- Email: [email protected]
The Relationship of Prostate Gland Volume to Extended Needle Biopsy on Prostate Cancer Detection
0022-5347/03/1691-0130/0 THE JOURNAL OF UROLOGY® Copyright © 2003 by AMERICAN UROLOGICAL ASSOCIATION
Vol. 169, 130 –135, January 2003 Printed in U.S.A.
DOI: 10.1097/01.ju.0000034153.49106.b7
THE RELATIONSHIP OF PROSTATE GLAND VOLUME TO EXTENDED NEEDLE BIOPSY ON PROSTATE CANCER DETECTION JEAN O. UNG, IGNACIO F. SAN FRANCISCO, MEREDITH M. REGAN, WILLIAM C. DEWOLF ARIA F. OLUMI
AND
From the Division of Urologic Surgery and Biometrics Center, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts
ABSTRACT
Purpose: We investigated the relationship between prostate volume and cancer detection by needle biopsy, and determined the effect of an increased number of cores on the sampling error of needle biopsy on large prostate glands. Materials and Methods: The study cohort included 750 consecutive patients who underwent first time transrectal ultrasound guided prostate needle biopsy from January 1995 to August 2001. Prostate volumes were divided into quartiles (13 to 34, 34.1 to 45, 45.1 to 64 and 64.1 to 244 cc). Multivariate analysis controlling for age, prostate specific antigen (PSA) and biopsy indication was performed to determine the effect of the number of cores and prostate volume on prostate cancer detection. Results: Patients diagnosed with prostate cancer were older (p ⫽ 0.0035) and had higher PSA levels (p ⫽ 0.0002) than those with no cancer on biopsy. Decreasing cancer detection rates were seen with increasing prostate volume (p ⫽ 0.0074). The OR of detection for each additional core was 0.99 (95% CI 0.93, 1.06), suggesting that increasing the number of biopsy cores did not increase the rate of prostate cancer detection. Multivariate analysis revealed that patients with larger prostates had the same, or possibly lower, cancer detection rate as the number of biopsy cores was increased. Patients with larger prostates were older (p ⬍0.0001), had higher PSA levels (p ⬍0.0001) and were even more likely to have undergone biopsy for increased PSA rather than abnormal digital rectal examination alone (p ⬍0.0001). Conclusions: Our study suggests that the lower cancer detection rate for men with large prostates may be due to a decrease in the use of increased serum PSA for prostate cancer detection in larger prostates in addition to other factors such as sampling error. Increased serum PSA levels in cases of larger prostates, although a risk factor for prostate cancer warranting biopsy, may also be due to nonmalignant sources such as benign prostatic hyperplasia. KEY WORDS: prostatic neoplasms, prostate-specific antigen, prostatic hyperplasia, biopsy
The early diagnosis of prostate cancer has evolved with the widespread use of serum prostate specific antigen (PSA) testing and transrectal ultrasound in addition to digital rectal examination. Transrectal ultrasound guided prostate needle biopsies have become the method of choice for the detection of prostate cancer. The sextant biopsy method, first proposed by Hodge et al in 1989, is still the standard strategy of systematic prostate biopsy.1 However, the optimal number of biopsies needed to identify all patients with prostate cancer at the earliest stage possible for optimal treatment, outcome and survival is not known. Stricker et al described a mathematical model based on Bayes’ theorem of conditional probability which demonstrated a correlation between the percent volume of prostate cancer within a prostate gland and the probability of a positive biopsy.2 The probability of detecting prostate cancer for a fixed percent cancer volume increases as the number of biopsies increases. However, this model assumes a uniform distribution of prostate carcinoma throughout the prostate gland and a random distribution of biopsies. Most data support a modified distribution of prostate carcinoma when the majority of tumors occur in the peripheral zone, as the transition zone is better known as the site of origin of benign
prostatic hyperplasia (BPH),3 and most biopsies are taken accordingly. Eskew et al reported that a systematic 5-region prostate biopsy of 13 to 18 cores is superior to the sextant method for diagnosing prostate cancer, with 35% more cases of prostate cancer detected.4 Several other authors have shown that the yield of sextant biopsy decreases with increasing prostate volume.5– 8 They have concluded that this finding is due to a sampling error that occurs in men with larger prostate glands. With no evidence to suggest differing intrinsic prostate cancer rates based on prostate volume alone,5, 9 additional biopsies could be necessary for larger prostate glands to detect prostate cancer. We further investigate the relationship between prostate gland volume and cancer detection by needle biopsy. The significance of sampling error with the sextant biopsy method has been noted. However, further study is necessary to evaluate the effect of extended needle biopsy on cancer detection rates with different prostate volumes. Vashi et al proposed a mathematical model for predicting the number of cores needed to detect clinically significant tumors at an early enough stage to provide the best chance of decreasing mortality.10 They recommend anywhere from 2 to 20 biopsies depending on prostate volume and patient age but their study is based on a mathematical model and lacks clinical data for correlation. Our study is based on clinical data from a single urologist (W. C. D.) and includes patients suspected
Accepted for publication May 24, 2002. Supported by the Hershey Family Prostate Cancer Foundation, Beth Israel Deaconess Medical Center. 130
PROSTATE CANCER DETECTION BY EXTENDED NEEDLE BIOPSY
131
FIG. 1. Distribution of prostate volumes by results of prostate needle biopsy. Median 45 cc, range 13 to 244 cc, total number 750 patients. Horizontal axis labels indicate upper bound of range (for example, less than 30 indicates 20 to 29.9).
to be at risk for clinically localized prostate cancer who subsequently underwent transrectal ultrasound guided prostate needle biopsy. METHODS
Patient selection. Between January 1995 and August 2001, 767 consecutive patients seen by a single urologist (W. C. D.) at our institution for suspicion of clinically localized prostate cancer underwent initial transrectal ultrasound guided prostate needle biopsy. Indications for biopsy included increased or increasing PSA and/or abnormal digital rectal examination. Transrectal ultrasound prostate volume was calculated using a computer generated elliptical estimation of 0.52 ⫻ length ⫻ width ⫻ height. The distribution of prostate volumes of patients undergoing prostate needle biopsy is illustrated in figure 1. Patients were excluded from study if prostate volume was not calculated during transrectal ultrasound or if biopsy was not completed. To focus our study on patients suspected to have clinically localized prostate cancer, those with metastatic disease on subsequent evaluation after biopsy were also excluded from study. Variables of patient age, date of biopsy, PSA, digital rectal examination, indication for biopsy, prostate volume, number of cores and presence of prostate cancer on biopsy were recorded for each patient in the remaining 750 patients. Prostate biopsy. The median number of cores was 12 (range 6 to 18). Biopsies were taken more laterally in the peripheral zone than those of the standard sextant biopsy.1 Figure 2 shows the distribution of cores for patients undergoing our standard 12-core biopsy. More cores were taken in areas of increased suspicion by digital rectal examination or transrectal ultrasound. Transition zone biopsies were not performed, as previous literature has shown a limited increase in cancer detection rates and noticeably increased incidence of complications such as gross hematuria.4, 11 Each biopsy core was labeled according to location on the prostate and sent separately for histological review. Biopsy was considered positive if adenocarcinoma was diagnosed. Patients with high grade prostatic intraepithelial neoplasia, atypia or dysplasia were considered to have a biopsy negative for prostate cancer, as the risk for concomitant but missed prostate cancer remains controversial for these histological findings, especially in the setting of extended needle biopsy.12, 13 Statistical methods. Variables were summarized as median, interquartile range (25th to 75th percentiles) and range of values, or as number and percent of patients. Before anal-
ysis the continuous variables of age, PSA and prostate volume were categorized by quartiles, and the number of biopsy cores was continuous or categorized as fewer than 9, 9 to 12 or greater than 12 cores. The univariate associations of clinical variables with prostate cancer detection were determined using logistic regression and the associations of clinical variables with prostate volume quartiles were determined from contingency tables. All p values resulted from 2-sided likelihood ratio chi-square tests. Multivariate logistic regression analysis controlling for age, PSA and biopsy indication was performed to determine the significance of the interaction between the number of cores and prostate volume with respect to prostate cancer detection. All statistical analyses were conducted using SAS version 8.0 (SAS Institute, Inc., Cary, North Carolina). RESULTS
A total of 767 patients underwent initial transrectal ultrasound guided prostate needle biopsy between January 1995 and August 2001. Excluded from the study were 10 patients who did not have prostate volume calculated, 3 who could not
FIG. 2. Cartoon of posterior view of prostate with areas of biopsy (cross marks) during median 12-core biopsy. Broken line represents mid lobe parasagittal line of biopsies described by Hodge et al.1
132
PROSTATE CANCER DETECTION BY EXTENDED NEEDLE BIOPSY
tolerate the biopsy and 4 who were found to have metastatic disease. A total of 750 patients were included in the study and were evenly distributed across the 6 3⁄4-year period (table 1). The cohort of 750 patients had a median age of 64 years (range 35 to 94), a median PSA of 4.4 ng./ml. (range 0.3 to 67.0) and a median prostate volume of 45 cc (range 13 to 244). Indications for biopsy included an increased or increasing PSA alone in 330 patients (44.0%), abnormal digital rectal examination alone in 226 (30.1%) and combination of PSA greater than 4.0 ng./ml. and abnormal digital rectal examination in 188 (25.1%). Digital rectal examination results were not available for 6 patients (0.8%). Of the 750 patients evaluated 253 (33.7%) were diagnosed with prostate cancer by initial prostate needle biopsy, which was clinical stage T1c in 94 (37.2%) and T2 in 157 (62.1%). Digital rectal examination results were not available for 2 patients (0.8%) diagnosed with prostate cancer. Prostate cancer detection rates did not differ across the years of the study (p ⫽ 0.87). Patients diagnosed with prostate cancer were significantly older (p ⫽ 0.0035), had higher PSA levels (p ⫽ 0.0002) and had smaller prostate glands (p ⫽ 0.0074) than those without cancer (table 1). Cancer detection rates were highest among patients with a combination of increased PSA and abnormal digital rectal examination (p ⬍0.0001), and the positive predictive value of the combination was 50.0%. Decreasing prostate cancer detection rates were noted with increasing prostate volume (p ⫽ 0.0074) as detection rates decreased from 40% to 24% with increasing quartiles of prostate volumes (table 2). Detection rates did not differ according to the number of cores (p ⫽ 0.77) as illustrated in figure 3. The OR of prostate cancer detection as biopsy numbers were increased from 6 to 18 was estimated to be 0.99 for each additional core taken during prostate biopsy (95% CI 0.93, 1.06), essentially suggesting no change, which is statistically defined as OR ⫽ 1 in prostate cancer detection rates. Thus, increasing the number of biopsy cores from 6 to 18 did not increase prostate cancer detection rates in univariate analysis. Table 2 summarizes the prostate cancer detection rates and other clinical and biopsy characteristics by prostate volume divided into quartiles. Patients with larger prostates were significantly older (p ⬍0.0001), had higher PSA levels (p ⬍0.0001) and were more likely to have undergone biopsy for increased serum PSA or combination of increased PSA and abnormal digital rectal examination rather than abnormal digital rectal examination alone (p ⬍0.0001). The number of cores did not differ across quartiles of prostate volume (p ⫽ 0.19) as illustrated in figure 4, suggesting that prostate
size did not influence the number of biopsies taken by the urologist performing the biopsies (W. C. D.). PSA density was also a significant predictor of prostate cancer detection (p ⬍0.0001, table 1). PSA density decreased as prostate volume increased (p ⬍0.0001), although PSA levels increased (p ⬍0.0001, table 2). We also observed an increase in prostate cancer detection across all quartiles of prostate volume as PSA density increased (p ⫽ 0.0098), which was most marked in the lowest volume quartile (fig. 5). A multivariate logistic regression analysis that included age, PSA or PSA density, biopsy indication, prostate volume and number of cores confirmed that each variable, except for number of cores (p ⫽ 0.87), was a significant predictor of prostate cancer detection (each p ⬍0.001). Under this analysis, increasing the number of cores from less than 9 to greater than 12 did not increase the overall prostate cancer detection rate. To determine whether lower cancer detection in larger prostates could be overcome by increasing the number of cores, we examined the relationship between quartiles of prostate gland volume, number of cores and prostate cancer detection. We determined that the number of cores did not significantly influence prostate cancer detection rates when stratified across quartiles of prostate volume (p ⫽ 0.36), again suggesting that increasing the number of cores taken at the time of prostate needle biopsy did not increase the prostate cancer detection rate in patients with larger prostate glands. This finding was surprising to us, particularly since other investigators had previously suggested an increase in cancer detection rates with increasing numbers of biopsy cores.4, 8 Instead, while we did observe slightly increased detection rates for the lower 2 quartiles of prostate volume, we actually observed slightly decreased detection rates for the upper 2 quartiles of prostate volume as the number of cores increased (fig. 6). When analyzing our 750-patient cohort for biopsies of 13 or more cores the odds of detection were 1.55 times the odds when biopsies of 8 or fewer cores were performed for the smaller prostates (OR ⫽ 1.55, 95% CI 0.82, 2.95). However, for the larger prostates the odds of detection with biopsies of 13 or more cores was approximately half of the odds of detection when 8 or fewer cores were taken (OR ⫽ 0.51, 95% CI 0.24, 1.10). None of these differences was statistically significant as confidence intervals overlapped with OR ⫽ 1. Overall, there was no statistically significant increase or decrease in prostate cancer detection rates with added cores, and no statistically significant interaction between number of cores and prostate volume with respect to cancer detection.
TABLE 1. Comparison of clinical characteristics of the 750 patients according to biopsy outcome Pos. Biopsy No. pts. (%) Biopsy yr.: 1995 1996 1997 1998 1999 2000 2001 Median pt. age (interquartile range) Median ng./ml. PSA (interquartile range) No. biopsy indication (%): PSA only Digital rectal examination only PSA plus digital rectal examination Unknown Median cc prostate vol. (interquartile range) Median ng./ml./cc PSA density (interquartile range) Median No. cores (interquartile range) * Univariate p value from logistic regression.
253 (33.7) 42 (35.0) 38 (32.8) 49 (37.1) 30 (34.5) 30 (28.0) 36 (33.6) 28 (34.6) 65 (59,72) 4.9 (3.6,7.1) 94 (28.5) 63 (27.9) 94 (50.0) 2 (N/A) 43 (31,57) 0.11 (.07,.17) 12 (10,13)
Neg. Biopsy 497 (66.3) 78 (65.0) 78 (67.2) 83 (62.9) 57 (65.5) 77 (72.0) 71 (66.4) 53 (65.4) 63 (56,70) 4.2 (2.3,5.9) 236 (71.5) 163 (72.1) 94 (50.0) 4 (N/A) 47 (36,67) 0.07 (.06,.11) 12 (10,13)
p Value* — 0.87
0.0035 0.0002 ⬍0.0001
0.0074 ⬍0.0001 0.77
133
PROSTATE CANCER DETECTION BY EXTENDED NEEDLE BIOPSY TABLE 2. Clinical and biopsy characteristics of patients according to quartiles of prostate volume Vol. Quartile 1 Vol. range (cc) No. pts. (%) Median pt. age (interquartile range) Median ng./ml. PSA (interquartile range) Median ng./ml./cc PSA density (interquartile range) No. biopsy indication (%) PSA only Digital rectal examination only PSA plus digital rectal examination Unknown Median No. cores (interquartile range) No. biopsy results (%): Pos. Neg. * Univariate p value from chi-square tests.
13–34 194 (25.9) 59 (53,67) 3.1 (1.1,4.9) 0.11 (.04,.19)
2 34.1–45 184 (24.5) 63 (57,69) 4.1 (2.5,5.3) 0.10 (.06,.14)
3 45.1–64 189 (25.2) 64 (59,71) 4.4 (3.3,5.9) 0.08 (.06,.11)
4 64.1–244 183 (24.4) 68 (62,73) 6.2 (4.6,8.8) 0.07 (.06,.10)
p Value*
— p ⬍0.0001 p ⬍0.0001 p ⬍0.0001 p ⬍0.0001
63 (32.5) 94 (48.5) 36 (18.6) 1 (0.5) 12 (10,13) 78 (40.2) 116 (59.8)
77 (41.8) 65 (35.3) 42 (22.8) 0 (0) 11 (9,13) 65 (35.3) 119 (64.7)
85 (45.0) 52 (27.5) 48 (25.4) 4 (2.1) 12 (10,13) 66 (34.9) 123 (65.1)
105 (57.4) 15 (8.2) 62 (33.9) 1 (0.5) 11 (10,13)
p ⫽ 0.194 p ⫽ 0.0074
44 (24.0) 139 (76.0)
FIG. 3. Prostate cancer detection rates based on number of cores categorized as less than 9, 9 to 12 and greater than 12. Increasing number of cores did not increase prostate cancer detection rates (univariate analysis p ⫽ 0.77).
FIG. 4. Median number of cores taken at biopsy based on prostate volume categorized by quartiles. 1 (13 to 34 cc), 2 (34.1 to 45 cc), 3 (45.1 to 64 cc) and 4 (64.1 to 244 cc). Broken lines indicate 25th to 75th percentile interquartile range of number of cores. Number of cores did not vary according to prostate size (univariate analysis p ⫽ 0.19).
DISCUSSION
Early detection and treatment of prostate cancer when it is still localized allow the greatest chance for cure. As screening for prostate cancer with serum PSA testing has become more
widespread, lesions detected by digital rectal examination have become less common and cancers are detected at an earlier stage.14 Mathematical models suggest that taking more biopsies increases the chance of prostate cancer detec-
134
PROSTATE CANCER DETECTION BY EXTENDED NEEDLE BIOPSY
FIG. 5. Median PSA density of patients with and without prostate cancer based on prostate volume categorized by quartiles 1 (13 to 34 cc), 2 (34.1 to 45 cc), 3 (45.1 to 64 cc) and 4 (64.1 to 244 cc). Broken lines indicate 25th to 75th percentile interquartile range of PSA density.
FIG. 6. Estimated prostate cancer detection rates across prostate volume quartiles stratified by number of cores and adjusting for age, PSA and biopsy indication in multivariate model. Number of cores did not influence prostate cancer detection rates (p ⫽ 0.36) across prostate volume quartiles 1 (13 to 34 cc), 2 (34.1 to 45 cc), 3 (45.1 to 64 cc) and 4 (64.1 to 244 cc).
tion.2, 10 However, in clinical practice there are a number of variables that predetermine risk of cancer and likelihood of cancer detection. Prostate size is an important variable in the diagnosis of prostate cancer since it has a direct role in the relative amount of tissue that is sampled per biopsy core. Although larger prostates should not have a lower incidence of prostate cancer, our data showed lower cancer detection rates with larger prostate glands, consistent with previously published sextant biopsy data.5– 8 These previous studies sampled prostates only along the mid lobe parasagittal line, which we now know misses a large portion of the peripheral zone.3, 4 In fact, Hodge et al have revised his original sextant biopsy method laterally to better sample the peripheral zone and improve cancer detection rates.1, 15 Thus, location of biopsy cores has an important role in detecting cancers. The biopsy cores of our study sample this lateral peripheral zone. The number of previous biopsies also changes the probability of finding prostate cancer.8, 16 The issues of when to repeat biopsy and in whom remain controversial. Although repeat biopsies have been shown to increase the overall detection rate of prostate cancer, we kept our patient population more homogeneous by including only those undergoing biopsy for the first time.
As the median number of biopsies in our study is 12 cores, our chance of missing a clinically significant cancer is lower than that of the sextant biopsy, especially with the support in the literature on extended needle biopsy.4, 8, 10, 13 We hoped to show an improvement in cancer detection with more cores, especially in the larger prostates where sampling error could have a role. However, our yields of cancer detection did not increase despite increased sampling with more cores. In fact, as illustrated in figure 6, we noted an estimated but not statistically significant decrease in cancer detection rates with an increased number of cores with larger prostates. However, with smaller prostates we noted an estimated, although again not statistically significant, increase in detection with increased core sampling. Thus, our data suggest that sampling error does not explain the decreased cancer detection rate for larger prostates. Despite increasing the number of cores from as low as 6 to as high as 18, we were not able to improve our cancer detection rate in large prostates. Several issues regarding our lack of an effect of biopsy number on cancer detection rates require further discussion. Although we limited our study to first time biopsies to avoid the bias of increased detection in patients undergoing repeat biopsies, perhaps this altered our patient population. However, our overall cancer detection rate of 33.7% is similar
PROSTATE CANCER DETECTION BY EXTENDED NEEDLE BIOPSY
to that of other studies.4 – 8, 16, 17 It is also possible that our range of numbers of cores is still not wide enough to see an effect on cancer detection rates. However, large numbers of biopsy cores are difficult for patients to tolerate without increased analgesia, anesthesia or sedation and put them at increased risk for complications. A further increase in the number of cores would approach the numbers used in saturation needle biopsy reports.13 Our end point of prostate cancer detection and not incidence may have affected the results of our study. For example, more than two-thirds of men with high PSA levels did not have prostate cancer and their biopsy results will be negative regardless of the number of cores taken. Similarly, in men with prostate cancer, once a certain number of cores have been taken to diagnose the cancer, additional cores will not alter the result. In patients with a high suspicion of prostate cancer and a small prostate, thus decreasing the risk of sampling error, there may be little difference in cancer detection rates within a range of 6 to 18 cores. Finally, although our study involved a large cohort of 750 patients, it is a retrospective analysis and not a prospective, randomized study. A larger cohort may be needed to find a statistically significant trend. However, even in a prospective randomized trial Naughton et al found no difference in cancer rates with 6 versus 12 biopsy cores.17 Without clear evidence for sampling error in our study, the differences in patients with increasing prostate size undergoing biopsy must be scrutinized. As expected, patients with larger prostates are older, consistent with autopsy studies showing an increased incidence and prevalence of BPH with age.9 They also have higher serum PSA levels than those with smaller prostates. As more patients in the current era of PSA screening undergo biopsy for increased or increasing PSA levels without palpable abnormalities, one must keep in mind that PSA is only organ specific and not cancer specific.9, 18 Our data demonstrate that patients with larger prostates are more likely to undergo biopsy for increased PSA than those with smaller prostates, suggesting that many of these men undergo biopsy for BPH. Improving the use of serum PSA by stratifying it into BPH specific PSA and cancer specific PSA may ultimately assist the clinician in recommending prostate needle biopsy with improved specificity.19, 20
3.
4. 5. 6. 7. 8.
9. 10. 11. 12.
13.
14.
15. 16.
CONCLUSIONS
Previous investigators have proposed that sampling error accounts for the lower prostate cancer detection rate of larger prostate glands. Our study suggests that the lower cancer detection rate may be due to a decrease in the use of increased serum PSA for prostate cancer detection in larger prostates in addition to other factors such as sampling error. Increased serum PSA, although a known risk factor for prostate cancer warranting biopsy, may also be due to nonmalignant sources in the asymptomatic patient, such as BPH, especially in larger prostates.
17.
18.
19.
REFERENCES
1. Hodge, K. K., McNeal, J. E., Terris, M. K. and Stamey, T. A.: Random systematic versus directed ultrasound guided transrectal core biopsies of the prostate. J Urol, 142: 71, 1989 2. Stricker, H. J., Ruddock, L. J., Wan, J. and Belville, W. D.:
20.
135
Detection of non-palpable prostate cancer. A mathematical and laboratory model. Br J Urol, 71: 43, 1993 McNeal, J. E., Redwine, E. A., Freiha, F. S. and Stamey, T. A.: Zonal distribution of prostatic adenocarcinoma: correlation with histologic pattern and direction of spread. Am J Surg Pathol, 12: 897, 1988 Eskew, L. A., Bare, R. L. and McCullough, D. L.: Systematic 5 region prostate biopsy is superior to sextant method for diagnosing carcinoma of the prostate. J Urol, 157: 199, 1997 Uzzo, R. G., Wei, J. T., Waldbaum, R. S., Perlmutter, A. P., Byrne, J. C. and Vaughan, E. D., Jr.: The influence of prostate size on cancer detection. Urology, 46: 831, 1995 Karakiewicz, P. I., Bazinet, M., Aprikian, A. G., Trudel, C., Aronson, S., Nachabe´ , M. et al: Outcome of sextant biopsy according to gland volume. Urology, 49: 55, 1997 Letran, J. L., Meyer, G. E., Loberiza, F. R. and Brawer, M. K.: The effect of prostate volume on the yield of needle biopsy. J Urol, 160: 1718, 1998 Levine, M. A., Ittman, M., Melamed, J. and Lepor, H.: Two consecutive sets of transrectal ultrasound guided sextant biopsies of the prostate for the detection of prostate cancer. J Urol, 159: 471, 1998 Bostwick, D. G., Cooner, W. H., Denis, L., Jones, G. W., Scardino, P. T. and Murphy, G. P.: The association of benign prostatic hyperplasia and cancer of the prostate. Cancer, 70: 291, 1992 Vashi, A. R., Wojno, K. J., Gillespie, B. and Oesterling, J. E.: A model for the number of cores per prostate biopsy based on patient age and prostate gland volume. J Urol, 159: 920, 1998 Lui, P. D., Terris, M. K., McNeal, J. E. and Stamey, T. A.: Indications for ultrasound guided transition zone biopsies in the detection of prostate cancer. J Urol, 153: 1000, 1995 Borboroglu, P. G., Sur, R. L., Roberts, J. L. and Amling, C. L.: Repeat biopsy strategy in patients with atypical small acinar proliferation or high grade prostatic intraepithelial neoplasia on initial prostate needle biopsy. J Urol, 166: 866, 2001 Stewart, C. S., Leibovich, B. C., Weaver, A. L. and Lieber, M. M.: Prostate cancer diagnosis using a saturation needle biopsy technique after previous negative sextant biopsies. J Urol, 166: 86, 2001 Han, M., Partin, A. W., Piantadosi, S., Epstein, J. I. and Walsh, P. C.: Era specific biochemical recurrence-free survival following radical prostatectomy for clinically localized prostate cancer. J Urol, 166: 416, 2001 Stamey, T. A.: Making the most out of six systematic sextant biopsies. Urology, 45: 2, 1995 Djavan, B., Ravery, V., Zlotta, A., Dobronski, P., Dobrovits, M., Fakhari, M. et al: Prospective evaluation of prostate cancer detected on biopsies 1, 2, 3 and 4: when should we stop? J Urol, 166: 1679, 2001 Naughton, C. K., Miller, D. C., Mager, D. E., Ornstein, D. K. and Catalona, W. J.: A prospective randomized trial comparing 6 versus 12 prostate needle biopsy cores: impact on cancer detection. J Urol, 164: 388, 2000 Kibel, A. S., Krithivas, K., Shamel, L. B., Kantoff, P. W. and DeWolf, W. C.: Constitutive expression of high levels of prostate-specific antigen in the absence of prostate carcinoma. Urology, 48: 741, 1996 Mikolajczyk, S. D., Millar, L. S., Wang, T. J., Rittenhouse, H. G., Wolfert, R. L., Marks, L. S. et al: “BPSA,” a specific molecular form of free prostate-specific antigen, is found predominantly in the transition zone of patients with nodular benign prostatic hyperplasia. Urology, 55: 41, 2000 Mikolajczyk, S. D., Marker, K. M., Millar, L. S., Kumar, A., Saedi, M. S., Payne, J. K. et al: A truncated precursor form of prostate-specific antigen is a more specific serum marker of prostate cancer. Cancer Res, 61: 6958, 2001
Vol. 169, 130 –135, January 2003 Printed in U.S.A.
DOI: 10.1097/01.ju.0000034153.49106.b7
THE RELATIONSHIP OF PROSTATE GLAND VOLUME TO EXTENDED NEEDLE BIOPSY ON PROSTATE CANCER DETECTION JEAN O. UNG, IGNACIO F. SAN FRANCISCO, MEREDITH M. REGAN, WILLIAM C. DEWOLF ARIA F. OLUMI
AND
From the Division of Urologic Surgery and Biometrics Center, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts
ABSTRACT
Purpose: We investigated the relationship between prostate volume and cancer detection by needle biopsy, and determined the effect of an increased number of cores on the sampling error of needle biopsy on large prostate glands. Materials and Methods: The study cohort included 750 consecutive patients who underwent first time transrectal ultrasound guided prostate needle biopsy from January 1995 to August 2001. Prostate volumes were divided into quartiles (13 to 34, 34.1 to 45, 45.1 to 64 and 64.1 to 244 cc). Multivariate analysis controlling for age, prostate specific antigen (PSA) and biopsy indication was performed to determine the effect of the number of cores and prostate volume on prostate cancer detection. Results: Patients diagnosed with prostate cancer were older (p ⫽ 0.0035) and had higher PSA levels (p ⫽ 0.0002) than those with no cancer on biopsy. Decreasing cancer detection rates were seen with increasing prostate volume (p ⫽ 0.0074). The OR of detection for each additional core was 0.99 (95% CI 0.93, 1.06), suggesting that increasing the number of biopsy cores did not increase the rate of prostate cancer detection. Multivariate analysis revealed that patients with larger prostates had the same, or possibly lower, cancer detection rate as the number of biopsy cores was increased. Patients with larger prostates were older (p ⬍0.0001), had higher PSA levels (p ⬍0.0001) and were even more likely to have undergone biopsy for increased PSA rather than abnormal digital rectal examination alone (p ⬍0.0001). Conclusions: Our study suggests that the lower cancer detection rate for men with large prostates may be due to a decrease in the use of increased serum PSA for prostate cancer detection in larger prostates in addition to other factors such as sampling error. Increased serum PSA levels in cases of larger prostates, although a risk factor for prostate cancer warranting biopsy, may also be due to nonmalignant sources such as benign prostatic hyperplasia. KEY WORDS: prostatic neoplasms, prostate-specific antigen, prostatic hyperplasia, biopsy
The early diagnosis of prostate cancer has evolved with the widespread use of serum prostate specific antigen (PSA) testing and transrectal ultrasound in addition to digital rectal examination. Transrectal ultrasound guided prostate needle biopsies have become the method of choice for the detection of prostate cancer. The sextant biopsy method, first proposed by Hodge et al in 1989, is still the standard strategy of systematic prostate biopsy.1 However, the optimal number of biopsies needed to identify all patients with prostate cancer at the earliest stage possible for optimal treatment, outcome and survival is not known. Stricker et al described a mathematical model based on Bayes’ theorem of conditional probability which demonstrated a correlation between the percent volume of prostate cancer within a prostate gland and the probability of a positive biopsy.2 The probability of detecting prostate cancer for a fixed percent cancer volume increases as the number of biopsies increases. However, this model assumes a uniform distribution of prostate carcinoma throughout the prostate gland and a random distribution of biopsies. Most data support a modified distribution of prostate carcinoma when the majority of tumors occur in the peripheral zone, as the transition zone is better known as the site of origin of benign
prostatic hyperplasia (BPH),3 and most biopsies are taken accordingly. Eskew et al reported that a systematic 5-region prostate biopsy of 13 to 18 cores is superior to the sextant method for diagnosing prostate cancer, with 35% more cases of prostate cancer detected.4 Several other authors have shown that the yield of sextant biopsy decreases with increasing prostate volume.5– 8 They have concluded that this finding is due to a sampling error that occurs in men with larger prostate glands. With no evidence to suggest differing intrinsic prostate cancer rates based on prostate volume alone,5, 9 additional biopsies could be necessary for larger prostate glands to detect prostate cancer. We further investigate the relationship between prostate gland volume and cancer detection by needle biopsy. The significance of sampling error with the sextant biopsy method has been noted. However, further study is necessary to evaluate the effect of extended needle biopsy on cancer detection rates with different prostate volumes. Vashi et al proposed a mathematical model for predicting the number of cores needed to detect clinically significant tumors at an early enough stage to provide the best chance of decreasing mortality.10 They recommend anywhere from 2 to 20 biopsies depending on prostate volume and patient age but their study is based on a mathematical model and lacks clinical data for correlation. Our study is based on clinical data from a single urologist (W. C. D.) and includes patients suspected
Accepted for publication May 24, 2002. Supported by the Hershey Family Prostate Cancer Foundation, Beth Israel Deaconess Medical Center. 130
PROSTATE CANCER DETECTION BY EXTENDED NEEDLE BIOPSY
131
FIG. 1. Distribution of prostate volumes by results of prostate needle biopsy. Median 45 cc, range 13 to 244 cc, total number 750 patients. Horizontal axis labels indicate upper bound of range (for example, less than 30 indicates 20 to 29.9).
to be at risk for clinically localized prostate cancer who subsequently underwent transrectal ultrasound guided prostate needle biopsy. METHODS
Patient selection. Between January 1995 and August 2001, 767 consecutive patients seen by a single urologist (W. C. D.) at our institution for suspicion of clinically localized prostate cancer underwent initial transrectal ultrasound guided prostate needle biopsy. Indications for biopsy included increased or increasing PSA and/or abnormal digital rectal examination. Transrectal ultrasound prostate volume was calculated using a computer generated elliptical estimation of 0.52 ⫻ length ⫻ width ⫻ height. The distribution of prostate volumes of patients undergoing prostate needle biopsy is illustrated in figure 1. Patients were excluded from study if prostate volume was not calculated during transrectal ultrasound or if biopsy was not completed. To focus our study on patients suspected to have clinically localized prostate cancer, those with metastatic disease on subsequent evaluation after biopsy were also excluded from study. Variables of patient age, date of biopsy, PSA, digital rectal examination, indication for biopsy, prostate volume, number of cores and presence of prostate cancer on biopsy were recorded for each patient in the remaining 750 patients. Prostate biopsy. The median number of cores was 12 (range 6 to 18). Biopsies were taken more laterally in the peripheral zone than those of the standard sextant biopsy.1 Figure 2 shows the distribution of cores for patients undergoing our standard 12-core biopsy. More cores were taken in areas of increased suspicion by digital rectal examination or transrectal ultrasound. Transition zone biopsies were not performed, as previous literature has shown a limited increase in cancer detection rates and noticeably increased incidence of complications such as gross hematuria.4, 11 Each biopsy core was labeled according to location on the prostate and sent separately for histological review. Biopsy was considered positive if adenocarcinoma was diagnosed. Patients with high grade prostatic intraepithelial neoplasia, atypia or dysplasia were considered to have a biopsy negative for prostate cancer, as the risk for concomitant but missed prostate cancer remains controversial for these histological findings, especially in the setting of extended needle biopsy.12, 13 Statistical methods. Variables were summarized as median, interquartile range (25th to 75th percentiles) and range of values, or as number and percent of patients. Before anal-
ysis the continuous variables of age, PSA and prostate volume were categorized by quartiles, and the number of biopsy cores was continuous or categorized as fewer than 9, 9 to 12 or greater than 12 cores. The univariate associations of clinical variables with prostate cancer detection were determined using logistic regression and the associations of clinical variables with prostate volume quartiles were determined from contingency tables. All p values resulted from 2-sided likelihood ratio chi-square tests. Multivariate logistic regression analysis controlling for age, PSA and biopsy indication was performed to determine the significance of the interaction between the number of cores and prostate volume with respect to prostate cancer detection. All statistical analyses were conducted using SAS version 8.0 (SAS Institute, Inc., Cary, North Carolina). RESULTS
A total of 767 patients underwent initial transrectal ultrasound guided prostate needle biopsy between January 1995 and August 2001. Excluded from the study were 10 patients who did not have prostate volume calculated, 3 who could not
FIG. 2. Cartoon of posterior view of prostate with areas of biopsy (cross marks) during median 12-core biopsy. Broken line represents mid lobe parasagittal line of biopsies described by Hodge et al.1
132
PROSTATE CANCER DETECTION BY EXTENDED NEEDLE BIOPSY
tolerate the biopsy and 4 who were found to have metastatic disease. A total of 750 patients were included in the study and were evenly distributed across the 6 3⁄4-year period (table 1). The cohort of 750 patients had a median age of 64 years (range 35 to 94), a median PSA of 4.4 ng./ml. (range 0.3 to 67.0) and a median prostate volume of 45 cc (range 13 to 244). Indications for biopsy included an increased or increasing PSA alone in 330 patients (44.0%), abnormal digital rectal examination alone in 226 (30.1%) and combination of PSA greater than 4.0 ng./ml. and abnormal digital rectal examination in 188 (25.1%). Digital rectal examination results were not available for 6 patients (0.8%). Of the 750 patients evaluated 253 (33.7%) were diagnosed with prostate cancer by initial prostate needle biopsy, which was clinical stage T1c in 94 (37.2%) and T2 in 157 (62.1%). Digital rectal examination results were not available for 2 patients (0.8%) diagnosed with prostate cancer. Prostate cancer detection rates did not differ across the years of the study (p ⫽ 0.87). Patients diagnosed with prostate cancer were significantly older (p ⫽ 0.0035), had higher PSA levels (p ⫽ 0.0002) and had smaller prostate glands (p ⫽ 0.0074) than those without cancer (table 1). Cancer detection rates were highest among patients with a combination of increased PSA and abnormal digital rectal examination (p ⬍0.0001), and the positive predictive value of the combination was 50.0%. Decreasing prostate cancer detection rates were noted with increasing prostate volume (p ⫽ 0.0074) as detection rates decreased from 40% to 24% with increasing quartiles of prostate volumes (table 2). Detection rates did not differ according to the number of cores (p ⫽ 0.77) as illustrated in figure 3. The OR of prostate cancer detection as biopsy numbers were increased from 6 to 18 was estimated to be 0.99 for each additional core taken during prostate biopsy (95% CI 0.93, 1.06), essentially suggesting no change, which is statistically defined as OR ⫽ 1 in prostate cancer detection rates. Thus, increasing the number of biopsy cores from 6 to 18 did not increase prostate cancer detection rates in univariate analysis. Table 2 summarizes the prostate cancer detection rates and other clinical and biopsy characteristics by prostate volume divided into quartiles. Patients with larger prostates were significantly older (p ⬍0.0001), had higher PSA levels (p ⬍0.0001) and were more likely to have undergone biopsy for increased serum PSA or combination of increased PSA and abnormal digital rectal examination rather than abnormal digital rectal examination alone (p ⬍0.0001). The number of cores did not differ across quartiles of prostate volume (p ⫽ 0.19) as illustrated in figure 4, suggesting that prostate
size did not influence the number of biopsies taken by the urologist performing the biopsies (W. C. D.). PSA density was also a significant predictor of prostate cancer detection (p ⬍0.0001, table 1). PSA density decreased as prostate volume increased (p ⬍0.0001), although PSA levels increased (p ⬍0.0001, table 2). We also observed an increase in prostate cancer detection across all quartiles of prostate volume as PSA density increased (p ⫽ 0.0098), which was most marked in the lowest volume quartile (fig. 5). A multivariate logistic regression analysis that included age, PSA or PSA density, biopsy indication, prostate volume and number of cores confirmed that each variable, except for number of cores (p ⫽ 0.87), was a significant predictor of prostate cancer detection (each p ⬍0.001). Under this analysis, increasing the number of cores from less than 9 to greater than 12 did not increase the overall prostate cancer detection rate. To determine whether lower cancer detection in larger prostates could be overcome by increasing the number of cores, we examined the relationship between quartiles of prostate gland volume, number of cores and prostate cancer detection. We determined that the number of cores did not significantly influence prostate cancer detection rates when stratified across quartiles of prostate volume (p ⫽ 0.36), again suggesting that increasing the number of cores taken at the time of prostate needle biopsy did not increase the prostate cancer detection rate in patients with larger prostate glands. This finding was surprising to us, particularly since other investigators had previously suggested an increase in cancer detection rates with increasing numbers of biopsy cores.4, 8 Instead, while we did observe slightly increased detection rates for the lower 2 quartiles of prostate volume, we actually observed slightly decreased detection rates for the upper 2 quartiles of prostate volume as the number of cores increased (fig. 6). When analyzing our 750-patient cohort for biopsies of 13 or more cores the odds of detection were 1.55 times the odds when biopsies of 8 or fewer cores were performed for the smaller prostates (OR ⫽ 1.55, 95% CI 0.82, 2.95). However, for the larger prostates the odds of detection with biopsies of 13 or more cores was approximately half of the odds of detection when 8 or fewer cores were taken (OR ⫽ 0.51, 95% CI 0.24, 1.10). None of these differences was statistically significant as confidence intervals overlapped with OR ⫽ 1. Overall, there was no statistically significant increase or decrease in prostate cancer detection rates with added cores, and no statistically significant interaction between number of cores and prostate volume with respect to cancer detection.
TABLE 1. Comparison of clinical characteristics of the 750 patients according to biopsy outcome Pos. Biopsy No. pts. (%) Biopsy yr.: 1995 1996 1997 1998 1999 2000 2001 Median pt. age (interquartile range) Median ng./ml. PSA (interquartile range) No. biopsy indication (%): PSA only Digital rectal examination only PSA plus digital rectal examination Unknown Median cc prostate vol. (interquartile range) Median ng./ml./cc PSA density (interquartile range) Median No. cores (interquartile range) * Univariate p value from logistic regression.
253 (33.7) 42 (35.0) 38 (32.8) 49 (37.1) 30 (34.5) 30 (28.0) 36 (33.6) 28 (34.6) 65 (59,72) 4.9 (3.6,7.1) 94 (28.5) 63 (27.9) 94 (50.0) 2 (N/A) 43 (31,57) 0.11 (.07,.17) 12 (10,13)
Neg. Biopsy 497 (66.3) 78 (65.0) 78 (67.2) 83 (62.9) 57 (65.5) 77 (72.0) 71 (66.4) 53 (65.4) 63 (56,70) 4.2 (2.3,5.9) 236 (71.5) 163 (72.1) 94 (50.0) 4 (N/A) 47 (36,67) 0.07 (.06,.11) 12 (10,13)
p Value* — 0.87
0.0035 0.0002 ⬍0.0001
0.0074 ⬍0.0001 0.77
133
PROSTATE CANCER DETECTION BY EXTENDED NEEDLE BIOPSY TABLE 2. Clinical and biopsy characteristics of patients according to quartiles of prostate volume Vol. Quartile 1 Vol. range (cc) No. pts. (%) Median pt. age (interquartile range) Median ng./ml. PSA (interquartile range) Median ng./ml./cc PSA density (interquartile range) No. biopsy indication (%) PSA only Digital rectal examination only PSA plus digital rectal examination Unknown Median No. cores (interquartile range) No. biopsy results (%): Pos. Neg. * Univariate p value from chi-square tests.
13–34 194 (25.9) 59 (53,67) 3.1 (1.1,4.9) 0.11 (.04,.19)
2 34.1–45 184 (24.5) 63 (57,69) 4.1 (2.5,5.3) 0.10 (.06,.14)
3 45.1–64 189 (25.2) 64 (59,71) 4.4 (3.3,5.9) 0.08 (.06,.11)
4 64.1–244 183 (24.4) 68 (62,73) 6.2 (4.6,8.8) 0.07 (.06,.10)
p Value*
— p ⬍0.0001 p ⬍0.0001 p ⬍0.0001 p ⬍0.0001
63 (32.5) 94 (48.5) 36 (18.6) 1 (0.5) 12 (10,13) 78 (40.2) 116 (59.8)
77 (41.8) 65 (35.3) 42 (22.8) 0 (0) 11 (9,13) 65 (35.3) 119 (64.7)
85 (45.0) 52 (27.5) 48 (25.4) 4 (2.1) 12 (10,13) 66 (34.9) 123 (65.1)
105 (57.4) 15 (8.2) 62 (33.9) 1 (0.5) 11 (10,13)
p ⫽ 0.194 p ⫽ 0.0074
44 (24.0) 139 (76.0)
FIG. 3. Prostate cancer detection rates based on number of cores categorized as less than 9, 9 to 12 and greater than 12. Increasing number of cores did not increase prostate cancer detection rates (univariate analysis p ⫽ 0.77).
FIG. 4. Median number of cores taken at biopsy based on prostate volume categorized by quartiles. 1 (13 to 34 cc), 2 (34.1 to 45 cc), 3 (45.1 to 64 cc) and 4 (64.1 to 244 cc). Broken lines indicate 25th to 75th percentile interquartile range of number of cores. Number of cores did not vary according to prostate size (univariate analysis p ⫽ 0.19).
DISCUSSION
Early detection and treatment of prostate cancer when it is still localized allow the greatest chance for cure. As screening for prostate cancer with serum PSA testing has become more
widespread, lesions detected by digital rectal examination have become less common and cancers are detected at an earlier stage.14 Mathematical models suggest that taking more biopsies increases the chance of prostate cancer detec-
134
PROSTATE CANCER DETECTION BY EXTENDED NEEDLE BIOPSY
FIG. 5. Median PSA density of patients with and without prostate cancer based on prostate volume categorized by quartiles 1 (13 to 34 cc), 2 (34.1 to 45 cc), 3 (45.1 to 64 cc) and 4 (64.1 to 244 cc). Broken lines indicate 25th to 75th percentile interquartile range of PSA density.
FIG. 6. Estimated prostate cancer detection rates across prostate volume quartiles stratified by number of cores and adjusting for age, PSA and biopsy indication in multivariate model. Number of cores did not influence prostate cancer detection rates (p ⫽ 0.36) across prostate volume quartiles 1 (13 to 34 cc), 2 (34.1 to 45 cc), 3 (45.1 to 64 cc) and 4 (64.1 to 244 cc).
tion.2, 10 However, in clinical practice there are a number of variables that predetermine risk of cancer and likelihood of cancer detection. Prostate size is an important variable in the diagnosis of prostate cancer since it has a direct role in the relative amount of tissue that is sampled per biopsy core. Although larger prostates should not have a lower incidence of prostate cancer, our data showed lower cancer detection rates with larger prostate glands, consistent with previously published sextant biopsy data.5– 8 These previous studies sampled prostates only along the mid lobe parasagittal line, which we now know misses a large portion of the peripheral zone.3, 4 In fact, Hodge et al have revised his original sextant biopsy method laterally to better sample the peripheral zone and improve cancer detection rates.1, 15 Thus, location of biopsy cores has an important role in detecting cancers. The biopsy cores of our study sample this lateral peripheral zone. The number of previous biopsies also changes the probability of finding prostate cancer.8, 16 The issues of when to repeat biopsy and in whom remain controversial. Although repeat biopsies have been shown to increase the overall detection rate of prostate cancer, we kept our patient population more homogeneous by including only those undergoing biopsy for the first time.
As the median number of biopsies in our study is 12 cores, our chance of missing a clinically significant cancer is lower than that of the sextant biopsy, especially with the support in the literature on extended needle biopsy.4, 8, 10, 13 We hoped to show an improvement in cancer detection with more cores, especially in the larger prostates where sampling error could have a role. However, our yields of cancer detection did not increase despite increased sampling with more cores. In fact, as illustrated in figure 6, we noted an estimated but not statistically significant decrease in cancer detection rates with an increased number of cores with larger prostates. However, with smaller prostates we noted an estimated, although again not statistically significant, increase in detection with increased core sampling. Thus, our data suggest that sampling error does not explain the decreased cancer detection rate for larger prostates. Despite increasing the number of cores from as low as 6 to as high as 18, we were not able to improve our cancer detection rate in large prostates. Several issues regarding our lack of an effect of biopsy number on cancer detection rates require further discussion. Although we limited our study to first time biopsies to avoid the bias of increased detection in patients undergoing repeat biopsies, perhaps this altered our patient population. However, our overall cancer detection rate of 33.7% is similar
PROSTATE CANCER DETECTION BY EXTENDED NEEDLE BIOPSY
to that of other studies.4 – 8, 16, 17 It is also possible that our range of numbers of cores is still not wide enough to see an effect on cancer detection rates. However, large numbers of biopsy cores are difficult for patients to tolerate without increased analgesia, anesthesia or sedation and put them at increased risk for complications. A further increase in the number of cores would approach the numbers used in saturation needle biopsy reports.13 Our end point of prostate cancer detection and not incidence may have affected the results of our study. For example, more than two-thirds of men with high PSA levels did not have prostate cancer and their biopsy results will be negative regardless of the number of cores taken. Similarly, in men with prostate cancer, once a certain number of cores have been taken to diagnose the cancer, additional cores will not alter the result. In patients with a high suspicion of prostate cancer and a small prostate, thus decreasing the risk of sampling error, there may be little difference in cancer detection rates within a range of 6 to 18 cores. Finally, although our study involved a large cohort of 750 patients, it is a retrospective analysis and not a prospective, randomized study. A larger cohort may be needed to find a statistically significant trend. However, even in a prospective randomized trial Naughton et al found no difference in cancer rates with 6 versus 12 biopsy cores.17 Without clear evidence for sampling error in our study, the differences in patients with increasing prostate size undergoing biopsy must be scrutinized. As expected, patients with larger prostates are older, consistent with autopsy studies showing an increased incidence and prevalence of BPH with age.9 They also have higher serum PSA levels than those with smaller prostates. As more patients in the current era of PSA screening undergo biopsy for increased or increasing PSA levels without palpable abnormalities, one must keep in mind that PSA is only organ specific and not cancer specific.9, 18 Our data demonstrate that patients with larger prostates are more likely to undergo biopsy for increased PSA than those with smaller prostates, suggesting that many of these men undergo biopsy for BPH. Improving the use of serum PSA by stratifying it into BPH specific PSA and cancer specific PSA may ultimately assist the clinician in recommending prostate needle biopsy with improved specificity.19, 20
3.
4. 5. 6. 7. 8.
9. 10. 11. 12.
13.
14.
15. 16.
CONCLUSIONS
Previous investigators have proposed that sampling error accounts for the lower prostate cancer detection rate of larger prostate glands. Our study suggests that the lower cancer detection rate may be due to a decrease in the use of increased serum PSA for prostate cancer detection in larger prostates in addition to other factors such as sampling error. Increased serum PSA, although a known risk factor for prostate cancer warranting biopsy, may also be due to nonmalignant sources in the asymptomatic patient, such as BPH, especially in larger prostates.
17.
18.
19.
REFERENCES
1. Hodge, K. K., McNeal, J. E., Terris, M. K. and Stamey, T. A.: Random systematic versus directed ultrasound guided transrectal core biopsies of the prostate. J Urol, 142: 71, 1989 2. Stricker, H. J., Ruddock, L. J., Wan, J. and Belville, W. D.:
20.
135
Detection of non-palpable prostate cancer. A mathematical and laboratory model. Br J Urol, 71: 43, 1993 McNeal, J. E., Redwine, E. A., Freiha, F. S. and Stamey, T. A.: Zonal distribution of prostatic adenocarcinoma: correlation with histologic pattern and direction of spread. Am J Surg Pathol, 12: 897, 1988 Eskew, L. A., Bare, R. L. and McCullough, D. L.: Systematic 5 region prostate biopsy is superior to sextant method for diagnosing carcinoma of the prostate. J Urol, 157: 199, 1997 Uzzo, R. G., Wei, J. T., Waldbaum, R. S., Perlmutter, A. P., Byrne, J. C. and Vaughan, E. D., Jr.: The influence of prostate size on cancer detection. Urology, 46: 831, 1995 Karakiewicz, P. I., Bazinet, M., Aprikian, A. G., Trudel, C., Aronson, S., Nachabe´ , M. et al: Outcome of sextant biopsy according to gland volume. Urology, 49: 55, 1997 Letran, J. L., Meyer, G. E., Loberiza, F. R. and Brawer, M. K.: The effect of prostate volume on the yield of needle biopsy. J Urol, 160: 1718, 1998 Levine, M. A., Ittman, M., Melamed, J. and Lepor, H.: Two consecutive sets of transrectal ultrasound guided sextant biopsies of the prostate for the detection of prostate cancer. J Urol, 159: 471, 1998 Bostwick, D. G., Cooner, W. H., Denis, L., Jones, G. W., Scardino, P. T. and Murphy, G. P.: The association of benign prostatic hyperplasia and cancer of the prostate. Cancer, 70: 291, 1992 Vashi, A. R., Wojno, K. J., Gillespie, B. and Oesterling, J. E.: A model for the number of cores per prostate biopsy based on patient age and prostate gland volume. J Urol, 159: 920, 1998 Lui, P. D., Terris, M. K., McNeal, J. E. and Stamey, T. A.: Indications for ultrasound guided transition zone biopsies in the detection of prostate cancer. J Urol, 153: 1000, 1995 Borboroglu, P. G., Sur, R. L., Roberts, J. L. and Amling, C. L.: Repeat biopsy strategy in patients with atypical small acinar proliferation or high grade prostatic intraepithelial neoplasia on initial prostate needle biopsy. J Urol, 166: 866, 2001 Stewart, C. S., Leibovich, B. C., Weaver, A. L. and Lieber, M. M.: Prostate cancer diagnosis using a saturation needle biopsy technique after previous negative sextant biopsies. J Urol, 166: 86, 2001 Han, M., Partin, A. W., Piantadosi, S., Epstein, J. I. and Walsh, P. C.: Era specific biochemical recurrence-free survival following radical prostatectomy for clinically localized prostate cancer. J Urol, 166: 416, 2001 Stamey, T. A.: Making the most out of six systematic sextant biopsies. Urology, 45: 2, 1995 Djavan, B., Ravery, V., Zlotta, A., Dobronski, P., Dobrovits, M., Fakhari, M. et al: Prospective evaluation of prostate cancer detected on biopsies 1, 2, 3 and 4: when should we stop? J Urol, 166: 1679, 2001 Naughton, C. K., Miller, D. C., Mager, D. E., Ornstein, D. K. and Catalona, W. J.: A prospective randomized trial comparing 6 versus 12 prostate needle biopsy cores: impact on cancer detection. J Urol, 164: 388, 2000 Kibel, A. S., Krithivas, K., Shamel, L. B., Kantoff, P. W. and DeWolf, W. C.: Constitutive expression of high levels of prostate-specific antigen in the absence of prostate carcinoma. Urology, 48: 741, 1996 Mikolajczyk, S. D., Millar, L. S., Wang, T. J., Rittenhouse, H. G., Wolfert, R. L., Marks, L. S. et al: “BPSA,” a specific molecular form of free prostate-specific antigen, is found predominantly in the transition zone of patients with nodular benign prostatic hyperplasia. Urology, 55: 41, 2000 Mikolajczyk, S. D., Marker, K. M., Millar, L. S., Kumar, A., Saedi, M. S., Payne, J. K. et al: A truncated precursor form of prostate-specific antigen is a more specific serum marker of prostate cancer. Cancer Res, 61: 6958, 2001