- Email: [email protected]
Preoperative PSA and progression-free survival after radical prostatectomy for Stage T1c disease
ADULT UROLOGY
PREOPERATIVE PSA AND PROGRESSION-FREE SURVIVAL AFTER RADICAL PROSTATECTOMY FOR STAGE T1c DISEASE JO ANN V. ANTENOR, KIMBERLY A. ROEHL, SCOTT E. EGGENER, SHILAJIT D. KUNDU, MISOP HAN, AND WILLIAM J. CATALONA
ABSTRACT Objectives. To examine biochemical progression-free survival (PFS) rates as a function of preoperative prostate-specific antigen (PSA) in patients with clinical Stage T1c prostate cancer treated with radical prostatectomy. Controversy exists about whether performing prostate biopsies for PSA levels in the 2.6 to 4.0 ng/mL range provides a PFS advantage compared with detection at higher PSA ranges. Methods. A total of 2804 men with clinical Stage T1c prostate cancer were treated with radical retropubic prostatectomy and monitored prospectively. The study parameters included preoperative PSA level, pathologic tumor stage, and Gleason grade. Patients were grouped into four clinically relevant strata according to their preoperative PSA level: 2.6 to 4.0, 4.1 to 7.0, 7.1 to 10.0, and greater than 10 ng/mL. The primary outcome was the 10-year actuarial biochemical PFS estimate generated using the Kaplan-Meier method. We compared the strata using the log-rank test. Cancer progression rates were compared using the Cochran Armitage test for trend. The chi-square test was used to compare the pathologic parameters among the PSA strata. Results. Of the men with a preoperative PSA level of 2.6 to 4.0, 4.1 to 7.0, 7.1 to 10.0, and greater than 10.0 ng/mL, 81%, 74%, 72%, and 60%, respectively, had organ-confined disease (P ⫽ 0.001) and 23%, 28%, 35%, and 47%, respectively, had a pathologic Gleason grade of 7 or greater (P ⫽ 0.001). The corresponding 10-year PFS estimates were 88%, 80%, 76%, and 61% (P ⫽ 0.0001, for trend). Conclusions. Among men with clinical Stage T1c prostate cancer, those with a PSA level of 2.6 to 4.0 ng/mL had the greatest rate of organ-confined disease, lowest pathologic Gleason grade, and greatest 10-year PFS rate. UROLOGY 66: 156–160, 2005. © 2005 Elsevier Inc.
A
pproximately 22% to 25% of men with serum prostate-specific antigen (PSA) levels of 2.6 to 4.0 ng/mL have prostate cancer that can be detected by sextant ultrasound-guided needle biopsies.1,2 Cancers detected in this low PSA range are This study was supported in part by a grant from Beckman Coulter, Inc., Fullerton, California and by the Urological Research Foundation. W. J. Catalona is a study investigator funded in part by a grant from Beckman Coulter (Hybritech). From the Departments of Neurology, Psychiatry, and Surgery, Washington University School of Medicine, St. Louis, Missouri; and Department of Urology, Northwestern University Feinberg School of Medicine, Chicago, Illinois Reprint requests: William J. Catalona, M.D., Department of Urology, Northwestern University Feinberg School of Medicine, 675 North Saint Clair Street, Suite 20-150, Chicago, IL 60611. E-mail: [email protected] Submitted: July 28, 2004, accepted (with revisions): January 6, 2005 © 2005 ELSEVIER INC. 156
ALL RIGHTS RESERVED
virtually all clinically localized and approximately 80% are pathologically organ confined.3 Controversy exists about whether a progression-free survival (PFS) advantage exists to detecting prostate cancer in this PSA range compared with detecting these cancers at a higher PSA level.3–5 To evaluate this issue further, we determined the biochemical PFS estimates as a function of preoperative PSA level in a large cohort of patients with Stage T1c prostate cancer treated with radical retropubic prostatectomy (RRP). MATERIAL AND METHODS STUDY PROTOCOL The patients included in this study were either treated by one of us (W.J.C.) or had had cancer detected in a prostate cancer screening study (PSA Study) that had enrolled community volunteers from 1989 to 2001. Patients had chosen to be treated by RRP. All volunteers provided informed consent. 0090-4295/05/$30.00 doi:10.1016/j.urology.2005.01.008
TABLE I. Patient distribution by preoperative prostate-specific antigen level Preoperative PSA (ng/mL) Characteristic
2.6–4.0
4.1–7.0
7.1–10.0
>10.0
Men (%) Men entered in study before May 1995* (n) Mean age (yr) Follow-up (yr) ⱕ5.0 5.1–7.0 7.1–10.0 ⱖ10 Overall follow-up (mo) Median Range
538 (19) 0 62
1496 (54) 287 62
445 (16) 85 62
323 (12) 78 64
432 (81) 85 (16) 7 (1) 8 (2)
985 (67) 235 (16) 197 (13) 50 (3)
258 (59) 76 (17) 84 (19) 21 (5)
187 (58) 66 (21) 49 (15) 18 (6)
50 1–149
45 1–168
P Value
⬍0.0001†
35 1–149
41 1–153
⬍0.0001‡
KEY: PSA ⫽ prostate-specific antigen. Data presented as number of men, with percentage in parentheses, unless noted otherwise. * Only included men from PSA Study. † Chi-square test. ‡ Kruskal-Wallis test.
The volunteers included in the PSA Study had undergone serial PSA determinations at either 6-month or 1-year intervals, depending on the findings from their previous screening tests. Until May 1995, transrectal ultrasound-guided needle biopsy of the prostate was recommended for a serum PSA level greater than 4.0 ng/mL or digital rectal examination (DRE) findings that were suspicious for prostate cancer. After May 1995, biopsy was recommended for a PSA level greater than 2.5 ng/mL or suspicious DRE findings. At least sextant biopsies were routinely performed.6 After RRP, patients were monitored with a PSA test every 6 months and an annual DRE. A verified postoperative PSA level greater than 0.2 ng/mL was considered biochemical evidence of cancer progression.
CLINICAL AND PATHOLOGIC STAGING Clinical staging was performed as previously described.7 All study subjects were classified as having clinical Stage T1c disease. Men whose cancer was confined to the prostate with clear surgical margins were categorized as having pathologically organ-confined disease (Stage pT2 R0), and those with extraprostatic tumor extension (pT3a) and/or positive surgical margins (R1) and/or seminal vesicle (pT3b) or lymph node involvement (N1) were classified as having pathologically advanced disease. The Gleason grade was assigned as previously described for RRP specimens.7
STATISTICAL ANALYSIS All statistical analyses were performed using Statistical Analysis Systems, version 8.2 (SAS Institute, Cary, NC) and Statistical Package for Social Sciences, version 10.0, for Windows (SPSS, Chicago, Ill). The study parameters included preoperative PSA level, follow-up time, pathologic tumor stage, and Gleason sum. We divided patients into four preoperative PSA strata: 2.6 to 4.0, 4.1 to 7.0, 7.1 to 10.0, and greater than 10.0 ng/mL. The main outcome parameter was the 10-year biochemical PFS estimate. Actuarial probabilities of PFS were generated using the Kaplan-Meier method, and survival curves were compared using the log-rank test. The proportion of men with progression was compared using the Cochran Armitage test for trend. We used the chi-square test to compare the pathologic paramUROLOGY 66 (1), 2005
eters among the PSA strata, and the Kruskal-Wallis test to compare the follow-up time among the PSA strata.
RESULTS POPULATION CHARACTERISTICS Approximately 4000 men in our study population underwent RRP, of whom 2804 (70%) had been diagnosed with clinical Stage T1c disease and were eligible for our study. Of these men, 1731 (62%) were from W.J.C.’s RRP series and 1061 (38%) were from the PSA Study; 92% were white, 5% black, and 3% of another ethnicity. The men had a mean age of 62 years (range 36 to 81). The mean preoperative PSA level was 6.7 ng/mL (median 5.4, range 2.6 to 66.2). Approximately 19%, 54%, 16%, and 12% of men had a preoperative PSA level of 2.6 to 4.0, 4.1 to 7.0, 7.1 to 10.0, and greater than 10.0 ng/mL, respectively. The mean follow-up interval was 43 months (median 36, range 1 to 172; Table I). PATHOLOGIC STAGE AND GLEASON GRADE Of the men with a PSA level of 2.6 to 4.0, 4.1 to 7.0, 7.1 to 10.0, and greater than 10.0 ng/mL, 81%, 74%, 72%, and 60%, respectively, had pathologically organ-confined tumor, and 23%, 28%, 35%, and 47%, respectively, had a pathologic Gleason grade of 7 or greater. Using the criteria of Epstein et al.8 for harmless cancer, 3 (less than 1%) were detected in the lowest PSA range. None were seen in higher PSA ranges. Using the criteria of Ohori et al.9 for harmless cancer, 11%, 4%, 1%, and 0% of the cancers were clinically insignificant among the men with a PSA level of 2.6 to 4.0, 4.1 to 7.0, 7.1 to 10.0, and greater than 10.0 ng/mL, respectively (Table II). None of the cancers detected had lymph node metastases (data not shown). 157
TABLE II. Preoperative PSA range and pathologic tumor characteristics Preoperative PSA (ng/mL) Characteristic Pathologic stage Organ-confined disease (pT2 R0) Pathologically advanced with negative margins Pathologically advanced with positive margins Pathologically advanced with positive seminal vesicle involvement Gleason grade ⱕ6 3⫹4 4⫹3 7† ⱖ8 “Harmless” cancer Epstein8 criteria Ohori9 criteria
2.6–4.0
4.1–7.0
7.1–10.0
>10.0
423 (81) 21 (4) 75 (14)
1086 (74) 99 (7) 274 (19)
316 (72) 31 (7) 2 (19)
188 (60) 32 (10) 81 (25)
P Value ⬍0.0001*
3 (⬍1)
13 (1)
8 (2)
17 (5) ⬍0.0001*
415 (77) 41 (8) 5 (⬍1) 73 (14) 4 (⬍1) 3 (⬍1) 34 (11)
1074 (72) 233 (16) 40 (3) 101 (7) 48 (3) 0 (0) 15 (4)
291 (65) 86 (19) 25 (6) 24 (5) 19 (4) 0 (0) 1 (1)
170 (53) 79 (24) 21 (7) 33 (10) 20 (6) 0 (0) 0 (0)
0.14 ⬍0.0001*
KEY: PSA ⫽ prostate-specific antigen. Data presented as number of men, with percentage in parentheses. * Chi square test. † Cancer detected earlier in series were not graded as either 3 ⫹ 4 or 4 ⫹ 3.
ther age group or race showed no statistically significant differences in survival rates (data not shown). COMMENT
FIGURE 1. Ten-year PFS (PSA less than 0.3 ng/mL) probability by preoperative PSA group.
PREOPERATIVE PSA AND PFS RATES The 10-year PFS estimate for men with a preoperative PSA level of 2.6 to 4.0, 4.1 to 7.0, 7.1 to 10.0, and greater than 10.0 ng/mL was 88%, 80%, 76%, and 61%, respectively (Fig. 1 and Table III). Overall, 5% of men with a PSA level of 2.6 to 4.0 ng/mL developed progression, and 26% of men with a PSA level greater than 10.0 ng/mL did so (Table III). The log-rank test for trend comparing the PFS rates across the preoperative PSA strata was statistically significant, but the log-rank test comparing a PSA level of 2.6 to 4.0 ng/mL to one of 4.1 to 7.0 ng/mL did not quite reach statistical significance (Table III). Computing the 10-year PFS estimates for each preoperative PSA stratum by ei158
Our study is, to our knowledge, the largest published study to date demonstrating that among men with clinical Stage T1c prostate cancer, those with a PSA level of 2.6 to 4.0 ng/mL have the greatest rate of organ-confined tumors, lowest pathologic Gleason scores, and greatest 10-year PFS rate. Among men with normal DRE findings and a PSA level of 2.6 to 4.0 ng/mL, prostate cancer is detected by sextant biopsy in 22% to 25%.2 The goal of prostate cancer screening is to identify clinically significant tumors that are still organ confined and curable. The rationale for PSA screening at levels of 2.6 to 4.0 ng/mL derives from two observations. First, clinically significant disease as defined by the pathologic tumor features is common in these men, and second, the prevalence of high-risk features is lower compared with those found using the traditional PSA screening threshold of 4.0 ng/mL. In our study, potentially aggressive disease, defined as positive margins (⬃15%), extraprostatic tumor extension (⬃20%), and Gleason grade of 7 or greater (⬃25%), was common in men with a PSA level of 2.6 to 4.0 ng/mL, although less so than in men with a PSA level greater than 4.0 ng/mL (positive margins 21%, extraprostatic disease 31%, and Gleason grade of 7 or greater 37%). Among men with a PSA level of 2.6 to 4.0 ng/mL, 81% had pathologically organ-confined UROLOGY 66 (1), 2005
TABLE III. Preoperative PSA range and progression-free survival Preoperative PSA (ng/mL) 10-yr PFS (%) Progression (%) 5-yr PFS (%)
2.6–4.0
4.1–7.0
7.1–10.0
>10.0
P Value
88* (82–94) 5 92 (90–95)
80* (76–84) 9 89 (87–91)
76 (67–85) 13 87 (83–91)
61 (53–69) 26 75 (70–80)
0.0001† 0.001‡ 0.0001†
KEY: PSA ⫽ prostate-specific antigen; PFS ⫽ progression-free survival. Data in parentheses are 95% confidence intervals. * PFS for PSA 2.6 – 4.0 vs. 4.1–7.0 ng/mL (P ⫽ 0.06). † Log-rank test. ‡ Cochran-Armitage test for trend.
disease; men with greater PSA levels had a lower rate of organ-confined disease. A previous study by Catalona et al.2 showed that 22% of men who underwent biopsy for a PSA level between 2.6 and 4.0 ng/mL were diagnosed with cancer. All of these cancers were clinically localized and 81% were organ confined.2 Other investigators have also shown that men diagnosed with prostate cancer with a PSA level of less than 4.0 ng/mL have a greater incidence of favorable features, such as organ-confined disease,4,10 –14 cancer-free surgical margins,4,11–13 and lower cancer volume.10,11 An argument against lowering the PSA screening threshold for biopsy to less than 4.0 ng/mL is the possible risk of overdetecting clinically insignificant tumors. However, Krumholtz et al.11 showed that screening for prostate cancer in the 2.6 to 4.0 ng/mL PSA range resulted in more frequent detection of small, organ-confined tumors without overdetection of “clinically insignificant” or “clinically unimportant” cancers. Lowering the PSA cutoff to 2.6 ng/mL also might mean subjecting more men to unnecessary biopsies compared with using a 4.0 ng/mL PSA cutoff. Although avoiding unnecessary biopsies is desirable, studies have shown that by using a 4.0 ng/mL PSA cutoff, a substantial proportion of cancers that would have been detected at lower PSA cutoffs are either missed or detected later.2 The extent to which delayed detection affects outcomes has not been clearly defined; however, our results suggest that delayed detection might compromise PFS. In our study, a lower preoperative PSA level (2.6 to 4.0 ng/mL) was associated with a greater PFS rate. For each strata of preoperative PSA studied, as the PSA level increased, the PFS rate decreased. Other investigators have also evaluated the role of preoperative PSA and PFS after RRP. Our findings are similar to those of many other large prostatectomy series.3,12,14,15 Han et al.3 presented a nomogram for 845 men with clinical Stage T1c disease. They used a PSA level of greater than 0.2 ng/mL as a criterion for cancer progression, grouped patients on the basis UROLOGY 66 (1), 2005
of the preoperative PSA level (0 to 4.0, 4.1 to 10, 10.1 to 20, and greater than 20 ng/mL), and stratified patients by biopsy Gleason score. The PFS estimations at 3, 5, 7, and 10 years for men with a PSA level of 0 to 4.0 ng/mL were consistently greater for every Gleason grade compared with men with a PSA level of 4.1 to 10.0 ng/mL. Similar results were demonstrated among men with organconfined disease. Men with a PSA level of less than 4.0 ng/mL, regardless of Gleason grade, uniformly had greater PFS rates than men with a PSA level of 4.1 to 10.0 ng/mL. Shekarriz et al.12 examined the disease-free survival of men after RRP. For clinical Stage T1c and T2 cancer, the disease-free survival rate for men with a preoperative PSA level less than 4.0 ng/mL was 95% compared with 85% for those with a PSA level greater than 4.0 ng/mL. Hull et al.4 reported that men with a PSA level less than 4.0 ng/mL had an 89% 10-year PFS rate compared with 84% for men with a PSA level between 4.0 and 9.9 ng/mL. In contrast, D’Amico et al.5 assessed PFS after RRP and stratified by pathologic Gleason grade. For men with Gleason grade of 6 or less, no difference in PFS was noted on the basis of the preoperative PSA level. However, for men with a Gleason grade of 7 or greater, those in the lowest PSA range (2.6 to 4.0 ng/mL) had worse PFS than those with greater PSA levels (4.1 to 10.0 ng/mL). These results reflect the aggressive nature of high-grade cancer in some men with low PSA levels. The investigators suggested that lower serum testosterone levels may be associated with aggressive disease in the setting of low PSA levels. Alternatively, some poorly differentiated prostate cancers do not produce high levels of PSA until they are advanced, if ever. Stamey16 reported no material difference in progression among patients with a PSA level less than 9 ng/mL. He concluded that in this PSA range, the prognosis was unrelated to the PSA level. Several aspects of his study suggest possible alternative explanations for these results. First, his series included patients with palpable (Stage T2) tumors. Some palpable low-PSA producing tumors are 159
highly aggressive, and their inclusion would negatively bias the prognosis of patients in the low PSA range. The PFS estimates in his study were lower than observed in most contemporary studies, including ours. Second, the follow-up in his series was short; thus, some patients at the higher end of this PSA range who would be expected to be at greater risk of progression might not have had time to manifest biochemical evidence of progression. Third, PFS was plotted as a function of the logarithm of the PSA level. Thus, even if a linear relationship between PSA level and PFS existed, a curved function plot similar to that obtained by Stamey16 would result. Our study had several limitations. One was our predominantly white study group. The results of our study may not be applicable to men of other ethnic backgrounds. Another limitation was the variable follow-up intervals among the different PSA strata. Men with a PSA level of 2.6 to 4.0 ng/mL had the shortest follow-up; however, differences in PFS were also evident at 3 and 5 years of follow-up (3-year data not shown). Although differences in disease-specific and overall survival had not yet been demonstrated, PSA PFS is usually an intermediate-term marker for these endpoints. Finally, we recognize that our longitudinal follow-up study is not the optimal method for analyzing how lowering the PSA cutoff for screening of prostate cancer affects treatment outcomes. A randomized clinical trial using different PSA cutoffs would be more valid. Nevertheless, our large study cohort with long-term follow-up provides important insights into likely outcome trends. CONCLUSIONS Among men with clinical Stage T1c prostate cancer, those with a PSA level of 2.6 to 4.0 ng/mL have the greatest rate of organ-confined disease, lowest pathologic Gleason grade, and greatest 10-year PFS rate. REFERENCES 1. Babaian RJ, Johnston DA, Naccarato W, et al: The incidence of prostate cancer in a screening population with a serum prostate specific antigen between 2.5 and 4.0 ng/ml: relation to biopsy strategy. J Urol 165: 757–760, 2001.
160
2. Catalona WJ, Smith DS, and Ornstein DK: Prostate cancer detection in men with serum PSA concentrations of 2.6 to 4.0 ng/mL and benign prostate examination: enhancement of specificity with free PSA measurements. JAMA 277: 1452– 1455, 1997. 3. Han M, Partin AW, Zahurak M, et al: Biochemical (prostate specific antigen) recurrence probability following radical prostatectomy for clinically localized prostate cancer. J Urol 169: 517–523, 2003. 4. Hull GW, Rabbani F, Abbas F, et al: Cancer control with radical prostatectomy alone in 1,000 consecutive patients. J Urol 167: 528 –534, 2002. 5. D’Amico AV, Chen MH, Malkowicz SB, et al: Lower prostate specific antigen outcome than expected following radical prostatectomy in patients with high grade prostate and a prostatic specific antigen level of 4 ng/ml. or less. J Urol 167: 2025–2030, 2002. 6. Smith DS, Humphrey PA, and Catalona WJ: The early detection of prostate carcinoma with prostate specific antigen: the Washington University experience. Cancer 80: 1852–1856, 1997. 7. Catalona WJ, Smith DS, Ratliff TL, et al: Measurement of prostate-specific antigen in serum as a screening test for prostate cancer. N Engl J Med 324: 1156 –1161, 1991. 8. Epstein JI, Walsh PC, Carmichael M, et al: Pathologic and clinical findings to predict tumor extent of nonpalpable (stage T1c) prostate cancer. JAMA 271: 368 –374, 1994. 9. Ohori M, Wheeler TM, Dunn JK, et al: The pathological features and prognosis of prostate cancer detectable with current diagnostic tests. J Urol 152: 1714 –1720, 1994. 10. Schroder FH, van der Cruijsen-Koeter I, de Koning HJ, et al: Prostate cancer detection at low prostate specific antigen. J Urol 163: 806 – 812, 2000. 11. Krumholtz JS, Carvalhal GF, Ramos CG, et al: Prostatespecific antigen cutoff of 2.6 ng/mL for prostate cancer screening is associated with favorable pathologic tumor features. Urology 60: 469 – 473, 2002. 12. Shekarriz B, Upadhyay J, Bianco FJ Jr, et al: Impact of preoperative serum PSA level from 0 to 10 ng/ml on pathological findings and disease-free survival after radical prostatectomy. Prostate 48: 136 –143, 2001. 13. Berger AP, Volgger H, Rogatsch H, et al: Screening with low PSA cutoff values results in low rates of positive surgical margins in radical prostatectomy specimens. Prostate 53: 241–245, 2002. 14. Aleman M, Karakiewicz PI, Kupelian P, et al: Age and PSA predict likelihood of organ-confined disease in men presenting with PSA less than 10 ng/mL: implications for screening. Urology 62: 70 –74, 2003. 15. Freedland SJ, Terris MK, Csathy GS, et al: Preoperative model for predicting prostate specific antigen recurrence after radical prostatectomy using percent of biopsy tissue with cancer, biopsy Gleason grade and serum prostate specific antigen. J Urol 171: 2215–2220, 2004. 16. Stamey TA: Preoperative serum prostate-specific antigen (PSA) below 10 g/l predicts neither the presence of prostate cancer nor the rate of postoperative PSA failure. Clin Chem 47: 631– 634, 2001.
UROLOGY 66 (1), 2005
PREOPERATIVE PSA AND PROGRESSION-FREE SURVIVAL AFTER RADICAL PROSTATECTOMY FOR STAGE T1c DISEASE JO ANN V. ANTENOR, KIMBERLY A. ROEHL, SCOTT E. EGGENER, SHILAJIT D. KUNDU, MISOP HAN, AND WILLIAM J. CATALONA
ABSTRACT Objectives. To examine biochemical progression-free survival (PFS) rates as a function of preoperative prostate-specific antigen (PSA) in patients with clinical Stage T1c prostate cancer treated with radical prostatectomy. Controversy exists about whether performing prostate biopsies for PSA levels in the 2.6 to 4.0 ng/mL range provides a PFS advantage compared with detection at higher PSA ranges. Methods. A total of 2804 men with clinical Stage T1c prostate cancer were treated with radical retropubic prostatectomy and monitored prospectively. The study parameters included preoperative PSA level, pathologic tumor stage, and Gleason grade. Patients were grouped into four clinically relevant strata according to their preoperative PSA level: 2.6 to 4.0, 4.1 to 7.0, 7.1 to 10.0, and greater than 10 ng/mL. The primary outcome was the 10-year actuarial biochemical PFS estimate generated using the Kaplan-Meier method. We compared the strata using the log-rank test. Cancer progression rates were compared using the Cochran Armitage test for trend. The chi-square test was used to compare the pathologic parameters among the PSA strata. Results. Of the men with a preoperative PSA level of 2.6 to 4.0, 4.1 to 7.0, 7.1 to 10.0, and greater than 10.0 ng/mL, 81%, 74%, 72%, and 60%, respectively, had organ-confined disease (P ⫽ 0.001) and 23%, 28%, 35%, and 47%, respectively, had a pathologic Gleason grade of 7 or greater (P ⫽ 0.001). The corresponding 10-year PFS estimates were 88%, 80%, 76%, and 61% (P ⫽ 0.0001, for trend). Conclusions. Among men with clinical Stage T1c prostate cancer, those with a PSA level of 2.6 to 4.0 ng/mL had the greatest rate of organ-confined disease, lowest pathologic Gleason grade, and greatest 10-year PFS rate. UROLOGY 66: 156–160, 2005. © 2005 Elsevier Inc.
A
pproximately 22% to 25% of men with serum prostate-specific antigen (PSA) levels of 2.6 to 4.0 ng/mL have prostate cancer that can be detected by sextant ultrasound-guided needle biopsies.1,2 Cancers detected in this low PSA range are This study was supported in part by a grant from Beckman Coulter, Inc., Fullerton, California and by the Urological Research Foundation. W. J. Catalona is a study investigator funded in part by a grant from Beckman Coulter (Hybritech). From the Departments of Neurology, Psychiatry, and Surgery, Washington University School of Medicine, St. Louis, Missouri; and Department of Urology, Northwestern University Feinberg School of Medicine, Chicago, Illinois Reprint requests: William J. Catalona, M.D., Department of Urology, Northwestern University Feinberg School of Medicine, 675 North Saint Clair Street, Suite 20-150, Chicago, IL 60611. E-mail: [email protected] Submitted: July 28, 2004, accepted (with revisions): January 6, 2005 © 2005 ELSEVIER INC. 156
ALL RIGHTS RESERVED
virtually all clinically localized and approximately 80% are pathologically organ confined.3 Controversy exists about whether a progression-free survival (PFS) advantage exists to detecting prostate cancer in this PSA range compared with detecting these cancers at a higher PSA level.3–5 To evaluate this issue further, we determined the biochemical PFS estimates as a function of preoperative PSA level in a large cohort of patients with Stage T1c prostate cancer treated with radical retropubic prostatectomy (RRP). MATERIAL AND METHODS STUDY PROTOCOL The patients included in this study were either treated by one of us (W.J.C.) or had had cancer detected in a prostate cancer screening study (PSA Study) that had enrolled community volunteers from 1989 to 2001. Patients had chosen to be treated by RRP. All volunteers provided informed consent. 0090-4295/05/$30.00 doi:10.1016/j.urology.2005.01.008
TABLE I. Patient distribution by preoperative prostate-specific antigen level Preoperative PSA (ng/mL) Characteristic
2.6–4.0
4.1–7.0
7.1–10.0
>10.0
Men (%) Men entered in study before May 1995* (n) Mean age (yr) Follow-up (yr) ⱕ5.0 5.1–7.0 7.1–10.0 ⱖ10 Overall follow-up (mo) Median Range
538 (19) 0 62
1496 (54) 287 62
445 (16) 85 62
323 (12) 78 64
432 (81) 85 (16) 7 (1) 8 (2)
985 (67) 235 (16) 197 (13) 50 (3)
258 (59) 76 (17) 84 (19) 21 (5)
187 (58) 66 (21) 49 (15) 18 (6)
50 1–149
45 1–168
P Value
⬍0.0001†
35 1–149
41 1–153
⬍0.0001‡
KEY: PSA ⫽ prostate-specific antigen. Data presented as number of men, with percentage in parentheses, unless noted otherwise. * Only included men from PSA Study. † Chi-square test. ‡ Kruskal-Wallis test.
The volunteers included in the PSA Study had undergone serial PSA determinations at either 6-month or 1-year intervals, depending on the findings from their previous screening tests. Until May 1995, transrectal ultrasound-guided needle biopsy of the prostate was recommended for a serum PSA level greater than 4.0 ng/mL or digital rectal examination (DRE) findings that were suspicious for prostate cancer. After May 1995, biopsy was recommended for a PSA level greater than 2.5 ng/mL or suspicious DRE findings. At least sextant biopsies were routinely performed.6 After RRP, patients were monitored with a PSA test every 6 months and an annual DRE. A verified postoperative PSA level greater than 0.2 ng/mL was considered biochemical evidence of cancer progression.
CLINICAL AND PATHOLOGIC STAGING Clinical staging was performed as previously described.7 All study subjects were classified as having clinical Stage T1c disease. Men whose cancer was confined to the prostate with clear surgical margins were categorized as having pathologically organ-confined disease (Stage pT2 R0), and those with extraprostatic tumor extension (pT3a) and/or positive surgical margins (R1) and/or seminal vesicle (pT3b) or lymph node involvement (N1) were classified as having pathologically advanced disease. The Gleason grade was assigned as previously described for RRP specimens.7
STATISTICAL ANALYSIS All statistical analyses were performed using Statistical Analysis Systems, version 8.2 (SAS Institute, Cary, NC) and Statistical Package for Social Sciences, version 10.0, for Windows (SPSS, Chicago, Ill). The study parameters included preoperative PSA level, follow-up time, pathologic tumor stage, and Gleason sum. We divided patients into four preoperative PSA strata: 2.6 to 4.0, 4.1 to 7.0, 7.1 to 10.0, and greater than 10.0 ng/mL. The main outcome parameter was the 10-year biochemical PFS estimate. Actuarial probabilities of PFS were generated using the Kaplan-Meier method, and survival curves were compared using the log-rank test. The proportion of men with progression was compared using the Cochran Armitage test for trend. We used the chi-square test to compare the pathologic paramUROLOGY 66 (1), 2005
eters among the PSA strata, and the Kruskal-Wallis test to compare the follow-up time among the PSA strata.
RESULTS POPULATION CHARACTERISTICS Approximately 4000 men in our study population underwent RRP, of whom 2804 (70%) had been diagnosed with clinical Stage T1c disease and were eligible for our study. Of these men, 1731 (62%) were from W.J.C.’s RRP series and 1061 (38%) were from the PSA Study; 92% were white, 5% black, and 3% of another ethnicity. The men had a mean age of 62 years (range 36 to 81). The mean preoperative PSA level was 6.7 ng/mL (median 5.4, range 2.6 to 66.2). Approximately 19%, 54%, 16%, and 12% of men had a preoperative PSA level of 2.6 to 4.0, 4.1 to 7.0, 7.1 to 10.0, and greater than 10.0 ng/mL, respectively. The mean follow-up interval was 43 months (median 36, range 1 to 172; Table I). PATHOLOGIC STAGE AND GLEASON GRADE Of the men with a PSA level of 2.6 to 4.0, 4.1 to 7.0, 7.1 to 10.0, and greater than 10.0 ng/mL, 81%, 74%, 72%, and 60%, respectively, had pathologically organ-confined tumor, and 23%, 28%, 35%, and 47%, respectively, had a pathologic Gleason grade of 7 or greater. Using the criteria of Epstein et al.8 for harmless cancer, 3 (less than 1%) were detected in the lowest PSA range. None were seen in higher PSA ranges. Using the criteria of Ohori et al.9 for harmless cancer, 11%, 4%, 1%, and 0% of the cancers were clinically insignificant among the men with a PSA level of 2.6 to 4.0, 4.1 to 7.0, 7.1 to 10.0, and greater than 10.0 ng/mL, respectively (Table II). None of the cancers detected had lymph node metastases (data not shown). 157
TABLE II. Preoperative PSA range and pathologic tumor characteristics Preoperative PSA (ng/mL) Characteristic Pathologic stage Organ-confined disease (pT2 R0) Pathologically advanced with negative margins Pathologically advanced with positive margins Pathologically advanced with positive seminal vesicle involvement Gleason grade ⱕ6 3⫹4 4⫹3 7† ⱖ8 “Harmless” cancer Epstein8 criteria Ohori9 criteria
2.6–4.0
4.1–7.0
7.1–10.0
>10.0
423 (81) 21 (4) 75 (14)
1086 (74) 99 (7) 274 (19)
316 (72) 31 (7) 2 (19)
188 (60) 32 (10) 81 (25)
P Value ⬍0.0001*
3 (⬍1)
13 (1)
8 (2)
17 (5) ⬍0.0001*
415 (77) 41 (8) 5 (⬍1) 73 (14) 4 (⬍1) 3 (⬍1) 34 (11)
1074 (72) 233 (16) 40 (3) 101 (7) 48 (3) 0 (0) 15 (4)
291 (65) 86 (19) 25 (6) 24 (5) 19 (4) 0 (0) 1 (1)
170 (53) 79 (24) 21 (7) 33 (10) 20 (6) 0 (0) 0 (0)
0.14 ⬍0.0001*
KEY: PSA ⫽ prostate-specific antigen. Data presented as number of men, with percentage in parentheses. * Chi square test. † Cancer detected earlier in series were not graded as either 3 ⫹ 4 or 4 ⫹ 3.
ther age group or race showed no statistically significant differences in survival rates (data not shown). COMMENT
FIGURE 1. Ten-year PFS (PSA less than 0.3 ng/mL) probability by preoperative PSA group.
PREOPERATIVE PSA AND PFS RATES The 10-year PFS estimate for men with a preoperative PSA level of 2.6 to 4.0, 4.1 to 7.0, 7.1 to 10.0, and greater than 10.0 ng/mL was 88%, 80%, 76%, and 61%, respectively (Fig. 1 and Table III). Overall, 5% of men with a PSA level of 2.6 to 4.0 ng/mL developed progression, and 26% of men with a PSA level greater than 10.0 ng/mL did so (Table III). The log-rank test for trend comparing the PFS rates across the preoperative PSA strata was statistically significant, but the log-rank test comparing a PSA level of 2.6 to 4.0 ng/mL to one of 4.1 to 7.0 ng/mL did not quite reach statistical significance (Table III). Computing the 10-year PFS estimates for each preoperative PSA stratum by ei158
Our study is, to our knowledge, the largest published study to date demonstrating that among men with clinical Stage T1c prostate cancer, those with a PSA level of 2.6 to 4.0 ng/mL have the greatest rate of organ-confined tumors, lowest pathologic Gleason scores, and greatest 10-year PFS rate. Among men with normal DRE findings and a PSA level of 2.6 to 4.0 ng/mL, prostate cancer is detected by sextant biopsy in 22% to 25%.2 The goal of prostate cancer screening is to identify clinically significant tumors that are still organ confined and curable. The rationale for PSA screening at levels of 2.6 to 4.0 ng/mL derives from two observations. First, clinically significant disease as defined by the pathologic tumor features is common in these men, and second, the prevalence of high-risk features is lower compared with those found using the traditional PSA screening threshold of 4.0 ng/mL. In our study, potentially aggressive disease, defined as positive margins (⬃15%), extraprostatic tumor extension (⬃20%), and Gleason grade of 7 or greater (⬃25%), was common in men with a PSA level of 2.6 to 4.0 ng/mL, although less so than in men with a PSA level greater than 4.0 ng/mL (positive margins 21%, extraprostatic disease 31%, and Gleason grade of 7 or greater 37%). Among men with a PSA level of 2.6 to 4.0 ng/mL, 81% had pathologically organ-confined UROLOGY 66 (1), 2005
TABLE III. Preoperative PSA range and progression-free survival Preoperative PSA (ng/mL) 10-yr PFS (%) Progression (%) 5-yr PFS (%)
2.6–4.0
4.1–7.0
7.1–10.0
>10.0
P Value
88* (82–94) 5 92 (90–95)
80* (76–84) 9 89 (87–91)
76 (67–85) 13 87 (83–91)
61 (53–69) 26 75 (70–80)
0.0001† 0.001‡ 0.0001†
KEY: PSA ⫽ prostate-specific antigen; PFS ⫽ progression-free survival. Data in parentheses are 95% confidence intervals. * PFS for PSA 2.6 – 4.0 vs. 4.1–7.0 ng/mL (P ⫽ 0.06). † Log-rank test. ‡ Cochran-Armitage test for trend.
disease; men with greater PSA levels had a lower rate of organ-confined disease. A previous study by Catalona et al.2 showed that 22% of men who underwent biopsy for a PSA level between 2.6 and 4.0 ng/mL were diagnosed with cancer. All of these cancers were clinically localized and 81% were organ confined.2 Other investigators have also shown that men diagnosed with prostate cancer with a PSA level of less than 4.0 ng/mL have a greater incidence of favorable features, such as organ-confined disease,4,10 –14 cancer-free surgical margins,4,11–13 and lower cancer volume.10,11 An argument against lowering the PSA screening threshold for biopsy to less than 4.0 ng/mL is the possible risk of overdetecting clinically insignificant tumors. However, Krumholtz et al.11 showed that screening for prostate cancer in the 2.6 to 4.0 ng/mL PSA range resulted in more frequent detection of small, organ-confined tumors without overdetection of “clinically insignificant” or “clinically unimportant” cancers. Lowering the PSA cutoff to 2.6 ng/mL also might mean subjecting more men to unnecessary biopsies compared with using a 4.0 ng/mL PSA cutoff. Although avoiding unnecessary biopsies is desirable, studies have shown that by using a 4.0 ng/mL PSA cutoff, a substantial proportion of cancers that would have been detected at lower PSA cutoffs are either missed or detected later.2 The extent to which delayed detection affects outcomes has not been clearly defined; however, our results suggest that delayed detection might compromise PFS. In our study, a lower preoperative PSA level (2.6 to 4.0 ng/mL) was associated with a greater PFS rate. For each strata of preoperative PSA studied, as the PSA level increased, the PFS rate decreased. Other investigators have also evaluated the role of preoperative PSA and PFS after RRP. Our findings are similar to those of many other large prostatectomy series.3,12,14,15 Han et al.3 presented a nomogram for 845 men with clinical Stage T1c disease. They used a PSA level of greater than 0.2 ng/mL as a criterion for cancer progression, grouped patients on the basis UROLOGY 66 (1), 2005
of the preoperative PSA level (0 to 4.0, 4.1 to 10, 10.1 to 20, and greater than 20 ng/mL), and stratified patients by biopsy Gleason score. The PFS estimations at 3, 5, 7, and 10 years for men with a PSA level of 0 to 4.0 ng/mL were consistently greater for every Gleason grade compared with men with a PSA level of 4.1 to 10.0 ng/mL. Similar results were demonstrated among men with organconfined disease. Men with a PSA level of less than 4.0 ng/mL, regardless of Gleason grade, uniformly had greater PFS rates than men with a PSA level of 4.1 to 10.0 ng/mL. Shekarriz et al.12 examined the disease-free survival of men after RRP. For clinical Stage T1c and T2 cancer, the disease-free survival rate for men with a preoperative PSA level less than 4.0 ng/mL was 95% compared with 85% for those with a PSA level greater than 4.0 ng/mL. Hull et al.4 reported that men with a PSA level less than 4.0 ng/mL had an 89% 10-year PFS rate compared with 84% for men with a PSA level between 4.0 and 9.9 ng/mL. In contrast, D’Amico et al.5 assessed PFS after RRP and stratified by pathologic Gleason grade. For men with Gleason grade of 6 or less, no difference in PFS was noted on the basis of the preoperative PSA level. However, for men with a Gleason grade of 7 or greater, those in the lowest PSA range (2.6 to 4.0 ng/mL) had worse PFS than those with greater PSA levels (4.1 to 10.0 ng/mL). These results reflect the aggressive nature of high-grade cancer in some men with low PSA levels. The investigators suggested that lower serum testosterone levels may be associated with aggressive disease in the setting of low PSA levels. Alternatively, some poorly differentiated prostate cancers do not produce high levels of PSA until they are advanced, if ever. Stamey16 reported no material difference in progression among patients with a PSA level less than 9 ng/mL. He concluded that in this PSA range, the prognosis was unrelated to the PSA level. Several aspects of his study suggest possible alternative explanations for these results. First, his series included patients with palpable (Stage T2) tumors. Some palpable low-PSA producing tumors are 159
highly aggressive, and their inclusion would negatively bias the prognosis of patients in the low PSA range. The PFS estimates in his study were lower than observed in most contemporary studies, including ours. Second, the follow-up in his series was short; thus, some patients at the higher end of this PSA range who would be expected to be at greater risk of progression might not have had time to manifest biochemical evidence of progression. Third, PFS was plotted as a function of the logarithm of the PSA level. Thus, even if a linear relationship between PSA level and PFS existed, a curved function plot similar to that obtained by Stamey16 would result. Our study had several limitations. One was our predominantly white study group. The results of our study may not be applicable to men of other ethnic backgrounds. Another limitation was the variable follow-up intervals among the different PSA strata. Men with a PSA level of 2.6 to 4.0 ng/mL had the shortest follow-up; however, differences in PFS were also evident at 3 and 5 years of follow-up (3-year data not shown). Although differences in disease-specific and overall survival had not yet been demonstrated, PSA PFS is usually an intermediate-term marker for these endpoints. Finally, we recognize that our longitudinal follow-up study is not the optimal method for analyzing how lowering the PSA cutoff for screening of prostate cancer affects treatment outcomes. A randomized clinical trial using different PSA cutoffs would be more valid. Nevertheless, our large study cohort with long-term follow-up provides important insights into likely outcome trends. CONCLUSIONS Among men with clinical Stage T1c prostate cancer, those with a PSA level of 2.6 to 4.0 ng/mL have the greatest rate of organ-confined disease, lowest pathologic Gleason grade, and greatest 10-year PFS rate. REFERENCES 1. Babaian RJ, Johnston DA, Naccarato W, et al: The incidence of prostate cancer in a screening population with a serum prostate specific antigen between 2.5 and 4.0 ng/ml: relation to biopsy strategy. J Urol 165: 757–760, 2001.
160
2. Catalona WJ, Smith DS, and Ornstein DK: Prostate cancer detection in men with serum PSA concentrations of 2.6 to 4.0 ng/mL and benign prostate examination: enhancement of specificity with free PSA measurements. JAMA 277: 1452– 1455, 1997. 3. Han M, Partin AW, Zahurak M, et al: Biochemical (prostate specific antigen) recurrence probability following radical prostatectomy for clinically localized prostate cancer. J Urol 169: 517–523, 2003. 4. Hull GW, Rabbani F, Abbas F, et al: Cancer control with radical prostatectomy alone in 1,000 consecutive patients. J Urol 167: 528 –534, 2002. 5. D’Amico AV, Chen MH, Malkowicz SB, et al: Lower prostate specific antigen outcome than expected following radical prostatectomy in patients with high grade prostate and a prostatic specific antigen level of 4 ng/ml. or less. J Urol 167: 2025–2030, 2002. 6. Smith DS, Humphrey PA, and Catalona WJ: The early detection of prostate carcinoma with prostate specific antigen: the Washington University experience. Cancer 80: 1852–1856, 1997. 7. Catalona WJ, Smith DS, Ratliff TL, et al: Measurement of prostate-specific antigen in serum as a screening test for prostate cancer. N Engl J Med 324: 1156 –1161, 1991. 8. Epstein JI, Walsh PC, Carmichael M, et al: Pathologic and clinical findings to predict tumor extent of nonpalpable (stage T1c) prostate cancer. JAMA 271: 368 –374, 1994. 9. Ohori M, Wheeler TM, Dunn JK, et al: The pathological features and prognosis of prostate cancer detectable with current diagnostic tests. J Urol 152: 1714 –1720, 1994. 10. Schroder FH, van der Cruijsen-Koeter I, de Koning HJ, et al: Prostate cancer detection at low prostate specific antigen. J Urol 163: 806 – 812, 2000. 11. Krumholtz JS, Carvalhal GF, Ramos CG, et al: Prostatespecific antigen cutoff of 2.6 ng/mL for prostate cancer screening is associated with favorable pathologic tumor features. Urology 60: 469 – 473, 2002. 12. Shekarriz B, Upadhyay J, Bianco FJ Jr, et al: Impact of preoperative serum PSA level from 0 to 10 ng/ml on pathological findings and disease-free survival after radical prostatectomy. Prostate 48: 136 –143, 2001. 13. Berger AP, Volgger H, Rogatsch H, et al: Screening with low PSA cutoff values results in low rates of positive surgical margins in radical prostatectomy specimens. Prostate 53: 241–245, 2002. 14. Aleman M, Karakiewicz PI, Kupelian P, et al: Age and PSA predict likelihood of organ-confined disease in men presenting with PSA less than 10 ng/mL: implications for screening. Urology 62: 70 –74, 2003. 15. Freedland SJ, Terris MK, Csathy GS, et al: Preoperative model for predicting prostate specific antigen recurrence after radical prostatectomy using percent of biopsy tissue with cancer, biopsy Gleason grade and serum prostate specific antigen. J Urol 171: 2215–2220, 2004. 16. Stamey TA: Preoperative serum prostate-specific antigen (PSA) below 10 g/l predicts neither the presence of prostate cancer nor the rate of postoperative PSA failure. Clin Chem 47: 631– 634, 2001.
UROLOGY 66 (1), 2005