- Email: [email protected]
THE PERCENT OF BIOPSY CORES POSITIVE FOR CANCER IS A PREDICTOR OF ADVANCED PATHOLOGICAL STAGE AND POOR CLINICAL OUTCOMES IN PATIENTS TREATED WITH RADICAL PROSTATECTOMY
0022-5347/04/1716-2209/0 THE JOURNAL OF UROLOGY® Copyright © 2004 by AMERICAN UROLOGICAL ASSOCIATION
Vol. 171, 2209 –2214, June 2004 Printed in U.S.A.
DOI: 10.1097/01.ju.0000127730.78973.fe
THE PERCENT OF BIOPSY CORES POSITIVE FOR CANCER IS A PREDICTOR OF ADVANCED PATHOLOGICAL STAGE AND POOR CLINICAL OUTCOMES IN PATIENTS TREATED WITH RADICAL PROSTATECTOMY YAIR LOTAN,*,† SHAHROKH F. SHARIAT,* SEYED M. KHODDAMI, HOSSEIN SABOORIAN, KENNETH S. KOENEMAN, JEFFREY A. CADEDDU, ARTHUR I. SAGALOWSKY, JOHN D. MCCONNELL AND CLAUS G. ROEHRBORN From the Departments of Urology and Pathology (HS), University of Texas Southwestern Medical Center at Dallas, Dallas, Texas
ABSTRACT
Purpose: We examined if the percent of positive biopsies is associated with features of biologically aggressive prostate cancer, biochemical progression and development of distant metastases in patients undergoing radical prostatectomy (RP). Materials and Methods: Multivariate analyses of preoperative features in 605 consecutive patients who underwent RP for clinically localized disease were evaluated to determine the association between the percent positive biopsy cores (PosBx), pathological stage and grade, and biochemical progression following RP. The percent of PosBx cores was defined using the formula, (number of positive biopsy cores/total number of biopsy cores) ⫻ 100. Results: The mean number of biopsy cores and percent PosBx cores ⫾ SE was 8.8 ⫾ 6.0 and 31.4 ⫾ 21.1, respectively. Higher percent PosBx was significantly associated with higher preoperative prostate specific antigen (PSA), extracapsular extension, seminal vesicle invasion, positive surgical margins, higher final Gleason sum, lymphovascular invasion, perineural invasion and metastases to regional lymph nodes. On multivariate analyses adjusted for the effects of standard preoperative features percent PosBx was associated with nonorgan confined disease, seminal vesicle invasion and biochemical progression after surgery (p ⫽ 0.049, 0.050 and 0.006, respectively). Percent PosBx retained its independent association with PSA progression after adjustment for the effects of postoperative pathological features (p ⫽ 0.015). Higher percent PosBx was associated with shorter PSA doubling time after PSA progression, and an increased risk of distant metastases and overall mortality (p ⫽ 0.039, 0.001 and 0.018, respectively). Conclusions: Percent PosBx is associated with established pathological features, biochemical progression, distant metastases and overall death in patients who undergo RP for clinically localized disease. Percent PosBx should be included in preoperative predictive models for prognosticating outcomes after primary treatment and it may assist in selecting patients for inclusion in neoadjuvant and/or adjuvant therapy trials. KEY WORDS: prostate, prostatic neoplasms, biopsy, neoplasm metastasis
The ability to predict accurately the pathological features, biochemical progression, development of metastases and most importantly survival in patients treated for clinically localized disease would be useful for managing prostate cancer. Currently prostate specific antigen (PSA), biopsy Gleason score and clinical T stage are the most commonly used parameters to attempt to predict the risk of organ confined disease, lymph node involvement and PSA progression after treatment.1 Unfortunately 10% to 26% of patients with pathologically organ confined disease after radical prostatectomy experience clinical recurrences.2 Identifying patients in whom primary curative therapy is likely to fail would be helpful for selecting those best suited for clinical trials of early systemic intervention, sparing
men the morbidity associated with ineffective local therapy and/or predicting patient outcome.2, 3 To improve the preoperative prediction of patient stage groups have evaluated additional information available in initial prostate needle biopsies. Factors such as quantitative nuclear grade,4 total percent cancer in biopsy cores (PosBx),5 a combination of percent positive biopsy (greater than 50%), and the number of positive cores (NoPosBx)6 –9 have been found to improve the prediction of pathological stage. The percent cancer in a prostate biopsy is associated with the risk of extracapsular extension.10 The number of positive biopsies can improve the prediction of margin status11 and the linear extent of prostate cancer in biopsies is significantly related to tumor volume.12 However, the number of positive cores depends on the total number of cores taken, thus, limiting the wide applicability of threshold parameters. Several studies used the amount and distribution of Gleason grade 4 or 5 cancer in prostate biopsies to predict the risk of lymph node spread.13, 14 Furthermore, evaluating percent positive prostate biopsy cores has been shown to help predict the risk of PSA relapse15 and time to PSA failure after radical prostatectomy.16
Accepted for publication January 16, 2004. Study received institutional review board approval. Supported by the Austrian Program for Advanced Research and Technology. * Equal study contribution. † Correspondence: Department of Urology, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, Texas 75390-9110 (telephone: 214-648-0389; FAX: 214-648-8786; e-mail: [email protected]). 2209
2210
PERCENT POSITIVE CORES AND PROSTATE CANCER CLINICAL OUTCOMES
To confirm these findings and further assess the hypothesis that a proportion of men with biologically aggressive clinically localized prostate cancer have higher percent PosBx we investigated the association of percent PosBx with prostate cancer features and outcomes in a large cohort of consecutive patients with clinically localized prostate cancer who underwent radical prostatectomy.
MATERIALS AND METHODS
Patient population. All 630 patients admitted to 2 teaching hospitals who underwent radical retropubic prostatectomy with bilateral lymphadenectomy for clinically localized prostate cancer by surgeons at the department of urology at our institution during July 1994 through December 2002 were potential candidates for this analysis. The number of biopsy cores obtained and/or the number of positive biopsy cores was unavailable for 25 men, who were excluded from analyses. Serum total PSA was measured by the Hybritech Tandem-E assay (Hybritech, Inc., San Diego, California). Prostate biopsies from different regions of the prostate were submitted in separate containers and embedded in distinct cassettes. Percent PosBx was defined using the formula, NoPosBx/total number of biopsy cores (NoBx) ⫻ 100. Pathological examination. Staff pathologists at our institution examined all prostatectomy specimens pathologically, as previously described.17 Briefly, evaluation of the radical prostatectomy specimen was performed in accordance with the guidelines of the College of American Pathologists.18 Multiple, oriented quadrant sections from the entire prostate were processed (approximately 25 to 30 blocks). The 1997 American Joint Committee on Cancer/TNM system was used for pathological staging. Primary tumor grading was performed according to the Gleason system. To ensure outcome validity 2 clinicians read pathology reports while blinded to patient clinical parameters and the finding of the other reviewer. Interreader reliability measured using the intraclass correlation coefficient was greater than 0.95 for all pathological parameters. Postoperative followup. Patients generally were scheduled to have digital rectal examination and serum PSA evaluation postoperatively every 3 months for year 1, semiannually from years 2 through 5 and annually thereafter. Biochemical progression was defined as a sustained increase on 2 or more occasions of serum total PSA 0.2 ng/ml or greater and it was assigned to the date of the first value of 0.2 ng/ml or greater. Patients who received adjuvant radiation therapy before biochemical progression because of clinical or pathological characteristics were considered to have disease progression from the date of the first value of 0.2 ng/ml or greater. In patients who had biochemical progression postprogression serum PSA doubling time was calculated using the formula, PSA doubling time ⫽ ln (2) ⫻ T/[ln (final PSA)-ln (initial PSA)], where T represents time.19 All patients had at least 3 PSA measurements available after progression.
STATISTICAL ANALYSIS
Differences in NoBx, NoPosBx and percent PosBx between clinical and pathological characteristics were tested with the Mann-Whitney U test. Differences in NoBx, NoPosBx and percent PosBx across categorical variables with more than 2 categories were assessed using the Kruskal-Wallis test. Multiple comparisons were done using the Mann-Whitney U test when the overall test was significant at the 5% level. Spearman’s rank correlation coefficient was used to compare ordi-
nal and continuous variables. Logistic regression was used for multivariate analysis of binary outcome variables. The Kaplan-Meier method was used to calculate survival functions and differences were assessed with the log rank statistic. Multivariate survival analysis was performed with the Cox proportional hazard regression model. Preoperative PSA, NoBx, NoPosBx and percent PosBx had a skewed distribution and, therefore, they were modeled with a log transformation. Clinical stage was evaluated as T1 vs T2. Biopsy and radical prostatectomy Gleason sum were evaluated as grades 4 to 6 vs 7 vs 8 to 10. Statistical significance in this study was set as p ⱕ0.05. All reported p values are 2-sided. All analyses were performed with the statistical package SPSS, version 10.0 for Windows (SPSS, Inc., Chicago, Illinois). RESULTS
Association of NoBx, NoPosBx and percent PosBx with clinical and pathological characteristics. Table 1 lists the clinical and pathological characteristics of the 605 patients with radical prostatectomy, and associations with NoBx, NoPosBx and percent PosBx. The mean number of lymph nodes per patient ⫾ SE removed at radical prostatectomy was 7.5 ⫾ 5.3 (median 8). Mean patient age in this study was 60.4 ⫾ 6.7 years (median 60.9, range 40 to 75). Mean preoperative serum total PSA was 8.1 ⫾ 8.0 ng/ml (median 6.1, IQR 4.2, range 0.1 to 99.0). Mean NoBx was 8.8 ⫾ 6.0 (median 6.0, IQR 4.0, range 2 to 36); mean NoPosBx was 2.4 ⫾ 1.9 (median 2.0, IQR 2.0, range 1 to 15) and mean percent PosBx was 31.4% ⫾ 21.1% (median 25.0%, IQR 33.3%, range 2.8% to 100%). NoBx, NoPosBx and percent PosBx did not correlate with age at surgery ( ⫽ 0.001 and p ⫽ 0.841, ⫽ 0.007 and p ⫽ 0.871, and ⫽ 0.007 and p ⫽ 0.874, respectively). Higher NoBx, NoPosBx and percent PosBx were associated with clinical stage T2 and higher biopsy Gleason sum (p ⱕ0.033). However, only NoPosBx and percent PosBx were associated with extracapsular extension, seminal vesicle invasion, positive surgical margin status, higher final Gleason sum, lymphovascular invasion, perineural invasion and metastases to regional lymph nodes (p ⬍0.001). NoBx, NoPosBx and percent PosBx did not correlate with age at surgery (p ⫽ 0.841; 0.871 and 0.874). NoBx did not correlate with preoperative serum PSA (p ⫽ 0.324), whereas NoPosBx ( ⫽ 0.113, p ⫽ 0.006) and percent PosBx ( ⫽ 0.091, p ⫽ 0.049) did. On univariate logistic regression analyses higher NoPosBx and higher percent PosBx were associated with an increased risk of nonorgan confined disease (HR 0.444, 95% CI 0.335 to 0.589, p ⬍0.001 and HR 0.479, 95% CI 0.363 to 0.632, p ⬍0.001), seminal vesicle invasion (HR 2.384, 95% CI 1.563 to 3.637, p ⬍0.001 and HR 3.446, 95% CI 2.135 to 5.563, p ⬍0.001) and metastases to regional lymph nodes (HR 4.910, 95% CI 1.863 to 12.941, p ⫽ 0.001 and HR 6.382, 95% CI 1.902 to 21.414, p ⫽ 0.003, respectively). However, NoBx was not associated with nonorgan confined disease, seminal vesicle invasion or lymph node metastases (p ⫽ 0.064, 0.348 and 0.827, respectively). On preoperative multivariate analyses percent PosBx (p ⫽ 0.049 and 0.050), clinical stage (p ⫽ 0.004 and 0.010), preoperative PSA (p ⬍0.001) and biopsy Gleason sum (Ptrend ⫽ 0.034 and ⬍0.001) were associated with nonorgan confined disease and seminal vesicle invasion, respectively (table 2). On preoperative multivariate analysis adjusted for the effects of clinical stage, NoPosBx and percent PosBx only preoperative PSA and biopsy Gleason sum were associated with metastases to regional lymph nodes (p ⫽ 0.011 and Ptrend ⫽ 0.018, respectively). Association of NoBx, NoPosBx and percent PosBx with biochemical progression after radical prostatectomy. Overall only 13% of patients (78 of 605) had cancer progression at a
2211
PERCENT POSITIVE CORES AND PROSTATE CANCER CLINICAL OUTCOMES
TABLE 1. NoBx, NoPosBx and percent PosBx vs clinical and pathological features in 605 consecutive patients who underwent radical retropubic prostatectomy for clinically localized disease No. Pts (%)
Median NoBx (IQR)
408 (67) 197 (33)
6.0 (4.0) 8.0 (2.0)
399 (66) 168 (28) 36 (6)
Clinical stage: T1 T2 Biopsy Gleason sum:† 4–6 7 8–10 Extraprostatic extension only: Neg Pos Seminal vesicle involvement: Neg Pos Surgical margin status:‡ Neg Pos Final Gleason sum: 4–6 7 8–10 Lympho-vascular invasion: Neg Pos Perineural invasion: Neg Pos Regional lymph node metastases: Neg Pos * p ⬍0.001. † Not available in 2 patients. ‡ Not available on 1 patient.
Median NoPosBx (IQR)*
Median % PosBx (IQR)*
⬍0.001 (Mann-Whitney U test)
2.0 (2.0) 3.0 (3.0)
17.2 (16.7) 33.3 (32.8)
6.0 (3.0) 7.0 (4.0) 6.0 (2.0)
0.033 (Kruskal-Wallis test)
1.0 (2.0) 3.0 (2.0) 3.0 (1.0)
16.7 (16.7) 33.3 (31.7) 50.0 (40.8)
437 (72) 168 (28)
6.0 (4.0) 6.0 (4.0)
0.344 (Mann-Whitney U test)
2.0 (2.0) 3.0 (3.0)
20.0 (16.7) 33.3 (33.3)
549 (91) 55 (9)
6.0 (4.0) 6.0 (2.0)
0.132 (Mann-Whitney U test)
2.0 (2.0) 3.0 (2.0)
22.2 (23.3) 50.0 (32.1)
435 (72) 170 (28)
6.0 (3.0) 7.0 (4.0)
0.233 (Mann-Whitney U test)
2.0 (2.0) 3.0 (3.0)
22.2 (20.8) 33.3 (33.3)
245 (41) 300 (50) 60 (10)
6.0 (4.0) 6.0 (4.0) 6.0 (2.0)
0.693 (Kruskal-Wallis test)
1.0 (1.0) 2.0 (2.0) 3.0 (3.0)
16.7 (16.7) 28.6 (33.3) 38.8 (35.1)
573 (95) 32 (5)
6.0 (4.0) 7.0 (4.0)
0.736 (Mann-Whitney U test)
2.0 (2.0) 4.0 (3.0)
23.1 (25.6) 50.0 (30.3)
240 (40) 365 (60)
6.0 (3.0) 6.0 (4.0) 1.0 (1.0) 6.0 (4.0) 8.0 (3.0)
0.677 (Mann-Whitney U test)
1.0 (1.0) 3.0 (2.0)
16.7 (17.4) 33.3 (33.3)
0.408 (Mann-Whitney U test)
2.0 (2.0) 4.0 (2.0)
25.0 (28.8) 50.0 (57.6)
595 (98) 10 (2)
p Value
TABLE 2. Multivariate logistic regression analyses of preoperative features for prediction of organ confined disease and seminal vesicle invasion in 605 patients who underwent radical retropubic prostatectomy for clinically localized prostate cancer Organ Confined Disease
Preop serum PSA* NoPosBx % PosBx Clinical stage Biopsy Gleason sum:† 4–6 7 8–10 * Logarithmically transformed. † Not available on 2 patients.
Seminal Vesicle Invasion
HR
95% CI
p Value
HR
95% CI
p Value
0.433 0.894 0.641 0.546
0.316–0.594 0.532–1.355 0.406–0.999 0.363–0.823
⬍0.001 0.493 0.049 0.004 0.034
2.311 0.750 2.265 2.364
1.493–3.577 0.324–1.738 1.000–5.327 1.227–4.553
⬍0.001 0.502 0.050 0.010 ⬍0.001
1.000 1.094 0.405
Referent 0.699–1.713 0.182–0.902
1.000 1.459 6.693
Referent 0.691–3.081 2.642–16.950
median postoperative followup of 21.2 months (range 1.0 to 101.3). Overall PSA progression-free survival was 81.2% ⫾ 1.1% at 3 years and 77.6% ⫾ 1.4% at 5 years. Using the log rank test we found that the 6 patients with NoBx above the median did not have a different probability of PSA progression than those below the median (p ⫽ 0.2548). In contrast, patients with NoPosBx or percent PosBx above the median (2 and 25%, respectively) had an increased probability of PSA progression (p ⫽ 0.0002 and p ⬍0.0001, respectively, fig. 1). On univariate Cox proportional hazards regression analyses all standard preoperative and postoperative features, higher NoPosBx and higher percent PosBx were significantly associated with an increased risk of rapid biochemical progression after surgery (table 3). NoBx was not associated with PSA progression after radical prostatectomy on multivariate analysis (p ⫽ 0.052). On preoperative multivariate analysis only preoperative PSA, percent PosBx and biopsy Gleason sum were associated with biochemical progression (p ⬍0.001, 0.006, and Ptrend ⬍0.001, respectively, table 3). On postoperative multivariate analysis only preoperative PSA, percent PosBx, seminal vesicle invasion, lymph node metastases and final Gleason sum were associated with PSA progression after surgery (p ⬍0.001, 0.015, 0.025 and 0.026, and Ptrend ⫽ 0.013, respectively, table 3).
0.695 0.027
0.322 ⬍0.001
Association of NoBx, NoPosBx and percent PosBx with PSA doubling time and response to salvage radiation therapy. Seven patients did not undergo additional treatment after disease progression at last followup, 21 underwent local salvage radiation therapy at our institution and 43 underwent hormonal therapy as primary treatment for prostate cancer progression. Seven patients were lost to followup. PSA doubling time calculation was performed for 68 patients who had PSA progression prior to the administration of salvage therapy. PSA doubling time between radical prostatectomy failure and the initiation of salvage therapy inversely correlated with NoPosBx and percent PosBx ( ⫽ ⫺0.349, p ⫽ 0.044 and ⫽ ⫺0.396, p ⫽ 0.039, respectively) but not NoBx ( ⫽ ⫺0.79, p ⫽ 0.527). Median time from the start of salvage local radiation therapy to last followup in patients who had a favorable response was 43.7 months (range 4.4 to 71.1). Overall PSA progression-free survival after salvage local radiation therapy was 45% ⫾ 11% at 2 years and 20% ⫾ 10% at 4 years. On univariate Cox proportional hazards regression analyses NoBx, NoPosBx and percent PosBx were not associated with biochemical progression after salvage radiation therapy (p ⫽ 0.187, 0.173 and 0.076, respectively). Association of NoBx, NoPosBx and percent PosBx with
2212
PERCENT POSITIVE CORES AND PROSTATE CANCER CLINICAL OUTCOMES
proportional hazards regression analyses higher NoPosBx (HR 3.749, 95% CI 1.202 to 11.688, p ⫽ 0.023) and higher percent PosBx (HR 5.746, 95% CI 1.351 to 24.439, p ⫽ 0.018) were associated with overall survival after surgery but NoBx was not (p ⫽ 0.576). DISCUSSION
FIG. 1. Kaplan-Meier estimates of PSA progression-free probability in 605 patients with clinically localized prostate cancer treated with radical prostatectomy, as stratified into groups above or below median. A, total number of positive biopsy cores (2). B, 25% positive biopsy cores.
clinically evident distant prostate cancer metastases and patient survival. Staging evaluation including bone scan or ProstaScint scan (Cytogen Corp., Princeton, New Jersey) scan was performed at our institution in 56 patients who had biochemical progression after radical prostatectomy. A total of 13 patients (2%) had bone metastases, as evidenced by imaging. Median followup in patients who experienced PSA progression after surgery but did not have metastases was 42.0 months (range 4 to 100). Using the log rank test patients with NoPosBx or percent PosBx above the median (2 and 25%, respectively) had an increased probability of distant metastasis (p ⫽ 0.0331 and 0.0054, respectively, fig. 2). The 6 patients with NoBx above the median did not have a different probability of distant metastases than those with NoBx below the median (p ⫽ 0.9432). On univariate Cox proportional hazards regression analyses higher NoPosBx (HR 3.307, 95% CI 1.450 to 7.542, p ⫽ 0.004) and higher percent PosBx (HR 5.850, 95% CI 2.036 to 16.815, p ⫽ 0.001) were associated with an increased risk of distant metastases after surgery but NoBx was not (p ⫽ 0.512). Overall survival data were available on 629 of the 630 patients. Seven of the 629 patients (1%) were dead of any cause at analysis. Median followup in patients who experienced PSA progression after surgery and were alive at last followup was 39.3 months (range 4 to 97). On univariate Cox
Knowledge of the true extent of prostate cancer is an important factor for determining appropriate treatment. The ability to determine accurately if a cancer is organ confined may help differentiate patients who can undergo nerve sparing radical prostatectomy and patients who are best candidates for brachytherapy or external beam radiation therapy. Furthermore, predicting which patients are at high risk for PSA failure may assist in selecting those who may benefit from neoadjuvant and/or adjuvant therapy. Current nomograms using preoperative serum PSA, clinical stage and Gleason score are insufficiently accurate to predict the patients in whom local therapy may fail.1 Even with pathologically confined disease 10% to 26% of patients experience clinical recurrences.2 In the current study we confirmed the strong association of percent PosBx with established features of biologically aggressive prostate cancer, such as tumor stage and histological differentiation, biochemical progression, distant metastases and overall mortality. On preoperative and postoperative multivariate analyses percent positive biopsy cores predicted an increased risk of nonorgan confined disease and seminal vesical invasion, and a higher risk of biochemical progression. The number and percent of positive biopsies were predictive of PSA failure and metastases on Kaplan-Meier analyses. Multiple studies have evaluated the association between initial needle biopsy characteristics and postoperative pathological features. Veltri et al found that quantitative nuclear grade features in prostate biopsy specimens were the strongest predictor of pathological stage in patients with clinical T1c when combined with biopsy Gleason score and complexed PSA density.4 D’Amico et al evaluated the predictive value of percent positive prostate biopsy in 960 patients and found a statistically significant correlation between increased percent positive biopsies, and increased extracapsular extension and seminal vesical invasion (p ⬍0.0001).20 Several studies assessed the area or length of cancer in prostate biopsies. Grossklaus et al found a correlation between percent tumor in the biopsy set and the risk of extracapsular extension (p ⬍0.01).10 Bismar et al found that total percent cancer in biopsy cores was significantly related to pathological T stage on multivariate analysis (p ⫽ 0.003).5 The linear extent of prostate cancer in biopsies is also significantly related to tumor volume.12 The disadvantage of measuring tumor area in relation to total biopsy core area is that it is quite labor intensive. The number of positive cores has been found to correlate with stage in multiple studies,6 –9 including ours. In patients with greater than 2 positive cores Kaplan-Meier curves demonstrated a decreased risk of PSA progression-free probability (p ⫽ 0.0002) and metastasis-free probability (p ⫽ 0.033, figs. 1 and 2). Because urologists are sampling a varying number of biopsy cores, it may be more difficult to use the number of positive cores in a uniform matter. The percent positive cores offers a more uniform parameter for comparing different biopsy strategies. As such, it is important to separately embed specimens to be able to assess accurately this percent. Beyond its usefulness for determining pathological stage information from needle biopsies has also been evaluated for predicting metastases to lymph nodes and biochemical progression after treatment with curative intent. Several studies used the amount and distribution of Gleason grade 4 or 5
2213
PERCENT POSITIVE CORES AND PROSTATE CANCER CLINICAL OUTCOMES
TABLE 3. Univariate and multivariate cox regression analyses of preoperative and postoperative features for prediction of biochemical progression in 605 patients who underwent radical retropubic prostatectomy for clinically localized prostate cancer Univariate
Preoperative serum PSA* NoPosBx % PosBx Clinical stage Biopsy Gleason sum:† 4–6 7 8–10 Extracapsular extension Seminal vesicle involvement‡ Surgical margin status Lymph node metastasis Final Gleason sum: 4–6 7 8–10 * Logarithmically transformed. † Not available on 2 patients. ‡ Not available on 1 patient.
Preop Multivariate
Postop Multivariate
HR
95% CI
p Value
HR
95% CI
p Value
HR
95% CI
p Value
3.830 1.786 2.486 2.863
2.804–5.230 1.284–2.485 1.727–3.579 1.821–4.501
⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001
3.039 0.521 2.591 1.586
2.198–4.202 0.264–1.025 1.306–5.141 0.953–2.639
⬍0.001 0.059 0.006 0.076 ⬍0.001
1.000 1.789 7.825 4.715 6.859 3.322 19.441
Referent 1.060–3.017 4.500–13.606 2.971–7.483 4.320–10.890 2.130–5.180 9.516–39.721
1.000 1.039 3.239 –
Referent 0.576–1.872 1.623–6.466 –
0.899 ⬍0.001 – – – – – – –
⬍0.001 0.066 0.015 – – – – – 0.092 0.025 0.127 0.026 0.013
Referent 1.165–4.161 6.004–21.967
– – – – – –
1.616–3.223 0.291–1.039 1.172–4.433 – – – – – 0.918–3.103 1.089–3.611 0.891–2.537 1.069–6.639
1.000 2.201 11.484
– – – – – –
2.282 0.550 2.279 – – – – – 1.687 1.983 1.503 2.888 1.000 0.944 3.226
Referent 0.461–1.933 1.068–7.122
0.029 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 0.015 ⬍0.001
FIG. 2. Kaplan-Meier estimates of metastasis-free probability in 605 patients with clinically localized prostate cancer treated with radical prostatectomy, as stratified into groups above or below median. A, total number of positive biopsy cores (2). B, 25% positive biopsy cores.
cancer in prostate biopsies to predict the risk of lymph node spread.13 As in our study and others, percent positive biopsies is a powerful predictor of biochemical progression. Presti
0.874 0.010
et al evaluated percent positive prostate biopsies for predicting the risk of relapse after radical prostatectomy in 109 patients.15 Using Kaplan-Meier curves they noted a statistically significant decrease in disease-free survival (p ⫽ 0.03) in patients with greater than 50% positive biopsies. Risk stratification using percent positive biopsies in conjunction with biopsy Gleason score, preoperative PSA and clinical stage16 or endorectal coil magnetic resonance imaging21 allows the identification of patients at increased risk for early PSA failure. This information was particularly useful for predicting failure in patients in an intermediate risk group (1992 American Joint Committee on Cancer T2b, biopsy Gleason sum 7 or PSA greater than 10 ng/ml and 20 ng/ml or less).16 Early PSA failure is an important end point because it predicts future poor clinical outcomes.22 Due to the important association of pathological features for predicting pathological stage and progression several studies have been performed to use these factors in predictive models. Veltri et al included multiple sextant pathology variables in a neural net and were able to predict organ confined status with greater than 90% accuracy.23 Ziada et al assessed the impact of different pretreatment variables on outcome in patients with clinically confined prostate cancer using an artificial neural network.24 Percent positive biopsy tissue improved the prediction of pathological stage and biochemical failure. The ability to enhance the prediction of stage may be of particular significance in patients undergoing radiation therapy. These patients do not have the benefit of pathological staging to assist in posttreatment clinical decision making. An improved prediction of patients at high risk for early PSA progression would allow the identification of patients for clinical trials of adjuvant therapy and possibly improved survival. Several limitations in this study should be considered. Followup in this study was limited. Some variables that were inconclusive due to limited statistical power may attain statistical significance if sample size, followup or the proportion of patients with biochemical failure or clinical progression were greater. In addition, the population in this study underwent radical prostatectomy and bilateral pelvic lymphadenectomy by multiple surgeons and specimens were evaluated by multiple pathologists. Most studies evaluating NoBx, NoPosBx and percent PosBx used cohorts operated on by a single surgeon that were assembled from populations at large academic hospitals serving as tertiary referral centers. While prognostic factors may perform well in the select group of patients treated at these centers, it remains to be determined whether they are applicable to the greater population
2214
PERCENT POSITIVE CORES AND PROSTATE CANCER CLINICAL OUTCOMES
of patients with prostate cancer. Similarly while it may be preferable for a single pathologist specialized in genitourinary pathology to review each prostatectomy specimen, the setup at our institution reflects real-world practice, in which local pathologists review tissue specimens and their interpretation is then used in clinical decision making with the patient. CONCLUSIONS
Percent positive prostate biopsy cores is a strong predictor of adverse pathological features, biochemical progression, distant metastases and overall survival in patients who undergo radical prostatectomy for clinically localized disease. Percent positive prostate biopsy cores should be included in preoperative predictive models for prognosticating outcomes after primary treatment and selecting patients for inclusion in neoadjuvant and/or adjuvant therapy trials. REFERENCES
1. Partin, A. W., Kattan, M. W., Subong, E. N., Walsh, P. C., Wojno, K. J., Oesterling, J. E. et al: Combination of prostate-specific antigen, clinical stage, and Gleason score to predict pathological stage of localized prostate cancer. A multi-institutional update. JAMA, 277: 1445, 1997 2. Lerner, S. E., Blute, M. L., Bergstralh, E. J., Bostwick, D. G., Eickholt, J. T. and Zincke, H.: Analysis of risk factors for progression in patients with pathologically confined prostate cancers after radical retropubic prostatectomy. J Urol, 156: 137, 1996 3. D’Amico, A. V., Whittington, R., Malkowicz, S. B., Schultz, D., Tomaszewski, J. E. and Wein, A.: Prostate-specific antigen failure despite pathologically organ-confined and marginnegative prostate cancer: the basis for an adjuvant therapy trial. J Clin Oncol, 15: 1465, 1997 4. Veltri, R. W., Miller, M. C., Mangold, L. A., O’Dowd, G. J., Epstein, J. I. and Partin, A. W.: Prediction of pathological stage in patients with clinical stage T1c prostate cancer: the new challenge. J Urol, 168: 100, 2002 5. Bismar, T. A., Lewis, J. S., Jr., Vollmer, R. T. and Humphrey, P. A.: Multiple measures of carcinoma extent versus perineural invasion in prostate needle biopsy tissue in prediction of pathologic stage in a screening population. Am J Surg Pathol, 27: 432, 2003 6. Wills, M. L., Sauvageot, J., Partin, A. W., Gurganus, R. and Epstein, J. I.: Ability of sextant biopsies to predict radical prostatectomy stage. Urology, 51: 759, 1998 7. Borirakchanyavat, S., Bhargava, V., Shinohara, K., Toke, A., Carroll, P. R. and Presti, J. C., Jr.: Systematic sextant biopsies in the prediction of extracapsular extension at radical prostatectomy. Urology, 50: 373, 1997 8. Peller, P. A., Young, D. C., Marmaduke, D. P., Marsh, W. L. and Badalament, R. A.: Sextant prostate biopsies. A histopathologic correlation with radical prostatectomy specimens. Cancer, 75: 530, 1995 9. Huland, H., Hammerer, P., Henke, R.-P. and Huland, E.: Preoperative prediction of tumor heterogeneity and recurrence after radical prostatectomy for localized prostatic carcinoma with digital rectal examination, prostate specific antigen and the results of 6 systematic biopsies. J Urol, 155: 1344, 1996 10. Grossklaus, D. J., Coffey, C. S., Shappell, S. B., Jack, G. S., Chang, S. S. and Cookson, M. S.: Percent of cancer in the biopsy set predicts pathological findings after prostatectomy. J Urol, 167: 2032, 2002
11. Tigrani, V. S., Bhargava, V., Shinohara, K. and Presti, J. C., Jr.: Number of positive systematic sextant biopsies predicts surgical margin status at radical prostatectomy. Urology, 54: 689, 1999 12. Lewis, J. S., Jr., Vollmer, R. T. and Humphrey, P. A.: Carcinoma extent in prostate needle biopsy tissue in the prediction of whole gland tumor volume in a screening population. Am J Clin Pathol, 118: 442, 2002 13. Conrad, S., Graefen, M., Pichlmeier, U., Henke, R.-P., Hammerer, P. G. and Huland, H.: Systematic sextant biopsies improve preoperative prediction of pelvic lymph node metastases in patients with clinically localized prostatic carcinoma. J Urol, 159: 2023, 1998 14. Haese, A., Chaudhari, M., Miller, M. C., Epstein, J. I., Huland, H., Palisaar, J. et al: Quantitative biopsy pathology for the prediction of pathologically organ-confined prostate carcinoma: a multiinstitutional validation study. Cancer, 97: 969, 2003 15. Presti, J. C., Jr., Shinohara, K., Bacchetti, P., Tigrani, V. and Bhargava, V.: Positive fraction of systematic biopsies predicts risk of relapse after radical prostatectomy. Urology, 52: 1079, 1998 16. D’Amico, A. V., Whittington, R., Malkowicz, S. B., Schultz, D., Fondurulia, J., Chen, M. H. et al: Clinical utility of the percentage of positive prostate biopsies in defining biochemical outcome after radical prostatectomy for patients with clinically localized prostate cancer. J Clin Oncol, 18: 1164, 2000 17. Shariat, S. F., Khoddami, S. M., Saboorian, H., Koeneman, K. S., Sagalowsky, A. I., Cadeddu, J. A. et al: Lymphovascular invasion is a pathologic feature of biologically aggressive disease in patients treated with radical prostatectomy. J Urol, 171: 1122, 2004 18. Henson, D. E., Hutter, R. V. and Farrow, G.: Practice protocol for the examination of specimens removed from patients with carcinoma of the prostate gland. A publication of the Cancer Committee, College of American Pathologists. Task Force on the Examination of Specimens Removed From Patients With Prostate Cancer. Arch Pathol Lab Med, 118: 779, 1994 19. Schmid, H. P., McNeal, J. E. and Stamey, T. A.: Observations on the doubling time of prostate cancer. The use of serial prostate-specific antigen in patients with untreated disease as a measure of increasing cancer volume. Cancer, 71: 2031, 1993 20. D’Amico, A. V., Whittington, R., Malkowicz, S. B., Schultz, D., Silver, B., Henry, L. et al: Investigating the clinical utility of the percent of positive prostate biopsies in predicting PSA outcome following local therapy for patients with clinically localized prostate cancer. Prostate Cancer Prostatic Dis, 3: 259, 2000 21. D’Amico, A. V., Whittington, R., Malkowicz, S. B., Wu, Y. H., Chen, M., Art, M. et al: Combination of the preoperative PSA level, biopsy gleason score, percentage of positive biopsies, and MRI T-stage to predict early PSA failure in men with clinically localized prostate cancer. Urology, 55: 572, 2000 22. Pound, C. R., Partin, A. W., Eisenberger, M. A., Chan, D. W., Pearson, J. D. and Walsh, P. C.: Natural history of progression after PSA elevation following radical prostatectomy. JAMA, 281: 1591, 1999 23. Veltri, R. W., Chaudhari, M., Miller, M. C., Poole, E. C., O’Dowd, G. J. and Partin, A. W.: Comparison of logistic regression and neural net modeling for prediction of prostate cancer pathologic stage. Clin Chem, 48: 1828, 2002 24. Ziada, A. M., Lisle, T. C., Snow, P. B., Levine, R. F., Miller, G. and Crawford, E. D.: Impact of different variables on the outcome of patients with clinically confined prostate carcinoma: prediction of pathologic stage and biochemical failure using an artificial neural network. Cancer, 91: 1653, 2001
Vol. 171, 2209 –2214, June 2004 Printed in U.S.A.
DOI: 10.1097/01.ju.0000127730.78973.fe
THE PERCENT OF BIOPSY CORES POSITIVE FOR CANCER IS A PREDICTOR OF ADVANCED PATHOLOGICAL STAGE AND POOR CLINICAL OUTCOMES IN PATIENTS TREATED WITH RADICAL PROSTATECTOMY YAIR LOTAN,*,† SHAHROKH F. SHARIAT,* SEYED M. KHODDAMI, HOSSEIN SABOORIAN, KENNETH S. KOENEMAN, JEFFREY A. CADEDDU, ARTHUR I. SAGALOWSKY, JOHN D. MCCONNELL AND CLAUS G. ROEHRBORN From the Departments of Urology and Pathology (HS), University of Texas Southwestern Medical Center at Dallas, Dallas, Texas
ABSTRACT
Purpose: We examined if the percent of positive biopsies is associated with features of biologically aggressive prostate cancer, biochemical progression and development of distant metastases in patients undergoing radical prostatectomy (RP). Materials and Methods: Multivariate analyses of preoperative features in 605 consecutive patients who underwent RP for clinically localized disease were evaluated to determine the association between the percent positive biopsy cores (PosBx), pathological stage and grade, and biochemical progression following RP. The percent of PosBx cores was defined using the formula, (number of positive biopsy cores/total number of biopsy cores) ⫻ 100. Results: The mean number of biopsy cores and percent PosBx cores ⫾ SE was 8.8 ⫾ 6.0 and 31.4 ⫾ 21.1, respectively. Higher percent PosBx was significantly associated with higher preoperative prostate specific antigen (PSA), extracapsular extension, seminal vesicle invasion, positive surgical margins, higher final Gleason sum, lymphovascular invasion, perineural invasion and metastases to regional lymph nodes. On multivariate analyses adjusted for the effects of standard preoperative features percent PosBx was associated with nonorgan confined disease, seminal vesicle invasion and biochemical progression after surgery (p ⫽ 0.049, 0.050 and 0.006, respectively). Percent PosBx retained its independent association with PSA progression after adjustment for the effects of postoperative pathological features (p ⫽ 0.015). Higher percent PosBx was associated with shorter PSA doubling time after PSA progression, and an increased risk of distant metastases and overall mortality (p ⫽ 0.039, 0.001 and 0.018, respectively). Conclusions: Percent PosBx is associated with established pathological features, biochemical progression, distant metastases and overall death in patients who undergo RP for clinically localized disease. Percent PosBx should be included in preoperative predictive models for prognosticating outcomes after primary treatment and it may assist in selecting patients for inclusion in neoadjuvant and/or adjuvant therapy trials. KEY WORDS: prostate, prostatic neoplasms, biopsy, neoplasm metastasis
The ability to predict accurately the pathological features, biochemical progression, development of metastases and most importantly survival in patients treated for clinically localized disease would be useful for managing prostate cancer. Currently prostate specific antigen (PSA), biopsy Gleason score and clinical T stage are the most commonly used parameters to attempt to predict the risk of organ confined disease, lymph node involvement and PSA progression after treatment.1 Unfortunately 10% to 26% of patients with pathologically organ confined disease after radical prostatectomy experience clinical recurrences.2 Identifying patients in whom primary curative therapy is likely to fail would be helpful for selecting those best suited for clinical trials of early systemic intervention, sparing
men the morbidity associated with ineffective local therapy and/or predicting patient outcome.2, 3 To improve the preoperative prediction of patient stage groups have evaluated additional information available in initial prostate needle biopsies. Factors such as quantitative nuclear grade,4 total percent cancer in biopsy cores (PosBx),5 a combination of percent positive biopsy (greater than 50%), and the number of positive cores (NoPosBx)6 –9 have been found to improve the prediction of pathological stage. The percent cancer in a prostate biopsy is associated with the risk of extracapsular extension.10 The number of positive biopsies can improve the prediction of margin status11 and the linear extent of prostate cancer in biopsies is significantly related to tumor volume.12 However, the number of positive cores depends on the total number of cores taken, thus, limiting the wide applicability of threshold parameters. Several studies used the amount and distribution of Gleason grade 4 or 5 cancer in prostate biopsies to predict the risk of lymph node spread.13, 14 Furthermore, evaluating percent positive prostate biopsy cores has been shown to help predict the risk of PSA relapse15 and time to PSA failure after radical prostatectomy.16
Accepted for publication January 16, 2004. Study received institutional review board approval. Supported by the Austrian Program for Advanced Research and Technology. * Equal study contribution. † Correspondence: Department of Urology, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, Texas 75390-9110 (telephone: 214-648-0389; FAX: 214-648-8786; e-mail: [email protected]). 2209
2210
PERCENT POSITIVE CORES AND PROSTATE CANCER CLINICAL OUTCOMES
To confirm these findings and further assess the hypothesis that a proportion of men with biologically aggressive clinically localized prostate cancer have higher percent PosBx we investigated the association of percent PosBx with prostate cancer features and outcomes in a large cohort of consecutive patients with clinically localized prostate cancer who underwent radical prostatectomy.
MATERIALS AND METHODS
Patient population. All 630 patients admitted to 2 teaching hospitals who underwent radical retropubic prostatectomy with bilateral lymphadenectomy for clinically localized prostate cancer by surgeons at the department of urology at our institution during July 1994 through December 2002 were potential candidates for this analysis. The number of biopsy cores obtained and/or the number of positive biopsy cores was unavailable for 25 men, who were excluded from analyses. Serum total PSA was measured by the Hybritech Tandem-E assay (Hybritech, Inc., San Diego, California). Prostate biopsies from different regions of the prostate were submitted in separate containers and embedded in distinct cassettes. Percent PosBx was defined using the formula, NoPosBx/total number of biopsy cores (NoBx) ⫻ 100. Pathological examination. Staff pathologists at our institution examined all prostatectomy specimens pathologically, as previously described.17 Briefly, evaluation of the radical prostatectomy specimen was performed in accordance with the guidelines of the College of American Pathologists.18 Multiple, oriented quadrant sections from the entire prostate were processed (approximately 25 to 30 blocks). The 1997 American Joint Committee on Cancer/TNM system was used for pathological staging. Primary tumor grading was performed according to the Gleason system. To ensure outcome validity 2 clinicians read pathology reports while blinded to patient clinical parameters and the finding of the other reviewer. Interreader reliability measured using the intraclass correlation coefficient was greater than 0.95 for all pathological parameters. Postoperative followup. Patients generally were scheduled to have digital rectal examination and serum PSA evaluation postoperatively every 3 months for year 1, semiannually from years 2 through 5 and annually thereafter. Biochemical progression was defined as a sustained increase on 2 or more occasions of serum total PSA 0.2 ng/ml or greater and it was assigned to the date of the first value of 0.2 ng/ml or greater. Patients who received adjuvant radiation therapy before biochemical progression because of clinical or pathological characteristics were considered to have disease progression from the date of the first value of 0.2 ng/ml or greater. In patients who had biochemical progression postprogression serum PSA doubling time was calculated using the formula, PSA doubling time ⫽ ln (2) ⫻ T/[ln (final PSA)-ln (initial PSA)], where T represents time.19 All patients had at least 3 PSA measurements available after progression.
STATISTICAL ANALYSIS
Differences in NoBx, NoPosBx and percent PosBx between clinical and pathological characteristics were tested with the Mann-Whitney U test. Differences in NoBx, NoPosBx and percent PosBx across categorical variables with more than 2 categories were assessed using the Kruskal-Wallis test. Multiple comparisons were done using the Mann-Whitney U test when the overall test was significant at the 5% level. Spearman’s rank correlation coefficient was used to compare ordi-
nal and continuous variables. Logistic regression was used for multivariate analysis of binary outcome variables. The Kaplan-Meier method was used to calculate survival functions and differences were assessed with the log rank statistic. Multivariate survival analysis was performed with the Cox proportional hazard regression model. Preoperative PSA, NoBx, NoPosBx and percent PosBx had a skewed distribution and, therefore, they were modeled with a log transformation. Clinical stage was evaluated as T1 vs T2. Biopsy and radical prostatectomy Gleason sum were evaluated as grades 4 to 6 vs 7 vs 8 to 10. Statistical significance in this study was set as p ⱕ0.05. All reported p values are 2-sided. All analyses were performed with the statistical package SPSS, version 10.0 for Windows (SPSS, Inc., Chicago, Illinois). RESULTS
Association of NoBx, NoPosBx and percent PosBx with clinical and pathological characteristics. Table 1 lists the clinical and pathological characteristics of the 605 patients with radical prostatectomy, and associations with NoBx, NoPosBx and percent PosBx. The mean number of lymph nodes per patient ⫾ SE removed at radical prostatectomy was 7.5 ⫾ 5.3 (median 8). Mean patient age in this study was 60.4 ⫾ 6.7 years (median 60.9, range 40 to 75). Mean preoperative serum total PSA was 8.1 ⫾ 8.0 ng/ml (median 6.1, IQR 4.2, range 0.1 to 99.0). Mean NoBx was 8.8 ⫾ 6.0 (median 6.0, IQR 4.0, range 2 to 36); mean NoPosBx was 2.4 ⫾ 1.9 (median 2.0, IQR 2.0, range 1 to 15) and mean percent PosBx was 31.4% ⫾ 21.1% (median 25.0%, IQR 33.3%, range 2.8% to 100%). NoBx, NoPosBx and percent PosBx did not correlate with age at surgery ( ⫽ 0.001 and p ⫽ 0.841, ⫽ 0.007 and p ⫽ 0.871, and ⫽ 0.007 and p ⫽ 0.874, respectively). Higher NoBx, NoPosBx and percent PosBx were associated with clinical stage T2 and higher biopsy Gleason sum (p ⱕ0.033). However, only NoPosBx and percent PosBx were associated with extracapsular extension, seminal vesicle invasion, positive surgical margin status, higher final Gleason sum, lymphovascular invasion, perineural invasion and metastases to regional lymph nodes (p ⬍0.001). NoBx, NoPosBx and percent PosBx did not correlate with age at surgery (p ⫽ 0.841; 0.871 and 0.874). NoBx did not correlate with preoperative serum PSA (p ⫽ 0.324), whereas NoPosBx ( ⫽ 0.113, p ⫽ 0.006) and percent PosBx ( ⫽ 0.091, p ⫽ 0.049) did. On univariate logistic regression analyses higher NoPosBx and higher percent PosBx were associated with an increased risk of nonorgan confined disease (HR 0.444, 95% CI 0.335 to 0.589, p ⬍0.001 and HR 0.479, 95% CI 0.363 to 0.632, p ⬍0.001), seminal vesicle invasion (HR 2.384, 95% CI 1.563 to 3.637, p ⬍0.001 and HR 3.446, 95% CI 2.135 to 5.563, p ⬍0.001) and metastases to regional lymph nodes (HR 4.910, 95% CI 1.863 to 12.941, p ⫽ 0.001 and HR 6.382, 95% CI 1.902 to 21.414, p ⫽ 0.003, respectively). However, NoBx was not associated with nonorgan confined disease, seminal vesicle invasion or lymph node metastases (p ⫽ 0.064, 0.348 and 0.827, respectively). On preoperative multivariate analyses percent PosBx (p ⫽ 0.049 and 0.050), clinical stage (p ⫽ 0.004 and 0.010), preoperative PSA (p ⬍0.001) and biopsy Gleason sum (Ptrend ⫽ 0.034 and ⬍0.001) were associated with nonorgan confined disease and seminal vesicle invasion, respectively (table 2). On preoperative multivariate analysis adjusted for the effects of clinical stage, NoPosBx and percent PosBx only preoperative PSA and biopsy Gleason sum were associated with metastases to regional lymph nodes (p ⫽ 0.011 and Ptrend ⫽ 0.018, respectively). Association of NoBx, NoPosBx and percent PosBx with biochemical progression after radical prostatectomy. Overall only 13% of patients (78 of 605) had cancer progression at a
2211
PERCENT POSITIVE CORES AND PROSTATE CANCER CLINICAL OUTCOMES
TABLE 1. NoBx, NoPosBx and percent PosBx vs clinical and pathological features in 605 consecutive patients who underwent radical retropubic prostatectomy for clinically localized disease No. Pts (%)
Median NoBx (IQR)
408 (67) 197 (33)
6.0 (4.0) 8.0 (2.0)
399 (66) 168 (28) 36 (6)
Clinical stage: T1 T2 Biopsy Gleason sum:† 4–6 7 8–10 Extraprostatic extension only: Neg Pos Seminal vesicle involvement: Neg Pos Surgical margin status:‡ Neg Pos Final Gleason sum: 4–6 7 8–10 Lympho-vascular invasion: Neg Pos Perineural invasion: Neg Pos Regional lymph node metastases: Neg Pos * p ⬍0.001. † Not available in 2 patients. ‡ Not available on 1 patient.
Median NoPosBx (IQR)*
Median % PosBx (IQR)*
⬍0.001 (Mann-Whitney U test)
2.0 (2.0) 3.0 (3.0)
17.2 (16.7) 33.3 (32.8)
6.0 (3.0) 7.0 (4.0) 6.0 (2.0)
0.033 (Kruskal-Wallis test)
1.0 (2.0) 3.0 (2.0) 3.0 (1.0)
16.7 (16.7) 33.3 (31.7) 50.0 (40.8)
437 (72) 168 (28)
6.0 (4.0) 6.0 (4.0)
0.344 (Mann-Whitney U test)
2.0 (2.0) 3.0 (3.0)
20.0 (16.7) 33.3 (33.3)
549 (91) 55 (9)
6.0 (4.0) 6.0 (2.0)
0.132 (Mann-Whitney U test)
2.0 (2.0) 3.0 (2.0)
22.2 (23.3) 50.0 (32.1)
435 (72) 170 (28)
6.0 (3.0) 7.0 (4.0)
0.233 (Mann-Whitney U test)
2.0 (2.0) 3.0 (3.0)
22.2 (20.8) 33.3 (33.3)
245 (41) 300 (50) 60 (10)
6.0 (4.0) 6.0 (4.0) 6.0 (2.0)
0.693 (Kruskal-Wallis test)
1.0 (1.0) 2.0 (2.0) 3.0 (3.0)
16.7 (16.7) 28.6 (33.3) 38.8 (35.1)
573 (95) 32 (5)
6.0 (4.0) 7.0 (4.0)
0.736 (Mann-Whitney U test)
2.0 (2.0) 4.0 (3.0)
23.1 (25.6) 50.0 (30.3)
240 (40) 365 (60)
6.0 (3.0) 6.0 (4.0) 1.0 (1.0) 6.0 (4.0) 8.0 (3.0)
0.677 (Mann-Whitney U test)
1.0 (1.0) 3.0 (2.0)
16.7 (17.4) 33.3 (33.3)
0.408 (Mann-Whitney U test)
2.0 (2.0) 4.0 (2.0)
25.0 (28.8) 50.0 (57.6)
595 (98) 10 (2)
p Value
TABLE 2. Multivariate logistic regression analyses of preoperative features for prediction of organ confined disease and seminal vesicle invasion in 605 patients who underwent radical retropubic prostatectomy for clinically localized prostate cancer Organ Confined Disease
Preop serum PSA* NoPosBx % PosBx Clinical stage Biopsy Gleason sum:† 4–6 7 8–10 * Logarithmically transformed. † Not available on 2 patients.
Seminal Vesicle Invasion
HR
95% CI
p Value
HR
95% CI
p Value
0.433 0.894 0.641 0.546
0.316–0.594 0.532–1.355 0.406–0.999 0.363–0.823
⬍0.001 0.493 0.049 0.004 0.034
2.311 0.750 2.265 2.364
1.493–3.577 0.324–1.738 1.000–5.327 1.227–4.553
⬍0.001 0.502 0.050 0.010 ⬍0.001
1.000 1.094 0.405
Referent 0.699–1.713 0.182–0.902
1.000 1.459 6.693
Referent 0.691–3.081 2.642–16.950
median postoperative followup of 21.2 months (range 1.0 to 101.3). Overall PSA progression-free survival was 81.2% ⫾ 1.1% at 3 years and 77.6% ⫾ 1.4% at 5 years. Using the log rank test we found that the 6 patients with NoBx above the median did not have a different probability of PSA progression than those below the median (p ⫽ 0.2548). In contrast, patients with NoPosBx or percent PosBx above the median (2 and 25%, respectively) had an increased probability of PSA progression (p ⫽ 0.0002 and p ⬍0.0001, respectively, fig. 1). On univariate Cox proportional hazards regression analyses all standard preoperative and postoperative features, higher NoPosBx and higher percent PosBx were significantly associated with an increased risk of rapid biochemical progression after surgery (table 3). NoBx was not associated with PSA progression after radical prostatectomy on multivariate analysis (p ⫽ 0.052). On preoperative multivariate analysis only preoperative PSA, percent PosBx and biopsy Gleason sum were associated with biochemical progression (p ⬍0.001, 0.006, and Ptrend ⬍0.001, respectively, table 3). On postoperative multivariate analysis only preoperative PSA, percent PosBx, seminal vesicle invasion, lymph node metastases and final Gleason sum were associated with PSA progression after surgery (p ⬍0.001, 0.015, 0.025 and 0.026, and Ptrend ⫽ 0.013, respectively, table 3).
0.695 0.027
0.322 ⬍0.001
Association of NoBx, NoPosBx and percent PosBx with PSA doubling time and response to salvage radiation therapy. Seven patients did not undergo additional treatment after disease progression at last followup, 21 underwent local salvage radiation therapy at our institution and 43 underwent hormonal therapy as primary treatment for prostate cancer progression. Seven patients were lost to followup. PSA doubling time calculation was performed for 68 patients who had PSA progression prior to the administration of salvage therapy. PSA doubling time between radical prostatectomy failure and the initiation of salvage therapy inversely correlated with NoPosBx and percent PosBx ( ⫽ ⫺0.349, p ⫽ 0.044 and ⫽ ⫺0.396, p ⫽ 0.039, respectively) but not NoBx ( ⫽ ⫺0.79, p ⫽ 0.527). Median time from the start of salvage local radiation therapy to last followup in patients who had a favorable response was 43.7 months (range 4.4 to 71.1). Overall PSA progression-free survival after salvage local radiation therapy was 45% ⫾ 11% at 2 years and 20% ⫾ 10% at 4 years. On univariate Cox proportional hazards regression analyses NoBx, NoPosBx and percent PosBx were not associated with biochemical progression after salvage radiation therapy (p ⫽ 0.187, 0.173 and 0.076, respectively). Association of NoBx, NoPosBx and percent PosBx with
2212
PERCENT POSITIVE CORES AND PROSTATE CANCER CLINICAL OUTCOMES
proportional hazards regression analyses higher NoPosBx (HR 3.749, 95% CI 1.202 to 11.688, p ⫽ 0.023) and higher percent PosBx (HR 5.746, 95% CI 1.351 to 24.439, p ⫽ 0.018) were associated with overall survival after surgery but NoBx was not (p ⫽ 0.576). DISCUSSION
FIG. 1. Kaplan-Meier estimates of PSA progression-free probability in 605 patients with clinically localized prostate cancer treated with radical prostatectomy, as stratified into groups above or below median. A, total number of positive biopsy cores (2). B, 25% positive biopsy cores.
clinically evident distant prostate cancer metastases and patient survival. Staging evaluation including bone scan or ProstaScint scan (Cytogen Corp., Princeton, New Jersey) scan was performed at our institution in 56 patients who had biochemical progression after radical prostatectomy. A total of 13 patients (2%) had bone metastases, as evidenced by imaging. Median followup in patients who experienced PSA progression after surgery but did not have metastases was 42.0 months (range 4 to 100). Using the log rank test patients with NoPosBx or percent PosBx above the median (2 and 25%, respectively) had an increased probability of distant metastasis (p ⫽ 0.0331 and 0.0054, respectively, fig. 2). The 6 patients with NoBx above the median did not have a different probability of distant metastases than those with NoBx below the median (p ⫽ 0.9432). On univariate Cox proportional hazards regression analyses higher NoPosBx (HR 3.307, 95% CI 1.450 to 7.542, p ⫽ 0.004) and higher percent PosBx (HR 5.850, 95% CI 2.036 to 16.815, p ⫽ 0.001) were associated with an increased risk of distant metastases after surgery but NoBx was not (p ⫽ 0.512). Overall survival data were available on 629 of the 630 patients. Seven of the 629 patients (1%) were dead of any cause at analysis. Median followup in patients who experienced PSA progression after surgery and were alive at last followup was 39.3 months (range 4 to 97). On univariate Cox
Knowledge of the true extent of prostate cancer is an important factor for determining appropriate treatment. The ability to determine accurately if a cancer is organ confined may help differentiate patients who can undergo nerve sparing radical prostatectomy and patients who are best candidates for brachytherapy or external beam radiation therapy. Furthermore, predicting which patients are at high risk for PSA failure may assist in selecting those who may benefit from neoadjuvant and/or adjuvant therapy. Current nomograms using preoperative serum PSA, clinical stage and Gleason score are insufficiently accurate to predict the patients in whom local therapy may fail.1 Even with pathologically confined disease 10% to 26% of patients experience clinical recurrences.2 In the current study we confirmed the strong association of percent PosBx with established features of biologically aggressive prostate cancer, such as tumor stage and histological differentiation, biochemical progression, distant metastases and overall mortality. On preoperative and postoperative multivariate analyses percent positive biopsy cores predicted an increased risk of nonorgan confined disease and seminal vesical invasion, and a higher risk of biochemical progression. The number and percent of positive biopsies were predictive of PSA failure and metastases on Kaplan-Meier analyses. Multiple studies have evaluated the association between initial needle biopsy characteristics and postoperative pathological features. Veltri et al found that quantitative nuclear grade features in prostate biopsy specimens were the strongest predictor of pathological stage in patients with clinical T1c when combined with biopsy Gleason score and complexed PSA density.4 D’Amico et al evaluated the predictive value of percent positive prostate biopsy in 960 patients and found a statistically significant correlation between increased percent positive biopsies, and increased extracapsular extension and seminal vesical invasion (p ⬍0.0001).20 Several studies assessed the area or length of cancer in prostate biopsies. Grossklaus et al found a correlation between percent tumor in the biopsy set and the risk of extracapsular extension (p ⬍0.01).10 Bismar et al found that total percent cancer in biopsy cores was significantly related to pathological T stage on multivariate analysis (p ⫽ 0.003).5 The linear extent of prostate cancer in biopsies is also significantly related to tumor volume.12 The disadvantage of measuring tumor area in relation to total biopsy core area is that it is quite labor intensive. The number of positive cores has been found to correlate with stage in multiple studies,6 –9 including ours. In patients with greater than 2 positive cores Kaplan-Meier curves demonstrated a decreased risk of PSA progression-free probability (p ⫽ 0.0002) and metastasis-free probability (p ⫽ 0.033, figs. 1 and 2). Because urologists are sampling a varying number of biopsy cores, it may be more difficult to use the number of positive cores in a uniform matter. The percent positive cores offers a more uniform parameter for comparing different biopsy strategies. As such, it is important to separately embed specimens to be able to assess accurately this percent. Beyond its usefulness for determining pathological stage information from needle biopsies has also been evaluated for predicting metastases to lymph nodes and biochemical progression after treatment with curative intent. Several studies used the amount and distribution of Gleason grade 4 or 5
2213
PERCENT POSITIVE CORES AND PROSTATE CANCER CLINICAL OUTCOMES
TABLE 3. Univariate and multivariate cox regression analyses of preoperative and postoperative features for prediction of biochemical progression in 605 patients who underwent radical retropubic prostatectomy for clinically localized prostate cancer Univariate
Preoperative serum PSA* NoPosBx % PosBx Clinical stage Biopsy Gleason sum:† 4–6 7 8–10 Extracapsular extension Seminal vesicle involvement‡ Surgical margin status Lymph node metastasis Final Gleason sum: 4–6 7 8–10 * Logarithmically transformed. † Not available on 2 patients. ‡ Not available on 1 patient.
Preop Multivariate
Postop Multivariate
HR
95% CI
p Value
HR
95% CI
p Value
HR
95% CI
p Value
3.830 1.786 2.486 2.863
2.804–5.230 1.284–2.485 1.727–3.579 1.821–4.501
⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001
3.039 0.521 2.591 1.586
2.198–4.202 0.264–1.025 1.306–5.141 0.953–2.639
⬍0.001 0.059 0.006 0.076 ⬍0.001
1.000 1.789 7.825 4.715 6.859 3.322 19.441
Referent 1.060–3.017 4.500–13.606 2.971–7.483 4.320–10.890 2.130–5.180 9.516–39.721
1.000 1.039 3.239 –
Referent 0.576–1.872 1.623–6.466 –
0.899 ⬍0.001 – – – – – – –
⬍0.001 0.066 0.015 – – – – – 0.092 0.025 0.127 0.026 0.013
Referent 1.165–4.161 6.004–21.967
– – – – – –
1.616–3.223 0.291–1.039 1.172–4.433 – – – – – 0.918–3.103 1.089–3.611 0.891–2.537 1.069–6.639
1.000 2.201 11.484
– – – – – –
2.282 0.550 2.279 – – – – – 1.687 1.983 1.503 2.888 1.000 0.944 3.226
Referent 0.461–1.933 1.068–7.122
0.029 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 0.015 ⬍0.001
FIG. 2. Kaplan-Meier estimates of metastasis-free probability in 605 patients with clinically localized prostate cancer treated with radical prostatectomy, as stratified into groups above or below median. A, total number of positive biopsy cores (2). B, 25% positive biopsy cores.
cancer in prostate biopsies to predict the risk of lymph node spread.13 As in our study and others, percent positive biopsies is a powerful predictor of biochemical progression. Presti
0.874 0.010
et al evaluated percent positive prostate biopsies for predicting the risk of relapse after radical prostatectomy in 109 patients.15 Using Kaplan-Meier curves they noted a statistically significant decrease in disease-free survival (p ⫽ 0.03) in patients with greater than 50% positive biopsies. Risk stratification using percent positive biopsies in conjunction with biopsy Gleason score, preoperative PSA and clinical stage16 or endorectal coil magnetic resonance imaging21 allows the identification of patients at increased risk for early PSA failure. This information was particularly useful for predicting failure in patients in an intermediate risk group (1992 American Joint Committee on Cancer T2b, biopsy Gleason sum 7 or PSA greater than 10 ng/ml and 20 ng/ml or less).16 Early PSA failure is an important end point because it predicts future poor clinical outcomes.22 Due to the important association of pathological features for predicting pathological stage and progression several studies have been performed to use these factors in predictive models. Veltri et al included multiple sextant pathology variables in a neural net and were able to predict organ confined status with greater than 90% accuracy.23 Ziada et al assessed the impact of different pretreatment variables on outcome in patients with clinically confined prostate cancer using an artificial neural network.24 Percent positive biopsy tissue improved the prediction of pathological stage and biochemical failure. The ability to enhance the prediction of stage may be of particular significance in patients undergoing radiation therapy. These patients do not have the benefit of pathological staging to assist in posttreatment clinical decision making. An improved prediction of patients at high risk for early PSA progression would allow the identification of patients for clinical trials of adjuvant therapy and possibly improved survival. Several limitations in this study should be considered. Followup in this study was limited. Some variables that were inconclusive due to limited statistical power may attain statistical significance if sample size, followup or the proportion of patients with biochemical failure or clinical progression were greater. In addition, the population in this study underwent radical prostatectomy and bilateral pelvic lymphadenectomy by multiple surgeons and specimens were evaluated by multiple pathologists. Most studies evaluating NoBx, NoPosBx and percent PosBx used cohorts operated on by a single surgeon that were assembled from populations at large academic hospitals serving as tertiary referral centers. While prognostic factors may perform well in the select group of patients treated at these centers, it remains to be determined whether they are applicable to the greater population
2214
PERCENT POSITIVE CORES AND PROSTATE CANCER CLINICAL OUTCOMES
of patients with prostate cancer. Similarly while it may be preferable for a single pathologist specialized in genitourinary pathology to review each prostatectomy specimen, the setup at our institution reflects real-world practice, in which local pathologists review tissue specimens and their interpretation is then used in clinical decision making with the patient. CONCLUSIONS
Percent positive prostate biopsy cores is a strong predictor of adverse pathological features, biochemical progression, distant metastases and overall survival in patients who undergo radical prostatectomy for clinically localized disease. Percent positive prostate biopsy cores should be included in preoperative predictive models for prognosticating outcomes after primary treatment and selecting patients for inclusion in neoadjuvant and/or adjuvant therapy trials. REFERENCES
1. Partin, A. W., Kattan, M. W., Subong, E. N., Walsh, P. C., Wojno, K. J., Oesterling, J. E. et al: Combination of prostate-specific antigen, clinical stage, and Gleason score to predict pathological stage of localized prostate cancer. A multi-institutional update. JAMA, 277: 1445, 1997 2. Lerner, S. E., Blute, M. L., Bergstralh, E. J., Bostwick, D. G., Eickholt, J. T. and Zincke, H.: Analysis of risk factors for progression in patients with pathologically confined prostate cancers after radical retropubic prostatectomy. J Urol, 156: 137, 1996 3. D’Amico, A. V., Whittington, R., Malkowicz, S. B., Schultz, D., Tomaszewski, J. E. and Wein, A.: Prostate-specific antigen failure despite pathologically organ-confined and marginnegative prostate cancer: the basis for an adjuvant therapy trial. J Clin Oncol, 15: 1465, 1997 4. Veltri, R. W., Miller, M. C., Mangold, L. A., O’Dowd, G. J., Epstein, J. I. and Partin, A. W.: Prediction of pathological stage in patients with clinical stage T1c prostate cancer: the new challenge. J Urol, 168: 100, 2002 5. Bismar, T. A., Lewis, J. S., Jr., Vollmer, R. T. and Humphrey, P. A.: Multiple measures of carcinoma extent versus perineural invasion in prostate needle biopsy tissue in prediction of pathologic stage in a screening population. Am J Surg Pathol, 27: 432, 2003 6. Wills, M. L., Sauvageot, J., Partin, A. W., Gurganus, R. and Epstein, J. I.: Ability of sextant biopsies to predict radical prostatectomy stage. Urology, 51: 759, 1998 7. Borirakchanyavat, S., Bhargava, V., Shinohara, K., Toke, A., Carroll, P. R. and Presti, J. C., Jr.: Systematic sextant biopsies in the prediction of extracapsular extension at radical prostatectomy. Urology, 50: 373, 1997 8. Peller, P. A., Young, D. C., Marmaduke, D. P., Marsh, W. L. and Badalament, R. A.: Sextant prostate biopsies. A histopathologic correlation with radical prostatectomy specimens. Cancer, 75: 530, 1995 9. Huland, H., Hammerer, P., Henke, R.-P. and Huland, E.: Preoperative prediction of tumor heterogeneity and recurrence after radical prostatectomy for localized prostatic carcinoma with digital rectal examination, prostate specific antigen and the results of 6 systematic biopsies. J Urol, 155: 1344, 1996 10. Grossklaus, D. J., Coffey, C. S., Shappell, S. B., Jack, G. S., Chang, S. S. and Cookson, M. S.: Percent of cancer in the biopsy set predicts pathological findings after prostatectomy. J Urol, 167: 2032, 2002
11. Tigrani, V. S., Bhargava, V., Shinohara, K. and Presti, J. C., Jr.: Number of positive systematic sextant biopsies predicts surgical margin status at radical prostatectomy. Urology, 54: 689, 1999 12. Lewis, J. S., Jr., Vollmer, R. T. and Humphrey, P. A.: Carcinoma extent in prostate needle biopsy tissue in the prediction of whole gland tumor volume in a screening population. Am J Clin Pathol, 118: 442, 2002 13. Conrad, S., Graefen, M., Pichlmeier, U., Henke, R.-P., Hammerer, P. G. and Huland, H.: Systematic sextant biopsies improve preoperative prediction of pelvic lymph node metastases in patients with clinically localized prostatic carcinoma. J Urol, 159: 2023, 1998 14. Haese, A., Chaudhari, M., Miller, M. C., Epstein, J. I., Huland, H., Palisaar, J. et al: Quantitative biopsy pathology for the prediction of pathologically organ-confined prostate carcinoma: a multiinstitutional validation study. Cancer, 97: 969, 2003 15. Presti, J. C., Jr., Shinohara, K., Bacchetti, P., Tigrani, V. and Bhargava, V.: Positive fraction of systematic biopsies predicts risk of relapse after radical prostatectomy. Urology, 52: 1079, 1998 16. D’Amico, A. V., Whittington, R., Malkowicz, S. B., Schultz, D., Fondurulia, J., Chen, M. H. et al: Clinical utility of the percentage of positive prostate biopsies in defining biochemical outcome after radical prostatectomy for patients with clinically localized prostate cancer. J Clin Oncol, 18: 1164, 2000 17. Shariat, S. F., Khoddami, S. M., Saboorian, H., Koeneman, K. S., Sagalowsky, A. I., Cadeddu, J. A. et al: Lymphovascular invasion is a pathologic feature of biologically aggressive disease in patients treated with radical prostatectomy. J Urol, 171: 1122, 2004 18. Henson, D. E., Hutter, R. V. and Farrow, G.: Practice protocol for the examination of specimens removed from patients with carcinoma of the prostate gland. A publication of the Cancer Committee, College of American Pathologists. Task Force on the Examination of Specimens Removed From Patients With Prostate Cancer. Arch Pathol Lab Med, 118: 779, 1994 19. Schmid, H. P., McNeal, J. E. and Stamey, T. A.: Observations on the doubling time of prostate cancer. The use of serial prostate-specific antigen in patients with untreated disease as a measure of increasing cancer volume. Cancer, 71: 2031, 1993 20. D’Amico, A. V., Whittington, R., Malkowicz, S. B., Schultz, D., Silver, B., Henry, L. et al: Investigating the clinical utility of the percent of positive prostate biopsies in predicting PSA outcome following local therapy for patients with clinically localized prostate cancer. Prostate Cancer Prostatic Dis, 3: 259, 2000 21. D’Amico, A. V., Whittington, R., Malkowicz, S. B., Wu, Y. H., Chen, M., Art, M. et al: Combination of the preoperative PSA level, biopsy gleason score, percentage of positive biopsies, and MRI T-stage to predict early PSA failure in men with clinically localized prostate cancer. Urology, 55: 572, 2000 22. Pound, C. R., Partin, A. W., Eisenberger, M. A., Chan, D. W., Pearson, J. D. and Walsh, P. C.: Natural history of progression after PSA elevation following radical prostatectomy. JAMA, 281: 1591, 1999 23. Veltri, R. W., Chaudhari, M., Miller, M. C., Poole, E. C., O’Dowd, G. J. and Partin, A. W.: Comparison of logistic regression and neural net modeling for prediction of prostate cancer pathologic stage. Clin Chem, 48: 1828, 2002 24. Ziada, A. M., Lisle, T. C., Snow, P. B., Levine, R. F., Miller, G. and Crawford, E. D.: Impact of different variables on the outcome of patients with clinically confined prostate carcinoma: prediction of pathologic stage and biochemical failure using an artificial neural network. Cancer, 91: 1653, 2001