY AND PERCENT FREE PROSTATE SPECIFIC ANTIGEN
FIG. 1. Percent fPSA distribution in our cohort was comparable to that reported in large, multicenter studies. Mean and median percent fPSA in study population was 16.1 ⫾ 7.24 and 14.9%, respectively.
FIG. 2. In patients with cancer detected exclusively on L6C percent fPSA distribution was almost identical to that in patients without cancer. Mean and median percent fPSA of cancers detected exclusively in L6C (17.2 ⫾ 6.29 and 15.7) was similar (p ⫽ 0.774) to that in biopsy negative cases (16.9 ⫾ 7.24 and 15.7) and higher (p ⫽ 0.024) than that of cancers detected by M6C (14.0 ⫾ 7.64 and 11.3, respectively). Mean percent fPSA of cancers detected on M6C was lower than that of biopsy negative cases (p ⫽ 0.005).
percent fPSA sensitivity was consistently lower when the S12C strategy was used. DISCUSSION
Disease screening programs commonly rely on the widespread use of a relatively inexpensive, noninvasive test to identify persons at high risk for a given disease. These screening tests show high sensitivity, usually at the expense of specificity. Persons with positive screening tests are then subjected to a confirmatory test that has higher specificity and yet is generally more costly and invasive. With biopsy strategies involving more than 10 cores recently becoming the standard of care in prostate cancer detection the sensitivity of the confirmatory test has increased and the characteristics of the prostate cancers identified have
changed.4, 5, 7–9 This prompted us to reevaluate the performance of percent fPSA when using S12C biopsy as opposed to sextant biopsy as a confirmatory test in prostate cancer screening. Sextant biopsy has been shown to detect only 70% of the cancers found with S12C biopsy.4, 5 Indeed, serial sextant biopsy studies show that up to 20% of men with PSA above 4 ng/ml have a positive repeat biopsy after a negative initial biopsy.6, 12 The positive biopsy rate for S12C biopsy approaches the cumulative positive biopsy rate reported in serial sextant biopsy studies.4 – 6, 12 The data presented in our study confirm that S12C biopsy detects at least 30% more cancers than sextant biopsy. In the group of patients evaluated in this study M6C biopsy detected cancer in 30% of the patients who underwent biopsy, while S12C biopsy detected cancer in 41% who underwent biopsy. Although it was not a strict, population based screening study, patients in this study had not recently undergone biopsy and they represented patients who typically might undergo percent fPSA testing to evaluate further their risk of prostate cancer prior to biopsy. All patients had negative DRE and PSA between 4 and 10 ng/ml. The pathological characteristics of the additional prostate cancers detected by S12C biopsy are quantitatively and qualitatively different from those of cancers detected simultaneously by sextant and S12C biopsy. The cancers identified exclusively by the additional lateral cores of a S12C biopsy are on average smaller and have lower Gleason scores than those identified on sextant plus lateral cores.9 Others have reported that, although S12 core biopsy identifies a greater number of significant cancers than sextant biopsy, the additional cancers identified are on average smaller and less aggressive.7–9 Our study confirms the observations of others that the additional cancers detected on S12C biopsy are smaller, have lower Gleason scores and lower PSAD, and higher percent fPSA than those detected with sextant biopsy. Nevertheless, as defined by commonly accepted criteria, 42% of the cancers detected exclusively on laterally directed cores were clinically significant compared to 100% of those detected on sextant biopsy and 87% of all cancers detected with S12C biopsy. In this study the addition of 6 laterally directed cores resulted in the detection of an additional 16 clinically significant prostate cancers. In our study, which is to our knowledge the first to examine carefully the effects of S12C biopsy on the performance of percent fPSA, 2 important findings emerged. 1) There is a nonstatistically significant trend toward lower sensitivity for each percent fPSA threshold value when the results of S12C biopsy are compared with those of sextant biopsy. 2) The specificity of percent fPSA value as a screening test remains the same for S12C and sextant biopsy for each threshold value. These 2 findings can be explained by the observation that the distribution of percent fPSA in patients with cancers
TABLE 2. Cancer characteristics by positive core location Mean M6C Pos Only ⫾ SD No. pts 32 Age 64 ⫾ 8.2 PSA (ng/ml) 6.1 ⫾ 1.7 % fPSA 17.8 ⫾ 8.0 PSAD (ng/ml/cc) 0.11 ⫾ 0.05 PSATZ vol (ng/ml/cc) 0.22 ⫾ 0.1 Prostate vol (cc) 67 ⫾ 40.9 TZ vol (cc) 40 ⫾ 31.5 Gleason sum 6.4 ⫾ 0.7 Max Gleason grade 3.3 ⫾ 0.5 Max mm Ca 2.5 ⫾ 2.3 Max % Ca 20.8 ⫾ 19.9 No. pos cores 1.25 ⫾ 0.57 * Equal variance not assumed. † Statistically significant to 95% level.
Mean L6C Pos Only ⫾ SD
2-Tailed Significance (independent sample t test)
Mean M6C Pos ⫾ SD
Mean L6C Pos Only ⫾ SD
2-Tailed Significance (independent sample t test)
38 65 ⫾ 7.8 6.1 ⫾ 1.5 17.2 ⫾ 6.3 0.12 ⫾ 0.05 0.26 ⫾ 0.2 59 ⫾ 23.8 32 ⫾ 17.6 6.2 ⫾ 0.6 3.2 ⫾ 0.4 3.7 ⫾ 3.4 29.7 ⫾ 24.1 1.13 ⫾ 0.34
0.649 0.989 0.759 0.580 0.307 0.365 0.230 0.144 0.139 0.085 0.093 0.307
104 64 ⫾ 8.5 6.1 ⫾ 1.5 14.1† ⫾ 7.6 0.15† ⫾ 0.08 0.53† ⫾ 1.2 51 ⫾ 31.7 27 ⫾ 24.9 6.6† ⫾ 0.7 3.5† ⫾ 0.5 5.3† ⫾ 3.7 45.1† ⫾ 30.2 1.73 ⫾ 1.12†
38 65 6.1 17.2† 0.12† 0.26† 59 32 6.2† 3.2† 3.6† 29.7† 1.13†
0.375 0.976 0.024† 0.004† 0.027† 0.096 0.190 0.001† 0.000† 0.018† 0.002† 0.000†
903
12-CORE BIOPSY AND PERCENT FREE PROSTATE SPECIFIC ANTIGEN
FIG. 3. ROC analysis of S12C biopsy vs M6C biopsy. Although it was not statistically significant, there was trend toward poorer overall performance of percent fPSA with S12C vs sextant biopsy (AUC 0.60, 95% CI 0.54 to 0.67 vs 0.66, 95% CI 0.58 to 0.73). As expected for relatively small study, ROC curves overlapped at extremes.
detected exclusively on L6Cs is similar to that in patients with a negative S12C biopsy (fig. 2). This means that additional cancers were found in a random group of patients who would have otherwise been deemed free of cancer by sextant biopsy. Therefore, the proportion of patients with a negative biopsy above (and below) any given threshold remains unchanged with the addition of 6 laterally directed cores. For this reason specificity at each threshold remained unchanged with the addition of 6 laterally directed cores despite the increased number of cancers detected. Our data demonstrate a small and yet consistent trend toward lower sensitivity at each threshold when comparing the performance of percent fPSA measured by S12C biopsy to the performance of percent fPSA measured by sextant biopsy. Because additional cancers detected exclusively on L6Cs had a distribution similar to that in
patients with negative biopsies and specificity at each biopsy threshold was unchanged when S12C was compared with sextant biopsy, the proportion of patients with additional positive biopsy results identified on S12C biopsy below any given threshold would be equal to the proportion of patients with negative biopsy results below that threshold or 1-specificity. Therefore, the number of S12C biopsy positive cases below any given threshold of percent fPSA can be defined by the formula, number positive below threshold S12C ⫽ [sensitivity (M6C) ⫻ total (M6C) positive] ⫹ (1 ⫺ specificity) ⫻ (total (M6C) positive) ⫻ 0.37, where 0.37 represents the fraction of additional cancers detected by S12C biopsy compared with sextant biopsy, sensitivity (M6C) represents the sensitivity at a given threshold for sextant biopsy and total (M6C) positive represents the number of cancers detected by sextant biopsy when all patients with PSA between 4 and 10 ng/ml are biopsied, irrespective of percent fPSA. Percent fPSA sensitivity is defined as the fraction of patients with positive biopsy below a given threshold. Given the mentioned formula, percent fPSA sensitivity with S12C biopsy at any given threshold can be predicted based on the known sensitivity and specificity of percent fPSA using sextant biopsy at that threshold and the fraction of additional cancers detected by S12C (0.37 in our series). For the formula, sensitivity(S12C) ⫽ [sensitivity(M6C) ⫹ (1 ⫺ specificity) ⫻ 0.37]/(1 ⫹ 0.37), the predicted sensitivity of percent fPSA when using S12C biopsy is in close agreement with empirically determined sensitivity (table 3). Because the difference in percent fPSA sensitivity whether S12C or sextant biopsy is used is small (between 1 and 5 percentage points), a study much larger than ours would be required to demonstrate with statistical significance that S12C biopsy decreases percent fPSA sensitivity. The main practical application of our findings is that 6 additional laterally directed cores affect only the sensitivity and not the specificity of percent fPSA at each biopsy threshold and the magnitude of this effect depends on the fraction of additional cancers detected by laterally directed cores, as indicated in the mentioned formula. When the fraction of additional cancers detected by S12C biopsy compared to sextant biopsy is in the 30% to 40% range, as in this and various other studies, the decrease in sensitivity for any given percent fPSA biopsy threshold is small.4, 5, 8
TABLE 3. Percent fPSA performance by biopsy strategy Sensitivity % fPSA Cutoff
M6C
S12C
Specificity* Predicted S12C†
M6C
% Pos Biopsy Probability S12C
M6C
S12C
16 or Less: % 70 65 65 47 47 36 47 No./total No. 57/82 74/114 91/195 76/163 57/159 74/159 95% CI 59–79 56–73 40–54 39–55 29–44 39–55 18 or Less: % 78 73 74 38 38 35 45 No./total No. 64/82 83/114 75/195 62/163 64/184 83/184 95% CI 68–86 64–80 31–45 31–46 28–42 38–52 20 or Less: % 84 81 81 29 29 34 45 No./total No. 69/82 92/114 56/195 47/163 69/205 92/205 95% CI 75–90 73–87 23–36 23–36 28–41 38–52 22 or Less: % 85 83 83 23 23 32 44 No./total No. 70/82 95/114 45/195 38/163 70/217 95/217 95% CI 76–91 75–89 18–29 17–30 26–38 38–51 25 or Less: % 90 89 89 13 13 30 42 No./total No. 74/82 102/114 25/195 21/163 74/244 102/244 95% CI 82–95 82–94 9–18 9–19 25–36 36–48 * Unnecessary biopsies avoided. † Calculated using the formula, sensitivity (S12C) ⫽ [sensitivity (M6C) ⫹ (1 ⫺ specificity) ⫻ 0.37]/(1 ⫹ 0.37), where 0.37 is fractional additional cancers detected by S12C compared to M6C or sextant biopsy.
904
12-CORE BIOPSY AND PERCENT FREE PROSTATE SPECIFIC ANTIGEN CONCLUSIONS
S12C biopsy detects more prostate cancers than sextant biopsy in men with benign DRE and PSA between 4 and 10 ng/ml. Prostate cancers detected exclusively on the medial or lateral 6 cores of S12C biopsy are smaller, and have lower PSAD, higher percent fPSA and lower Gleason sum than those detected on at least 1 medial and 1 laterally directed core. Nevertheless, the addition of 6 laterally directed cores identifies additional significant cancers that would have been missed using a sextant biopsy technique. Cancers identified exclusively in the additional laterally directed cores of S12C biopsy have a percent fPSA distribution similar to that in patients with negative biopsy. As a result, the specificity of percent fPSA remains the same for any given biopsy threshold value when S12C and sextant biopsy results are compared. In addition, there is a nonstatistically significant, and yet consistent and predictable trend toward lower sensitivity for each biopsy threshold value of percent fPSA when S12C rather than sextant biopsy is used as the validating test. REFERENCES
1. Gann, P. H., Ma, J., Catalona, W. J. and Stampfer, M. J.: Strategies combining total and percent free prostate specific antigen for detecting prostate cancer: a prospective evaluation. J Urol, 167: 2427, 2002 2. Catalona, W. J., Partin, A. W., Slawin, K. M., Brawer, M. K., Flanigan, R. C., Patel, A. et al: Use of the percentage of free prostate-specific antigen to enhance differentiation of prostate cancer from benign prostatic disease: a prospective multicenter clinical trial. JAMA, 279: 1542, 1998 3. Catalona, W. J., Smith, D. S., Wolfert, R. L., Wang, T. J., Rittenhouse, H. G., Ratliff, T. L. et al: Evaluation of percentage of free serum prostate-specific antigen to improve speci-
ficity of prostate cancer screening. JAMA, 274: 1214, 1995 4. Gore, J. L., Shariat, S. F., Miles, B. J., Kadmon, D., Jiang, N., Wheeler, T. M. et al: Optimal combinations of systematic sextant and laterally directed biopsies for the detection of prostate cancer. J Urol, 165: 1554, 2001 5. Presti, J. C., Jr., O’Dowd, G. J., Miller, M. C., Mattu, R. and Veltri, R. W.: Extended peripheral zone biopsy schemes increase cancer detection rates and minimize variance in prostate specific antigen and age related cancer rates: results of a community multi-practice study. J Urol, 169: 125, 2003 6. Roehl, K. A., Antenor, J. A. V. and Catalona, W. J.: Serial biopsy results in prostate cancer screening study. J Urol, 167: 2435, 2002 7. Chan, T. Y., Chan, D. Y., Lecksell, K., Stutzman, R. E. and Epstein, J. I.: Does increased needle biopsy sampling of the prostate detect a higher number of potentially insignificant tumors? J Urol, 166: 2181, 2001 8. Taylor, J. A., 3rd, Gancarczyk, K. J., Fant, G. V. and McLeod, D. G.: Increasing the number of core samples taken at prostate needle biopsy enhances the detection of clinically significant prostate cancer. Urology, 60: 841, 2002 9. Singh, H., Canto, E. I., Shariat, S. F., Kadmon, D., Miles, B. J., Wheeler, T. M. et al: Improved detection of clinically significant, curable prostate cancer with systematic 12-core biopsy. J Urol, 171: 1089, 2003 10. Singh, H., Canto, E. I., Shariat, S. F., Kadmon, D., Miles, B. J., Wheeler, T. M. et al: Six additional systematic lateral cores enhance sextant biopsy prediction of pathological features at radical prostatectomy. J Urol, 171: 204, 2003 11. Epstein, J. I., Walsh, P. C., Carmichael, M. and Brendler, C. B.: Pathologic and clinical findings to predict tumor extent of nonpalpable (stage T1c) prostate cancer. JAMA, 271: 368, 1994 12. Roehrborn, C. G., Pickens, G. J. and Sanders, J. S.: Diagnostic yield of repeated transrectal ultrasound-guided biopsies stratified by specific histopathologic diagnoses and prostate specific antigen levels. Urology, 47: 347, 1996
Vol. 172, 900 –904, September 2004 Printed in U.S.A.
DOI: 10.1097/01.ju.0000134619.72675.8d
EFFECTS OF SYSTEMATIC 12-CORE BIOPSY ON THE PERFORMANCE OF PERCENT FREE PROSTATE SPECIFIC ANTIGEN FOR PROSTATE CANCER DETECTION EDUARDO I. CANTO, HERB SINGH, SHAHROKH F. SHARIAT, DOV KADMON, BRIAN J. MILES, THOMAS M. WHEELER* AND KEVIN M. SLAWIN†,‡ From the Baylor Prostate Center, Scott Department of Urology (EIC, HS, DK, BJM, KMS) and Department of Pathology (TMW), Baylor College of Medicine and Methodist Hospital (TMW), Houston and Department of Urology, University of Texas Southwestern (SFS), Dallas, Texas
ABSTRACT
Purpose: The performance characteristics of percent free (f) prostate specific antigen (PSA) for differentiating between benign prostatic hyperplasia and prostate cancer were originally established using primarily sextant biopsy. We determined whether the addition of 6 laterally directed cores to the traditional sextant prostate biopsy affects the performance of percent fPSA. Materials and Methods: We retrospectively evaluated a cohort of 350 consecutive biopsies in men with negative digital rectal examinations and PSA between 4 and 10 ng/ml who underwent systematic 12 core biopsy (S12C) biopsy at Scott Department of Urology between March 1999 and January 2003. The effects of 6 additional, laterally directed biopsies on the sensitivity, specificity and area under the ROC curve for percent fPSA was evaluated in the 277 men in whom percent fPSA was measured. Results: Cancers detected exclusively in the 6 laterally directed cores were associated with percent fPSA values similar to those in patients with a benign S12C biopsy. This resulted in a modest and yet predictable decrease in the sensitivity of percent fPSA at each biopsy threshold value without affecting specificity. There was a nonstatistically significant decrease in the area under the ROC curve with the addition of 6 laterally directed cores to sextant biopsy (medial sextant cores 0.66 vs S12C 0.60). Conclusions: The 12 core biopsy strategies have a higher cancer detection rate than sextant biopsies and they are gaining widespread acceptance. The addition of 6 laterally directed cores to traditional sextant biopsy may result in a modest decrease in the sensitivity of percent fPSA at each selected biopsy threshold without affecting specificity. KEY WORDS: prostate, biopsy, prostatic neoplasms, prostate-specific antigen
The fraction of total prostate specific antigen (PSA) comprised of free (f) PSA expressed as a percent is Food and Drug Administration approved for use in the enhancement of PSA for differentiating prostate cancer from benign prostatic hyperplasia when PSA is in the range of 4 to 10 ng/ml. The area under the ROC curve (AUC) of percent fPSA as a screening tool has been reported to be about 0.70 to 0.74.1, 2 Large multicenter studies have shown that the use of percent fPSA can decrease the number of unnecessary biopsies by about 20%, while maintaining 95% sensitivity.1–3 However, the validated biopsy threshold for percent fPSA is based on cancer detection using primarily sextant biopsy.2 Systematic 12 core (S12C) biopsy that includes the standard sextant biopsy cores and an additional 6 laterally directed biopsy cores has a significantly higher rate of prostate cancer detection than traditional sextant biopsy.4, 5 We reported that the addition of 6 laterally directed cores to traditional sextant biopsy increases the prostate cancer detection rate from 31% to 43%,4 a rate that approaches the overall cancer detection rate seen in serial sextant biopsy studies.6 Recent reports suggest that S12C biopsy detects the Accepted for publication April 23, 2004. * Financial interest and/or other relationship with Pintex. † Correspondence: Scott Department of Urology, Baylor College of Medicine, Baylor Prostate Center, 6560 Fannin St., Suite 2100, Houston, Texas 77030 (telephone: 713-798-8670; FAX: 713-798-8030; e-mail: [email protected]). ‡ Financial interest and/or other relationship with GlaxoSmithKline and National Institutes of Health.
highest number of significant and insignificant cancers.7, 8 In an analysis of 176 patients who underwent S12C biopsy followed by radical retropubic prostatectomies at Baylor College of Medicine and Methodist Hospital, Houston, Texas 15 of the 86 cancers (17%) with tumor volume greater than 0.5 cc were detected in the laterally directed cores only and, therefore, they would have been missed with a traditional sextant biopsy strategy.9 As more men undergo yearly PSA based prostate cancer screening, a larger proportion of those who are candidates for prostate biopsy have PSA in the 4 to 10 ng/ml range. Many of these men undergo percent fPSA testing as an adjunct to further stratify the risk of prostate cancer. Although S12C biopsy detects a greater number of clinically significant cancers, the cancers detected exclusively with sextant or laterally directed cores of S12C biopsy are on average smaller and the associated serum percent fPSA is higher than that of those detected with sextant plus laterally directed cores.9 Since many urologists now routinely perform at least 10 biopsy cores during prostate biopsy, it is important to understand how the differences in the nature of cancers detected through the additional cores may affect percent fPSA performance. Because the Food and Drug Administration approved percent fPSA biopsy threshold value was established using sextant biopsy and additional cancers detected by the lateral cores of S12C biopsy have lower associated serum percent fPSA values,9 we hypothesized that the percent fPSA thresh-
900
901
12-CORE BIOPSY AND PERCENT FREE PROSTATE SPECIFIC ANTIGEN
old for S12C biopsy may need to be lowered to maintain the same sensitivity as that based on sextant biopsy. MATERIALS AND METHODS
Patient selection. We retrospectively evaluated a cohort of 350 men with a negative digital rectal examination (DRE) and PSA between 4 and 10 ng/ml who underwent screening S12C biopsy at the Scott Department of Urology between March 1999 and January 2003. Subjects who had undergone an ablative procedure (ie transurethral resection, transurethral microwave ablation, etc) within the preceding 10 years, had received any antiandrogen therapy or had undergone biopsy within the prior 2 years were excluded. Percent fPSA was available for 277 of the 350 patients. Prostate biopsy. Total prostate and transition zone volumes were determined by transrectal ultrasound using the prolate ellipsoid formula, [(maximal transverse diameter) ⫻ (anteroposterior diameter) ⫻ (longitudinal diameter in mid sagittal plane) ⫻ (/6)]. All patients received a Fleets enema and oral antibiotics prior to biopsy. Transrectal biopsies were performed under ultrasound guidance with an end fire transducer. Six medial cores (M6Cs) were obtained from the mid lobar parasagittal plane and 6 lateral cores were obtained just medial to the lateral edge of the prostate, as previously reported.10 Additional lesion directed cores were obtained at attending urologist discretion. Laboratory and pathological analyses. Biopsy cores were individually labeled and scored for Gleason grade and percent cancer. Lesion directed biopsy cores were excluded from these analyses. PSA and percent fPSA tests were performed using Hybritech Tandem-MP assays (Hybritech, Inc., San Diego, California). All PSA measurements were performed within 6 months of the biopsy date. No serum sample was collected within the first 6 weeks after biopsy. Statistical analyses. ROC curve analyses and t tests for equality of means were done with SPSS 10.0 (SPSS, Inc., Chicago, Illinois) and Medcalc (MedCalc Software, Mariakerke, Belgium) software. RESULTS
Patient characteristics. A cohort of 350 consecutive men with negative DRE, PSA between 4 and 10 ng/ml, no biopsy within the preceding 2 years and no history of prostate ablation or antiandrogen therapy underwent S12C biopsy at Scott Department of Urology between March 1999 and January 2003. Mean patient age, PSA and prostate volume were 63 years, 6.1 ng/ml and 63 cc, respectively. Percent fPSA was measured in 277 of the 350 men (79%). There was no statistically significant difference in mean age, PSA, PSA density (PSAD), PSA transition zone (TZ) density, prostate volume or transition zone volume between patients who did and did not undergo percent fPSA testing (table 1). The distribution of percent fPSA values in the study population was comparable to that reported in large multicenter studies with an overall positive sextant biopsy, and negative median sextant biopsy percent fPSA values of 14.9% (mean 16.1%), 11.3% (mean
14.0%) and 15.7% (mean 16.9%), respectively (figs. 1 and 2).1–3 Characteristics of cancers detected by each biopsy strategy. The cancer detection rate for all 350 patients with total PSA between 4 and 10 ng/ml was 30% (95% CI 25 to 35) and 41% (95% CI 36 to 46) for the M6C and S12C biopsy sets, respectively, a statistically significant difference (p ⬍0.05). A comparison of pathological and serological data on additional cancers detected on S12C biopsy (ie not detected by sextant biopsy) to all others (ie those detected on sextant biopsy irrespective of whether they also had cancer detected on laterally directed cores) revealed that cancers detected exclusively on laterally directed cores were smaller, and had lower Gleason scores and lower PSAD. Cancers detected exclusively on laterally directed cores were similar in pathological characteristics to cancers identified exclusively on medially directed cores (table 2). The addition of 6 laterally directed cores (L6Cs) resulted in the detection of an additional 38 cancers, of which 16 (42%) were clinically significant, as defined by biopsy findings of Gleason grade 4 or greater and tumor greater than 3 mm if only 1 core was positive.11 Percent fPSA distribution of cancers detected on lateral cores only. To compare percent fPSA values of the additional prostate cancer cases identified by S12C biopsy and missed by sextant biopsy we plotted the distribution of percent fPSA values in patients with positive biopsies in laterally directed cores only (fig. 2). In patients with cancer detected exclusively on laterally directed cores the distribution of percent fPSA was skewed toward higher percent fPSA. In fact, the percent fPSA distribution for cancers detected exclusively on laterally directed cores was almost identical to that in patients without cancer. The mean ⫾ SD and median percent fPSA of cancers detected exclusively with laterally directed cores (17.2 ⫾ 6.29 and 15.7) were similar (p ⫽ 0.774) to those in biopsy negative patients (16.9 ⫾ 7.24 and 15.7) and higher (p ⫽ 0.024) than those in patients with cancers detected by the traditional sextant biopsy scheme (14.0 ⫾ 7.64 and 11.3, respectively). As expected, the mean percent fPSA of cancers detected on M6C was lower than that in biopsy negative patients (p ⫽ 0.005). Percent fPSA performance by biopsy strategy. To determine the effects of an S12C biopsy strategy on the performance of percent fPSA as an adjunct to total PSA for prostate cancer screening we performed ROC analysis. Although it was not statistically significant, there was a trend toward a poorer overall percent fPSA performance with a S12C biopsy scheme, as demonstrated by a lower AUC than with sextant biopsy (0.60, 95% CI 0.54 to 0.67 vs 0.66, 95% CI 0.58 to 0.73). As expected for a relatively small study, the ROC curves overlapped at the extremes (fig. 3). To evaluate the effects of S12C biopsy on the sensitivity and specificity of percent fPSA we tabulated the sensitivity and specificity of percent fPSA for various biopsy threshold values (table 3). At each biopsy threshold value percent fPSA specificity was not affected by the biopsy strategy, whereas
TABLE 1. Patient clinical characteristics All Pts Mean No. pts Age 63 PSA (ng/ml) 6.1 % fPSA 16.0 PSAD (ng/ml/cc) 0.12 PSATZ vol (ng/ml/cc) 0.32 Prostate vol (cc) 63 TZ vol (cc) 36 * Equal variances not assumed.
Median
Range
% fPSA Available ⫾ SD
% fPSA Not Available ⫾ SD
2-Tailed Significance (independent sample t test)*
63 5.8 14.9 0.11 0.20 55 29
34–83 4.0–10.0 0.8–45.0 0.02–0.44 0.03–8.4 18–318 1–223
277 63 ⫾ 8.3 6.1 ⫾ 1.6 16.0 ⫾ 7.2 0.12 ⫾ 0.07 0.34 ⫾ 0.8 61 ⫾ 34.8 35 ⫾ 26.8
73 63 ⫾ 7.9 5.9 ⫾ 1.6
0.88 0.34
0.12 ⫾ 0.07 0.26 ⫾ 0.2 67 ⫾ 41.4 39 ⫾ 32.8
0.29 0.15 0.28 0.31
902
12-CORE BIOPSY AND PERCENT FREE PROSTATE SPECIFIC ANTIGEN
FIG. 1. Percent fPSA distribution in our cohort was comparable to that reported in large, multicenter studies. Mean and median percent fPSA in study population was 16.1 ⫾ 7.24 and 14.9%, respectively.
FIG. 2. In patients with cancer detected exclusively on L6C percent fPSA distribution was almost identical to that in patients without cancer. Mean and median percent fPSA of cancers detected exclusively in L6C (17.2 ⫾ 6.29 and 15.7) was similar (p ⫽ 0.774) to that in biopsy negative cases (16.9 ⫾ 7.24 and 15.7) and higher (p ⫽ 0.024) than that of cancers detected by M6C (14.0 ⫾ 7.64 and 11.3, respectively). Mean percent fPSA of cancers detected on M6C was lower than that of biopsy negative cases (p ⫽ 0.005).
percent fPSA sensitivity was consistently lower when the S12C strategy was used. DISCUSSION
Disease screening programs commonly rely on the widespread use of a relatively inexpensive, noninvasive test to identify persons at high risk for a given disease. These screening tests show high sensitivity, usually at the expense of specificity. Persons with positive screening tests are then subjected to a confirmatory test that has higher specificity and yet is generally more costly and invasive. With biopsy strategies involving more than 10 cores recently becoming the standard of care in prostate cancer detection the sensitivity of the confirmatory test has increased and the characteristics of the prostate cancers identified have
changed.4, 5, 7–9 This prompted us to reevaluate the performance of percent fPSA when using S12C biopsy as opposed to sextant biopsy as a confirmatory test in prostate cancer screening. Sextant biopsy has been shown to detect only 70% of the cancers found with S12C biopsy.4, 5 Indeed, serial sextant biopsy studies show that up to 20% of men with PSA above 4 ng/ml have a positive repeat biopsy after a negative initial biopsy.6, 12 The positive biopsy rate for S12C biopsy approaches the cumulative positive biopsy rate reported in serial sextant biopsy studies.4 – 6, 12 The data presented in our study confirm that S12C biopsy detects at least 30% more cancers than sextant biopsy. In the group of patients evaluated in this study M6C biopsy detected cancer in 30% of the patients who underwent biopsy, while S12C biopsy detected cancer in 41% who underwent biopsy. Although it was not a strict, population based screening study, patients in this study had not recently undergone biopsy and they represented patients who typically might undergo percent fPSA testing to evaluate further their risk of prostate cancer prior to biopsy. All patients had negative DRE and PSA between 4 and 10 ng/ml. The pathological characteristics of the additional prostate cancers detected by S12C biopsy are quantitatively and qualitatively different from those of cancers detected simultaneously by sextant and S12C biopsy. The cancers identified exclusively by the additional lateral cores of a S12C biopsy are on average smaller and have lower Gleason scores than those identified on sextant plus lateral cores.9 Others have reported that, although S12 core biopsy identifies a greater number of significant cancers than sextant biopsy, the additional cancers identified are on average smaller and less aggressive.7–9 Our study confirms the observations of others that the additional cancers detected on S12C biopsy are smaller, have lower Gleason scores and lower PSAD, and higher percent fPSA than those detected with sextant biopsy. Nevertheless, as defined by commonly accepted criteria, 42% of the cancers detected exclusively on laterally directed cores were clinically significant compared to 100% of those detected on sextant biopsy and 87% of all cancers detected with S12C biopsy. In this study the addition of 6 laterally directed cores resulted in the detection of an additional 16 clinically significant prostate cancers. In our study, which is to our knowledge the first to examine carefully the effects of S12C biopsy on the performance of percent fPSA, 2 important findings emerged. 1) There is a nonstatistically significant trend toward lower sensitivity for each percent fPSA threshold value when the results of S12C biopsy are compared with those of sextant biopsy. 2) The specificity of percent fPSA value as a screening test remains the same for S12C and sextant biopsy for each threshold value. These 2 findings can be explained by the observation that the distribution of percent fPSA in patients with cancers
TABLE 2. Cancer characteristics by positive core location Mean M6C Pos Only ⫾ SD No. pts 32 Age 64 ⫾ 8.2 PSA (ng/ml) 6.1 ⫾ 1.7 % fPSA 17.8 ⫾ 8.0 PSAD (ng/ml/cc) 0.11 ⫾ 0.05 PSATZ vol (ng/ml/cc) 0.22 ⫾ 0.1 Prostate vol (cc) 67 ⫾ 40.9 TZ vol (cc) 40 ⫾ 31.5 Gleason sum 6.4 ⫾ 0.7 Max Gleason grade 3.3 ⫾ 0.5 Max mm Ca 2.5 ⫾ 2.3 Max % Ca 20.8 ⫾ 19.9 No. pos cores 1.25 ⫾ 0.57 * Equal variance not assumed. † Statistically significant to 95% level.
Mean L6C Pos Only ⫾ SD
2-Tailed Significance (independent sample t test)
Mean M6C Pos ⫾ SD
Mean L6C Pos Only ⫾ SD
2-Tailed Significance (independent sample t test)
38 65 ⫾ 7.8 6.1 ⫾ 1.5 17.2 ⫾ 6.3 0.12 ⫾ 0.05 0.26 ⫾ 0.2 59 ⫾ 23.8 32 ⫾ 17.6 6.2 ⫾ 0.6 3.2 ⫾ 0.4 3.7 ⫾ 3.4 29.7 ⫾ 24.1 1.13 ⫾ 0.34
0.649 0.989 0.759 0.580 0.307 0.365 0.230 0.144 0.139 0.085 0.093 0.307
104 64 ⫾ 8.5 6.1 ⫾ 1.5 14.1† ⫾ 7.6 0.15† ⫾ 0.08 0.53† ⫾ 1.2 51 ⫾ 31.7 27 ⫾ 24.9 6.6† ⫾ 0.7 3.5† ⫾ 0.5 5.3† ⫾ 3.7 45.1† ⫾ 30.2 1.73 ⫾ 1.12†
38 65 6.1 17.2† 0.12† 0.26† 59 32 6.2† 3.2† 3.6† 29.7† 1.13†
0.375 0.976 0.024† 0.004† 0.027† 0.096 0.190 0.001† 0.000† 0.018† 0.002† 0.000†
903
12-CORE BIOPSY AND PERCENT FREE PROSTATE SPECIFIC ANTIGEN
FIG. 3. ROC analysis of S12C biopsy vs M6C biopsy. Although it was not statistically significant, there was trend toward poorer overall performance of percent fPSA with S12C vs sextant biopsy (AUC 0.60, 95% CI 0.54 to 0.67 vs 0.66, 95% CI 0.58 to 0.73). As expected for relatively small study, ROC curves overlapped at extremes.
detected exclusively on L6Cs is similar to that in patients with a negative S12C biopsy (fig. 2). This means that additional cancers were found in a random group of patients who would have otherwise been deemed free of cancer by sextant biopsy. Therefore, the proportion of patients with a negative biopsy above (and below) any given threshold remains unchanged with the addition of 6 laterally directed cores. For this reason specificity at each threshold remained unchanged with the addition of 6 laterally directed cores despite the increased number of cancers detected. Our data demonstrate a small and yet consistent trend toward lower sensitivity at each threshold when comparing the performance of percent fPSA measured by S12C biopsy to the performance of percent fPSA measured by sextant biopsy. Because additional cancers detected exclusively on L6Cs had a distribution similar to that in
patients with negative biopsies and specificity at each biopsy threshold was unchanged when S12C was compared with sextant biopsy, the proportion of patients with additional positive biopsy results identified on S12C biopsy below any given threshold would be equal to the proportion of patients with negative biopsy results below that threshold or 1-specificity. Therefore, the number of S12C biopsy positive cases below any given threshold of percent fPSA can be defined by the formula, number positive below threshold S12C ⫽ [sensitivity (M6C) ⫻ total (M6C) positive] ⫹ (1 ⫺ specificity) ⫻ (total (M6C) positive) ⫻ 0.37, where 0.37 represents the fraction of additional cancers detected by S12C biopsy compared with sextant biopsy, sensitivity (M6C) represents the sensitivity at a given threshold for sextant biopsy and total (M6C) positive represents the number of cancers detected by sextant biopsy when all patients with PSA between 4 and 10 ng/ml are biopsied, irrespective of percent fPSA. Percent fPSA sensitivity is defined as the fraction of patients with positive biopsy below a given threshold. Given the mentioned formula, percent fPSA sensitivity with S12C biopsy at any given threshold can be predicted based on the known sensitivity and specificity of percent fPSA using sextant biopsy at that threshold and the fraction of additional cancers detected by S12C (0.37 in our series). For the formula, sensitivity(S12C) ⫽ [sensitivity(M6C) ⫹ (1 ⫺ specificity) ⫻ 0.37]/(1 ⫹ 0.37), the predicted sensitivity of percent fPSA when using S12C biopsy is in close agreement with empirically determined sensitivity (table 3). Because the difference in percent fPSA sensitivity whether S12C or sextant biopsy is used is small (between 1 and 5 percentage points), a study much larger than ours would be required to demonstrate with statistical significance that S12C biopsy decreases percent fPSA sensitivity. The main practical application of our findings is that 6 additional laterally directed cores affect only the sensitivity and not the specificity of percent fPSA at each biopsy threshold and the magnitude of this effect depends on the fraction of additional cancers detected by laterally directed cores, as indicated in the mentioned formula. When the fraction of additional cancers detected by S12C biopsy compared to sextant biopsy is in the 30% to 40% range, as in this and various other studies, the decrease in sensitivity for any given percent fPSA biopsy threshold is small.4, 5, 8
TABLE 3. Percent fPSA performance by biopsy strategy Sensitivity % fPSA Cutoff
M6C
S12C
Specificity* Predicted S12C†
M6C
% Pos Biopsy Probability S12C
M6C
S12C
16 or Less: % 70 65 65 47 47 36 47 No./total No. 57/82 74/114 91/195 76/163 57/159 74/159 95% CI 59–79 56–73 40–54 39–55 29–44 39–55 18 or Less: % 78 73 74 38 38 35 45 No./total No. 64/82 83/114 75/195 62/163 64/184 83/184 95% CI 68–86 64–80 31–45 31–46 28–42 38–52 20 or Less: % 84 81 81 29 29 34 45 No./total No. 69/82 92/114 56/195 47/163 69/205 92/205 95% CI 75–90 73–87 23–36 23–36 28–41 38–52 22 or Less: % 85 83 83 23 23 32 44 No./total No. 70/82 95/114 45/195 38/163 70/217 95/217 95% CI 76–91 75–89 18–29 17–30 26–38 38–51 25 or Less: % 90 89 89 13 13 30 42 No./total No. 74/82 102/114 25/195 21/163 74/244 102/244 95% CI 82–95 82–94 9–18 9–19 25–36 36–48 * Unnecessary biopsies avoided. † Calculated using the formula, sensitivity (S12C) ⫽ [sensitivity (M6C) ⫹ (1 ⫺ specificity) ⫻ 0.37]/(1 ⫹ 0.37), where 0.37 is fractional additional cancers detected by S12C compared to M6C or sextant biopsy.
904
12-CORE BIOPSY AND PERCENT FREE PROSTATE SPECIFIC ANTIGEN CONCLUSIONS
S12C biopsy detects more prostate cancers than sextant biopsy in men with benign DRE and PSA between 4 and 10 ng/ml. Prostate cancers detected exclusively on the medial or lateral 6 cores of S12C biopsy are smaller, and have lower PSAD, higher percent fPSA and lower Gleason sum than those detected on at least 1 medial and 1 laterally directed core. Nevertheless, the addition of 6 laterally directed cores identifies additional significant cancers that would have been missed using a sextant biopsy technique. Cancers identified exclusively in the additional laterally directed cores of S12C biopsy have a percent fPSA distribution similar to that in patients with negative biopsy. As a result, the specificity of percent fPSA remains the same for any given biopsy threshold value when S12C and sextant biopsy results are compared. In addition, there is a nonstatistically significant, and yet consistent and predictable trend toward lower sensitivity for each biopsy threshold value of percent fPSA when S12C rather than sextant biopsy is used as the validating test. REFERENCES
1. Gann, P. H., Ma, J., Catalona, W. J. and Stampfer, M. J.: Strategies combining total and percent free prostate specific antigen for detecting prostate cancer: a prospective evaluation. J Urol, 167: 2427, 2002 2. Catalona, W. J., Partin, A. W., Slawin, K. M., Brawer, M. K., Flanigan, R. C., Patel, A. et al: Use of the percentage of free prostate-specific antigen to enhance differentiation of prostate cancer from benign prostatic disease: a prospective multicenter clinical trial. JAMA, 279: 1542, 1998 3. Catalona, W. J., Smith, D. S., Wolfert, R. L., Wang, T. J., Rittenhouse, H. G., Ratliff, T. L. et al: Evaluation of percentage of free serum prostate-specific antigen to improve speci-
ficity of prostate cancer screening. JAMA, 274: 1214, 1995 4. Gore, J. L., Shariat, S. F., Miles, B. J., Kadmon, D., Jiang, N., Wheeler, T. M. et al: Optimal combinations of systematic sextant and laterally directed biopsies for the detection of prostate cancer. J Urol, 165: 1554, 2001 5. Presti, J. C., Jr., O’Dowd, G. J., Miller, M. C., Mattu, R. and Veltri, R. W.: Extended peripheral zone biopsy schemes increase cancer detection rates and minimize variance in prostate specific antigen and age related cancer rates: results of a community multi-practice study. J Urol, 169: 125, 2003 6. Roehl, K. A., Antenor, J. A. V. and Catalona, W. J.: Serial biopsy results in prostate cancer screening study. J Urol, 167: 2435, 2002 7. Chan, T. Y., Chan, D. Y., Lecksell, K., Stutzman, R. E. and Epstein, J. I.: Does increased needle biopsy sampling of the prostate detect a higher number of potentially insignificant tumors? J Urol, 166: 2181, 2001 8. Taylor, J. A., 3rd, Gancarczyk, K. J., Fant, G. V. and McLeod, D. G.: Increasing the number of core samples taken at prostate needle biopsy enhances the detection of clinically significant prostate cancer. Urology, 60: 841, 2002 9. Singh, H., Canto, E. I., Shariat, S. F., Kadmon, D., Miles, B. J., Wheeler, T. M. et al: Improved detection of clinically significant, curable prostate cancer with systematic 12-core biopsy. J Urol, 171: 1089, 2003 10. Singh, H., Canto, E. I., Shariat, S. F., Kadmon, D., Miles, B. J., Wheeler, T. M. et al: Six additional systematic lateral cores enhance sextant biopsy prediction of pathological features at radical prostatectomy. J Urol, 171: 204, 2003 11. Epstein, J. I., Walsh, P. C., Carmichael, M. and Brendler, C. B.: Pathologic and clinical findings to predict tumor extent of nonpalpable (stage T1c) prostate cancer. JAMA, 271: 368, 1994 12. Roehrborn, C. G., Pickens, G. J. and Sanders, J. S.: Diagnostic yield of repeated transrectal ultrasound-guided biopsies stratified by specific histopathologic diagnoses and prostate specific antigen levels. Urology, 47: 347, 1996