- Email: [email protected]
Deciding whom to biopsy
Urologic Oncology: Seminars and Original Investigations 28 (2010) 542–545
Seminar article
Deciding whom to biopsy Christopher L. Amling, M.D.a,*, William J. Catalona, M.D.b, Eric A. Klein, M.D.c a
Division of Urology, Oregon Health and Science University, Portland, OR 97239, USA b Department of Urology, Northwestern University, Chicago, IL 60611, USA c Glickman Urological and Kidney Institute, Cleveland Clinic Foundation, Cleveland, OH 44195, USA
Abstract Biopsy results from the Prostate Cancer Prevention Trial (PCPT) showed that prostate cancer exists at all PSA levels and that a significant number of men with “normal” PSA levels have high grade cancer. These findings and the low specificity of total PSA in discriminating cancer from benign disease have added to the debate about how best to use PSA in selecting men for prostate biopsy. Lower PSA thresholds for consideration of biopsy, particularly in younger men, are advocated by some. PSA velocity measurements may assist in the identification of men most likely to harbor cancer, and lower PSA velocity thresholds may be more appropriate in younger men. A more individualized approach using a predictive model developed from PCPT biopsy results is promoted by others. While able to incorporate risk variables other than PSA, including new markers, this risk calculator does not include PSA velocity since this variable was not found to have independent predictive value in this model. This article will present differing viewpoints on selecting men for prostate biopsy, one advocating the use of a PSA cut-off or PSA velocity measure (Dr. Catalona) and the other arguing for the routine use of established risk nomograms (Dr. Klein). © 2010 Elsevier Inc. All rights reserved. Keywords: Prostate cancer; Prostate biopsy; Prostate specific antigen
Introduction New research on PSA and the advent of predictive nomograms have made selection of men for prostate biopsy a more complex decision than it was in the past. While a PSA threshold of 4 ng/ml has traditionally been accepted as the level at which to recommend prostate biopsy, more recently PSA kinetics and lower PSA thresholds have been advocated. Deciding whom to biopsy has become even more complicated with the publication of biopsy results from the Prostate Cancer Prevention Trial (PCPT) [1]. This study demonstrated that prostate cancer exists at all PSA levels and that instead of risk thresholds, PSA represents a continuum of risk. In addition, a significant number of men with “normal” PSA levels have high-grade cancers. Further analysis of these study results suggested that PSA by itself may not be the only factor to consider when selecting men for prostate biopsy. This led to the development of a biopsydetectable prostate cancer risk prediction tool, which combined several risk factors with PSA level in a predictive nomogram.
* Corresponding author. Tel.: ⫹1-503-418-9132; fax: ⫹1-503-346-1501. E-mail address: [email protected] (C.L. Amling). 1078-1439/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.urolonc.2010.05.012
Lower PSA thresholds are generally used to consider biopsy in younger men. It has also been recognized that a lower PSA velocity threshold may indicate the presence of prostate cancer at these lower PSA levels. While it is generally accepted that a PSA velocity of 0.75 ng/ml/y is a suspicious marker for prostate cancer in men with total PSA levels between 4 and 10 ng/ml, for men with PSA levels ⬍ 4 ng/ml, PSA velocities ⬎ 0.35 to 0.40 ng/ml/y may be more applicable [2– 4]. These findings and the low specificity of PSA for identifying men with prostate cancer has fueled debate about whether men should be selected for prostate biopsy based on a PSA cut-off level or whether biopsy candidates should instead be selected using recently established risk nomograms. In this article, the case for using a PSA cut-off for selection of men for prostate biopsy will be presented by Dr. Catalona and the argument for using predictive nomograms will be presented by Dr. Klein.
Biopsy candidates should be selected by PSA cut-off Although it is now clear from the PCPT that PSA level represents a continuum of prostate cancer risk, a PSA cutoff of 2.5 ng/ml may be the best threshold at which to recommend initial prostate biopsy. In the original large PSA
C.L. Amling et al. / Urologic Oncology: Seminars and Original Investigations 28 (2010) 542–545
screening study of 24,000 men screened at 6-month intervals, a nearly identical percent of men with PSAs between 2.6 and 4 ng/ml and 4 to 10 ng/ml were found upon biopsy to have prostate cancer [5]. A lower PSA threshold for recommending biopsy, particularly in younger men, may improve the clinical value of the PSA test as well. In a study of 6,691 men who underwent PSA-based screening for prostate cancer between 1995 and 2001, 705 (11%) underwent prostate biopsy [6]. Adjusting for verification bias significantly increased the overall diagnostic performance of the PSA test, particularly in younger men. If the threshold PSA value for undergoing biopsy were set at 4.1 ng/ml, 82% of cancers in younger men and 65% of cancers in older men would be missed. This and other studies validate the use of a lower PSA threshold (PSA 2.5 ng/ml) to select men for initial prostate biopsy. When cancer detected at PSA values between 2.6 and 4 ng/ml are compared with those detected at levels between 4 and 10 ng/ml, they have smaller volumes, are equally “significant,” and are more likely to be organ-confined. Krumholtz et al. evaluated the pathologic characteristics of screen-detected clinical stage T1 prostate cancers in 94 patients undergoing radical prostatectomy, comparing cancers detected in the 2.6 – 4 ng/ml range to the 4.1–10 ng/ml range [7]. Cancers detected at the 2.6 – 4 ng/ml PSA range had smaller volumes (1.1 vs. 1.8 cc) and were significantly more likely to be organ-confined (88% vs. 63%). Of note, there was no difference in the proportion of insignificant cancers detected (12% vs. 12% “insignificant”). Another study of the same screening population examined progression-free survival as a function of preoperative PSA level in clinical T1c radical prostatectomy patients [8]. The men with PSAs between 2.6 and 4 ng/ml had an 81% chance of having organ-confined disease compared with 74% in men with PSAs between 4.1 and 7 ng/ml. Ten-year progressionfree survival was also better in the men with lower preoperative PSA levels. The PCPT study confirmed that a significant percentage of men with PSA levels between 2.1 and 4 ng/ml are found to have cancer if biopsy is performed, and a significant number of these are found to have high grade disease [1]. In a follow-up study evaluating the pathologic features of these cancers in men who elected to undergo radical prostatectomy, Lucia et al. compared cancers with preoperative PSAs of 2.6 – 4 ng/ml to those with PSAs of 4 –10 ng/ml [9]. A similar number of cancers in these 2 groups were found to be high grade (Gleason score ⱖ 7) and only 18% of cancers in the 2.6 – 4 ng/ml PSA range met the criteria for insignificant disease. The pathologic characteristics of clinical T1c cancers were similarly compared between preoperative PSA groups 2.6 – 4 ng/ml and 4.1– 6 ng/ml in a series of 2,896 men undergoing radical prostatectomy at Johns Hopkins Hospital [10]. In this study, patients with PSA ⬍ 4 ng/ml had fewer positive margins, less risk of extraprostatic extension, and a lesser likelihood of having Gleason ⱖ 7 disease. As a whole, these studies suggest that detection of
543
prostate cancer at PSA thresholds below 4 ng/ml results in cancers more readily curable and, in most cases, these cancers represent significant disease. A PSA threshold of 2.5 ng/ml to select men for initial prostate biopsy is especially appropriate for younger men, who as a group have median PSA levels much lower than that. Loeb et al. studied baseline PSA levels in a group of approximately 14,000 men younger than 60 years of age who participated in a prostate cancer screening study between 1991 and 2001 [11]. Men aged 40 – 49 years were eligible for this study if they had a positive family history of prostate cancer or were of African-American heritage. Men older than 50 were screened without respect to risk factors. The median PSA level was 0.7 ng/ml for men aged 40 – 49 years and 0.9 ng/ml for men aged 50 –59 years. The risk of cancer was very low for men below the median PSA level while the risk increased as PSA increased above the median. A baseline PSA level between the median and 2.5 ng/ml was associated with a 14.6-fold and 7.6-fold increased risk of prostate cancer in men aged 40 – 49 years and 50 –59 years, respectively. Additionally, the risk of aggressive disease also increased as the PSA level increased above this median-for-age level. This study showed that in young men, a baseline PSA value is a strong predictor of prostate cancer, and this risk was proportional to the degree of PSA elevation above the median PSA for that age group. PSA velocity is also useful in selecting men with “lower” PSA levels who may be at higher risk for harboring cancer. It is important to realize that PSA velocity correlates with total PSA level and as such, lower PSA velocity thresholds should be used to select men for biopsy when total PSA is ⬍4 ng/ml. Carter et al. investigated the use of PSA velocity at normal PSA levels for potential detection of life-threatening prostate cancer during a window of curability [3]. Using the Baltimore Longitudinal Study of Aging population, 980 men had PSA velocity determined. The relative risks of prostate cancer death and prostate cancer-specific survival were stratified according to PSA velocity. PSA velocity measured 10 –15 years before prostate cancer diagnosis when most men had PSA levels below 4 ng/ml was associated with cancer-specific survival 25 years later. Survival was 92% among men with PSA velocity of 0.35 ng/ml/y or less, and 54% among men with PSA velocity above 0.35 ng/ml/y. Men with PSA velocity above 0.35 ng/ml/y had a higher relative risk of prostate cancer death as well. This study suggests that a PSA velocity threshold of 0.35 ng/ml may help identify men with life-threatening prostate cancer during a period when their PSA levels are associated with the presence of curable disease. Carter et al. also introduced the PSA velocity “risk count” concept as a way to potentially identify men at risk for the development of high risk disease [12]. Using this method, the number of times the PSA velocity exceeded a threshold (PSA velocity risk count) was determined in 717 men over 10 –20 years. The probability of high risk disease increased directly with the risk count, with a relative risk for developing high risk
544
C.L. Amling et al. / Urologic Oncology: Seminars and Original Investigations 28 (2010) 542–545
disease of 1.49 (95% CI, 1.29 –1.71) using a PSA velocity cut-point of 0.4 ng/ml/y. This method of PSA history interpretation could help to identify those men destined to develop high risk disease who would benefit from diagnosis and treatment of their prostate cancer at PSA levels more likely associated with curable disease. To summarize, it is recommended that PSA testing begin at age 40 years and then be repeated annually. If the PSA is higher than the age-specific median level, repeat testing is warranted to verify this. Should there be concerns that BPH might be confounding the PSA determination, a PSA density (cut-off ⬎ 0.1) or a percent free PSA (cut-off 10%) could be used to either increase or decrease the level of suspicion. To rule out prostatitis, a course of antibiotic therapy and repeat testing might be warranted, but may not be appropriate in all situations. If PSA elevation is confirmed and does not decrease substantially with antibiotics or decrease substantially on follow-up, a prostate biopsy is recommended. Alternatively, PSA could be monitored at 3– 6 month intervals to assess PSA velocity with a PSA velocity of 0.35 ng/ml/y being a suspicious threshold at which biopsy should be strongly recommended. These biopsy recommendations are consistent with the 2009 NCCN Prostate Cancer Early Detection guidelines for men with PSA between 2.6 and 4 ng/ml or PSA velocity ⱖ 0.35 ng/ml/y. When PSA is ⬍2.5 ng/ml, biopsy should be considered when PSA velocity is ⱖ0.35 ng/ml/y. In the future, measurements of [-2] proPSA will be useful in improving the accuracy of PSA testing. Proenzyme forms of prostate specific antigen have been reported to be a more prostate cancer-specific marker than total PSA level [13–15]. In a recently published prospective prostate cancer screening study of over 2,000 men with total PSA between 2.5 and 10 ng/ml, the percent [-2]proPSA (ratio of [-2]proPSA to free PSA) had significantly better specificity for prostate cancer detection compared with total PSA and percent free PSA levels [14]. Another recent European study showed similar results with percent [-2]proPSA demonstrating significantly increased prostate cancer predictive value and specificity compared to both total PSA and percent free PSA levels [15]. When available for routine use, this marker’s ability to improve discrimination between prostate cancer and benign disease will further help in the appropriate selection of men for prostate biopsy.
Biopsy candidates should be selected using nomograms It is important to recognize first that PSA-based prostate cancer screening has significant limitations. While PSA is prostate-specific, it is not prostate cancer-specific and most men with elevated PSA do not have prostate cancer. It is also important to distinguish between biologically significant vs. clinically relevant disease. Biologically significant disease is cancer destined to cause significant morbidity and mortality. Clinically relevant disease causes anxiety or leads
to treatment with the potential for treatment-related morbidity. In the context of prostate cancer screening, it is important to note that a small proportion of men truly benefit from treatment and it is the identification of these men that should be the goal when selecting men for prostate biopsy. The PCPT provided a large set of data useful for developing nomograms to use in the selection of men for prostate biopsy [1]. In this trial 18,882 men were randomly assigned to receive either finasteride or placebo and were followed up with annual digital rectal examinations (DRE) and PSA measurements. Prostate biopsy was recommended “for cause” if the result of either test was abnormal. After 7 years of study, men who had not been diagnosed with prostate cancer were asked to undergo an end-of-study biopsy. These men were biopsied at PSA levels ⱕ 4 ng/ml with a normal DRE result. These end-of-study biopsies in men with PSA levels ⬍ 4 ng/ml showed that there was a continuum of risk for prostate cancer detection that correlated directly with PSA level in this range. The risk of finding high grade disease was also continuous with PSA level in that high grade disease was more likely at higher PSA levels. Utilizing the biopsy results of 5,519 men from the placebo group who underwent prostate biopsy, a predictive model of prostate cancer detection was developed [16]. Logistic regression was used to model the risk of prostate cancer and high grade disease associated with age at biopsy, race, family history of prostate cancer, PSA level, PSA velocity, DRE result, and previous prostate biopsy. Risk equations were created from the estimated logistic regression models. Variables that predicted prostate cancer included higher PSA level, positive family history of prostate cancer, and abnormal digital rectal examination result. Previous negative prostate biopsy was associated with reduced risk. Of note, neither age at biopsy nor PSA velocity contributed independent prognostic information. Higher PSA level, abnormal DRE result, older age at biopsy and African-American race were predictive for high grade disease, whereas previous negative prostate biopsy reduced the risk. Since finasteride use was found to improve the accuracy of PSA for detection of any cancer as well as high grade cancer, the PCPT risk calculator also accounts for the effect of finasteride on PSA performance [17]. The predictive model utilizing PCPT biopsy results allows an individualized assessment of prostate cancer risk and the risk of high grade disease for men who are candidates for prostate biopsy. The advantages of a risk calculator like this are that it captures all of the predictive values and continuous variables such as age and PSA, it can be validated on multiple cohorts, and it is flexible, allowing rapid evaluation and the incorporation of new markers. The risk calculator can be accessed online at http:// deb.uthscsa.edu/URORiskCalc/Pages/calcs.jsp. A potential limitation of the PCPT risk calculator is that it was developed from a cohort of men who entered the trial with a PSA of 3.0 ng/ml or less, a normal DRE, and age 55 years or older. Most men presenting for prostate biopsy
C.L. Amling et al. / Urologic Oncology: Seminars and Original Investigations 28 (2010) 542–545
have an increased PSA level or an abnormal DRE and many are younger than 55 years of age. As such, it is important to validate risk calculators to determine if it can be widely applied and useful in men seeking care in the routine American urology practice setting. To validate the PCPT risk calculator, Eyre et al. utilized data from patients undergoing prostate biopsy by 12 urologists at 5 sites who were enrolled in the Early Detection Research Network (EDRN) study [18]. The PCPT risk calculator was validated by examining the area under the receiver operating characteristic (ROC) curve, sensitivity, specificity, and calibration comparing observed vs. predicted risk of prostate cancer detection. This analysis revealed that the EDRN and PCPT cohorts were significantly different. Cancer incidence was greater in the EDRN validation cohort (43%) compared with the PCPT cohort (22%). Furthermore, the EDRN participants were younger and more racially diverse than PCPT participants. Prostate cancer severity was worse in the EDRN cohort as well. Despite these differences, however, the PCPT risk calculator was superior to PSA alone for predicting cancer in the EDRN cohort, validating this risk calculator’s performance in a multi-center urology practice setting. An additional advantage of a risk calculator over a cutpoint threshold for determining who to biopsy is the ability to incorporate new markers. An example of this is the recent incorporation of prostate cancer gene 3 (PCA3) into the PCPT risk calculator to improve its diagnostic accuracy [19]. Based on a cohort of 522 men who underwent prostate biopsy with measurements of urinary PCA3, serum PSA, DRE, and biopsy history, this marker was incorporated into the risk assessment. External validation of the updated risk calculator was performed on the cohort of 443 European patients and compared with PCPT risk estimates. Incorporation of PCA3 improved the diagnostic accuracy of the PCPT risk calculator demonstrating that new markers for prostate cancer can be incorporated relatively easily into existing risk calculators to improve the ability to select men for prostate biopsy. In the future, additional biologic markers such as EPCA2, gene fusions, and metabolomics may be similarly incorporated into existing risk calculators. In summary, selection of men using a PSA cut-off, PSA density or PSA velocity threshold should really be something of the past as this selects men for biopsy based on the risk of any cancer. Currently, we should use continuous risk models to select men for biopsy. These include nomograms and risk calculators in which new markers can be incorporated as they are detected to improve their diagnostic accuracy. Using these methods, the risk of any biologically significant cancer can be better determined. In the future, individual risk assessment will be possible utilizing PSA at a young age and genetic testing in men who are candidates for biopsy. The hope then will be that the risk of a truly biologically significant cancer will be determined and that selective biopsy can be recommended. This assessment of risk should also lead to development of prostate cancer prevention strategies.
545
References [1] Thompson IM, Pauler DK, Goodman PJ, et al. Prevalence of prostate cancer among men with a prostate-specific antigen level ⱕ 4.0 ng per milliliter. N Engl J Med 2004;350:2239 – 46. [2] Carter HB, Pearson JD. PSA velocity for the diagnosis of early prostate cancer. A new concept Urol Clin North Am 1993;20:665–70. [3] Carter HB, Ferrucci L, Kettermann A, et al. Detection of life-threatening prostate cancer with prostate-specific antigen velocity during a window of curability. J Nat Cancer Inst 2006;98:1521–7. [4] Loeb S, Roehl KA, Helfand BT, et al. Can prostate specific antigen velocity thresholds decrease insignificant prostate cancer detection? J Urol 2010;183:112–7. [5] Smith DS, Catalona WJ. The nature of prostate cancer detected through prostate specific antigen based screening. J Urol 1994;152: 1732– 6. [6] Punglia RS, D’Amico AV, Catalona WJ, et al. Effect of verification bias on screening for prostate cancer by measurement of prostatespecific antigen. N Engl J Med 2003;349:335– 42. [7] Krumholtz JS, Carvalhal GF, Ramos CG, et al. Prostate-specific antigen cutoff of 2.6 ng/ml for prostate cancer screening is associated with favorable pathologic tumor features. Urol 2002;60:469 –73. [8] Antenor JV, Roehl KA, Eggener SE, et al. Preoperative PSA and progression-free survival after radical prostatectomy for stage T1c disease. Urol 2005;66:156 – 60. [9] Lucia MS, Darke AK, Goodman PJ, et al. Pathologic characteristics of cancers detected in the Prostate Cancer Prevention Trial: Implications for prostate cancer detection and chemoprevention. Cancer Prev Res 2008;1:167–73. [10] Makarov DV, Humphreys EB, Mangold LA, et al. Pathological outcomes and biochemical progression in men with T1c prostate cancer undergoing radical prostatectomy with prostate specific antigen 2.6 to 4.0 vs. 4.1 to 6.0 ng/ml. J Urol 2006;176:554 – 8. [11] Loeb S, Roehl KA, Antenor JV, et al. Baseline prostate-specific antigen compared with median prostate-specific antigen for age group as predictor of prostate cancer risk in men younger than 60 years. Urology 2006;67:316 –20. [12] Carter HB, Kettermann A, Ferrucci L, et al. Prostate-specific antigen velocity risk count assessment: A new concept for detection of lifethreatening prostate cancer during window of curability. Urology 2007;70:685–90. [13] Catalona WJ, Bartsch G, Rittenhouse HG, et al. Serum pro-prostate specific antigen preferentially detects aggressive prostate cancers in men with 2 to 4 ng/ml prostate specific antigen. J Urol 2004;171: 2239 – 44. [14] Le BV, Griffin CR, Loeb S, et al. [-2]Proenzyme prostate specific antigen is more accurate than total and free prostate specific antigen in differentiating prostate cancer from benign disease in a prospective prostate cancer screening study. J Urol 2010;183:1355–9. [15] Jansen FH, van Schaik RH, Kurstjens J, et al. Prostate-specific antigen (PSA) isoform p2PSA in combination with total PSA and free PSA improves diagnostic accuracy in prostate cancer detection. Eur Urol 2010 Feb 13 [Epub ahead of print]. [16] Thompson IM, Ankerst DP, Chi C, et al. Assessing prostate cancer risk: Results from the Prostate Cancer Prevention Trial. J Nat Cancer Inst 2006;98:529 –34. [17] Thompson IM, Chi C, Ankerst DP, et al. Effect of finasteride on the sensitivity of PSA for detecting prostate cancer. J Nat Cancer Inst 2006;98:1128 –33. [18] Eyre SJ, Ankerst DP, Wei JT, et al. Validation in a multiple urology practice cohort of the Prostate Cancer Prevention Trial calculator for predicting prostate cancer detection. J Urol 2009;182:2653– 8. [19] Ankerst DP, Groskopf J, Day JR, et al. Predicting prostate cancer risk through incorporation of prostate cancer gene 3. J Urol 2008;180: 1303– 8.
Seminar article
Deciding whom to biopsy Christopher L. Amling, M.D.a,*, William J. Catalona, M.D.b, Eric A. Klein, M.D.c a
Division of Urology, Oregon Health and Science University, Portland, OR 97239, USA b Department of Urology, Northwestern University, Chicago, IL 60611, USA c Glickman Urological and Kidney Institute, Cleveland Clinic Foundation, Cleveland, OH 44195, USA
Abstract Biopsy results from the Prostate Cancer Prevention Trial (PCPT) showed that prostate cancer exists at all PSA levels and that a significant number of men with “normal” PSA levels have high grade cancer. These findings and the low specificity of total PSA in discriminating cancer from benign disease have added to the debate about how best to use PSA in selecting men for prostate biopsy. Lower PSA thresholds for consideration of biopsy, particularly in younger men, are advocated by some. PSA velocity measurements may assist in the identification of men most likely to harbor cancer, and lower PSA velocity thresholds may be more appropriate in younger men. A more individualized approach using a predictive model developed from PCPT biopsy results is promoted by others. While able to incorporate risk variables other than PSA, including new markers, this risk calculator does not include PSA velocity since this variable was not found to have independent predictive value in this model. This article will present differing viewpoints on selecting men for prostate biopsy, one advocating the use of a PSA cut-off or PSA velocity measure (Dr. Catalona) and the other arguing for the routine use of established risk nomograms (Dr. Klein). © 2010 Elsevier Inc. All rights reserved. Keywords: Prostate cancer; Prostate biopsy; Prostate specific antigen
Introduction New research on PSA and the advent of predictive nomograms have made selection of men for prostate biopsy a more complex decision than it was in the past. While a PSA threshold of 4 ng/ml has traditionally been accepted as the level at which to recommend prostate biopsy, more recently PSA kinetics and lower PSA thresholds have been advocated. Deciding whom to biopsy has become even more complicated with the publication of biopsy results from the Prostate Cancer Prevention Trial (PCPT) [1]. This study demonstrated that prostate cancer exists at all PSA levels and that instead of risk thresholds, PSA represents a continuum of risk. In addition, a significant number of men with “normal” PSA levels have high-grade cancers. Further analysis of these study results suggested that PSA by itself may not be the only factor to consider when selecting men for prostate biopsy. This led to the development of a biopsydetectable prostate cancer risk prediction tool, which combined several risk factors with PSA level in a predictive nomogram.
* Corresponding author. Tel.: ⫹1-503-418-9132; fax: ⫹1-503-346-1501. E-mail address: [email protected] (C.L. Amling). 1078-1439/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.urolonc.2010.05.012
Lower PSA thresholds are generally used to consider biopsy in younger men. It has also been recognized that a lower PSA velocity threshold may indicate the presence of prostate cancer at these lower PSA levels. While it is generally accepted that a PSA velocity of 0.75 ng/ml/y is a suspicious marker for prostate cancer in men with total PSA levels between 4 and 10 ng/ml, for men with PSA levels ⬍ 4 ng/ml, PSA velocities ⬎ 0.35 to 0.40 ng/ml/y may be more applicable [2– 4]. These findings and the low specificity of PSA for identifying men with prostate cancer has fueled debate about whether men should be selected for prostate biopsy based on a PSA cut-off level or whether biopsy candidates should instead be selected using recently established risk nomograms. In this article, the case for using a PSA cut-off for selection of men for prostate biopsy will be presented by Dr. Catalona and the argument for using predictive nomograms will be presented by Dr. Klein.
Biopsy candidates should be selected by PSA cut-off Although it is now clear from the PCPT that PSA level represents a continuum of prostate cancer risk, a PSA cutoff of 2.5 ng/ml may be the best threshold at which to recommend initial prostate biopsy. In the original large PSA
C.L. Amling et al. / Urologic Oncology: Seminars and Original Investigations 28 (2010) 542–545
screening study of 24,000 men screened at 6-month intervals, a nearly identical percent of men with PSAs between 2.6 and 4 ng/ml and 4 to 10 ng/ml were found upon biopsy to have prostate cancer [5]. A lower PSA threshold for recommending biopsy, particularly in younger men, may improve the clinical value of the PSA test as well. In a study of 6,691 men who underwent PSA-based screening for prostate cancer between 1995 and 2001, 705 (11%) underwent prostate biopsy [6]. Adjusting for verification bias significantly increased the overall diagnostic performance of the PSA test, particularly in younger men. If the threshold PSA value for undergoing biopsy were set at 4.1 ng/ml, 82% of cancers in younger men and 65% of cancers in older men would be missed. This and other studies validate the use of a lower PSA threshold (PSA 2.5 ng/ml) to select men for initial prostate biopsy. When cancer detected at PSA values between 2.6 and 4 ng/ml are compared with those detected at levels between 4 and 10 ng/ml, they have smaller volumes, are equally “significant,” and are more likely to be organ-confined. Krumholtz et al. evaluated the pathologic characteristics of screen-detected clinical stage T1 prostate cancers in 94 patients undergoing radical prostatectomy, comparing cancers detected in the 2.6 – 4 ng/ml range to the 4.1–10 ng/ml range [7]. Cancers detected at the 2.6 – 4 ng/ml PSA range had smaller volumes (1.1 vs. 1.8 cc) and were significantly more likely to be organ-confined (88% vs. 63%). Of note, there was no difference in the proportion of insignificant cancers detected (12% vs. 12% “insignificant”). Another study of the same screening population examined progression-free survival as a function of preoperative PSA level in clinical T1c radical prostatectomy patients [8]. The men with PSAs between 2.6 and 4 ng/ml had an 81% chance of having organ-confined disease compared with 74% in men with PSAs between 4.1 and 7 ng/ml. Ten-year progressionfree survival was also better in the men with lower preoperative PSA levels. The PCPT study confirmed that a significant percentage of men with PSA levels between 2.1 and 4 ng/ml are found to have cancer if biopsy is performed, and a significant number of these are found to have high grade disease [1]. In a follow-up study evaluating the pathologic features of these cancers in men who elected to undergo radical prostatectomy, Lucia et al. compared cancers with preoperative PSAs of 2.6 – 4 ng/ml to those with PSAs of 4 –10 ng/ml [9]. A similar number of cancers in these 2 groups were found to be high grade (Gleason score ⱖ 7) and only 18% of cancers in the 2.6 – 4 ng/ml PSA range met the criteria for insignificant disease. The pathologic characteristics of clinical T1c cancers were similarly compared between preoperative PSA groups 2.6 – 4 ng/ml and 4.1– 6 ng/ml in a series of 2,896 men undergoing radical prostatectomy at Johns Hopkins Hospital [10]. In this study, patients with PSA ⬍ 4 ng/ml had fewer positive margins, less risk of extraprostatic extension, and a lesser likelihood of having Gleason ⱖ 7 disease. As a whole, these studies suggest that detection of
543
prostate cancer at PSA thresholds below 4 ng/ml results in cancers more readily curable and, in most cases, these cancers represent significant disease. A PSA threshold of 2.5 ng/ml to select men for initial prostate biopsy is especially appropriate for younger men, who as a group have median PSA levels much lower than that. Loeb et al. studied baseline PSA levels in a group of approximately 14,000 men younger than 60 years of age who participated in a prostate cancer screening study between 1991 and 2001 [11]. Men aged 40 – 49 years were eligible for this study if they had a positive family history of prostate cancer or were of African-American heritage. Men older than 50 were screened without respect to risk factors. The median PSA level was 0.7 ng/ml for men aged 40 – 49 years and 0.9 ng/ml for men aged 50 –59 years. The risk of cancer was very low for men below the median PSA level while the risk increased as PSA increased above the median. A baseline PSA level between the median and 2.5 ng/ml was associated with a 14.6-fold and 7.6-fold increased risk of prostate cancer in men aged 40 – 49 years and 50 –59 years, respectively. Additionally, the risk of aggressive disease also increased as the PSA level increased above this median-for-age level. This study showed that in young men, a baseline PSA value is a strong predictor of prostate cancer, and this risk was proportional to the degree of PSA elevation above the median PSA for that age group. PSA velocity is also useful in selecting men with “lower” PSA levels who may be at higher risk for harboring cancer. It is important to realize that PSA velocity correlates with total PSA level and as such, lower PSA velocity thresholds should be used to select men for biopsy when total PSA is ⬍4 ng/ml. Carter et al. investigated the use of PSA velocity at normal PSA levels for potential detection of life-threatening prostate cancer during a window of curability [3]. Using the Baltimore Longitudinal Study of Aging population, 980 men had PSA velocity determined. The relative risks of prostate cancer death and prostate cancer-specific survival were stratified according to PSA velocity. PSA velocity measured 10 –15 years before prostate cancer diagnosis when most men had PSA levels below 4 ng/ml was associated with cancer-specific survival 25 years later. Survival was 92% among men with PSA velocity of 0.35 ng/ml/y or less, and 54% among men with PSA velocity above 0.35 ng/ml/y. Men with PSA velocity above 0.35 ng/ml/y had a higher relative risk of prostate cancer death as well. This study suggests that a PSA velocity threshold of 0.35 ng/ml may help identify men with life-threatening prostate cancer during a period when their PSA levels are associated with the presence of curable disease. Carter et al. also introduced the PSA velocity “risk count” concept as a way to potentially identify men at risk for the development of high risk disease [12]. Using this method, the number of times the PSA velocity exceeded a threshold (PSA velocity risk count) was determined in 717 men over 10 –20 years. The probability of high risk disease increased directly with the risk count, with a relative risk for developing high risk
544
C.L. Amling et al. / Urologic Oncology: Seminars and Original Investigations 28 (2010) 542–545
disease of 1.49 (95% CI, 1.29 –1.71) using a PSA velocity cut-point of 0.4 ng/ml/y. This method of PSA history interpretation could help to identify those men destined to develop high risk disease who would benefit from diagnosis and treatment of their prostate cancer at PSA levels more likely associated with curable disease. To summarize, it is recommended that PSA testing begin at age 40 years and then be repeated annually. If the PSA is higher than the age-specific median level, repeat testing is warranted to verify this. Should there be concerns that BPH might be confounding the PSA determination, a PSA density (cut-off ⬎ 0.1) or a percent free PSA (cut-off 10%) could be used to either increase or decrease the level of suspicion. To rule out prostatitis, a course of antibiotic therapy and repeat testing might be warranted, but may not be appropriate in all situations. If PSA elevation is confirmed and does not decrease substantially with antibiotics or decrease substantially on follow-up, a prostate biopsy is recommended. Alternatively, PSA could be monitored at 3– 6 month intervals to assess PSA velocity with a PSA velocity of 0.35 ng/ml/y being a suspicious threshold at which biopsy should be strongly recommended. These biopsy recommendations are consistent with the 2009 NCCN Prostate Cancer Early Detection guidelines for men with PSA between 2.6 and 4 ng/ml or PSA velocity ⱖ 0.35 ng/ml/y. When PSA is ⬍2.5 ng/ml, biopsy should be considered when PSA velocity is ⱖ0.35 ng/ml/y. In the future, measurements of [-2] proPSA will be useful in improving the accuracy of PSA testing. Proenzyme forms of prostate specific antigen have been reported to be a more prostate cancer-specific marker than total PSA level [13–15]. In a recently published prospective prostate cancer screening study of over 2,000 men with total PSA between 2.5 and 10 ng/ml, the percent [-2]proPSA (ratio of [-2]proPSA to free PSA) had significantly better specificity for prostate cancer detection compared with total PSA and percent free PSA levels [14]. Another recent European study showed similar results with percent [-2]proPSA demonstrating significantly increased prostate cancer predictive value and specificity compared to both total PSA and percent free PSA levels [15]. When available for routine use, this marker’s ability to improve discrimination between prostate cancer and benign disease will further help in the appropriate selection of men for prostate biopsy.
Biopsy candidates should be selected using nomograms It is important to recognize first that PSA-based prostate cancer screening has significant limitations. While PSA is prostate-specific, it is not prostate cancer-specific and most men with elevated PSA do not have prostate cancer. It is also important to distinguish between biologically significant vs. clinically relevant disease. Biologically significant disease is cancer destined to cause significant morbidity and mortality. Clinically relevant disease causes anxiety or leads
to treatment with the potential for treatment-related morbidity. In the context of prostate cancer screening, it is important to note that a small proportion of men truly benefit from treatment and it is the identification of these men that should be the goal when selecting men for prostate biopsy. The PCPT provided a large set of data useful for developing nomograms to use in the selection of men for prostate biopsy [1]. In this trial 18,882 men were randomly assigned to receive either finasteride or placebo and were followed up with annual digital rectal examinations (DRE) and PSA measurements. Prostate biopsy was recommended “for cause” if the result of either test was abnormal. After 7 years of study, men who had not been diagnosed with prostate cancer were asked to undergo an end-of-study biopsy. These men were biopsied at PSA levels ⱕ 4 ng/ml with a normal DRE result. These end-of-study biopsies in men with PSA levels ⬍ 4 ng/ml showed that there was a continuum of risk for prostate cancer detection that correlated directly with PSA level in this range. The risk of finding high grade disease was also continuous with PSA level in that high grade disease was more likely at higher PSA levels. Utilizing the biopsy results of 5,519 men from the placebo group who underwent prostate biopsy, a predictive model of prostate cancer detection was developed [16]. Logistic regression was used to model the risk of prostate cancer and high grade disease associated with age at biopsy, race, family history of prostate cancer, PSA level, PSA velocity, DRE result, and previous prostate biopsy. Risk equations were created from the estimated logistic regression models. Variables that predicted prostate cancer included higher PSA level, positive family history of prostate cancer, and abnormal digital rectal examination result. Previous negative prostate biopsy was associated with reduced risk. Of note, neither age at biopsy nor PSA velocity contributed independent prognostic information. Higher PSA level, abnormal DRE result, older age at biopsy and African-American race were predictive for high grade disease, whereas previous negative prostate biopsy reduced the risk. Since finasteride use was found to improve the accuracy of PSA for detection of any cancer as well as high grade cancer, the PCPT risk calculator also accounts for the effect of finasteride on PSA performance [17]. The predictive model utilizing PCPT biopsy results allows an individualized assessment of prostate cancer risk and the risk of high grade disease for men who are candidates for prostate biopsy. The advantages of a risk calculator like this are that it captures all of the predictive values and continuous variables such as age and PSA, it can be validated on multiple cohorts, and it is flexible, allowing rapid evaluation and the incorporation of new markers. The risk calculator can be accessed online at http:// deb.uthscsa.edu/URORiskCalc/Pages/calcs.jsp. A potential limitation of the PCPT risk calculator is that it was developed from a cohort of men who entered the trial with a PSA of 3.0 ng/ml or less, a normal DRE, and age 55 years or older. Most men presenting for prostate biopsy
C.L. Amling et al. / Urologic Oncology: Seminars and Original Investigations 28 (2010) 542–545
have an increased PSA level or an abnormal DRE and many are younger than 55 years of age. As such, it is important to validate risk calculators to determine if it can be widely applied and useful in men seeking care in the routine American urology practice setting. To validate the PCPT risk calculator, Eyre et al. utilized data from patients undergoing prostate biopsy by 12 urologists at 5 sites who were enrolled in the Early Detection Research Network (EDRN) study [18]. The PCPT risk calculator was validated by examining the area under the receiver operating characteristic (ROC) curve, sensitivity, specificity, and calibration comparing observed vs. predicted risk of prostate cancer detection. This analysis revealed that the EDRN and PCPT cohorts were significantly different. Cancer incidence was greater in the EDRN validation cohort (43%) compared with the PCPT cohort (22%). Furthermore, the EDRN participants were younger and more racially diverse than PCPT participants. Prostate cancer severity was worse in the EDRN cohort as well. Despite these differences, however, the PCPT risk calculator was superior to PSA alone for predicting cancer in the EDRN cohort, validating this risk calculator’s performance in a multi-center urology practice setting. An additional advantage of a risk calculator over a cutpoint threshold for determining who to biopsy is the ability to incorporate new markers. An example of this is the recent incorporation of prostate cancer gene 3 (PCA3) into the PCPT risk calculator to improve its diagnostic accuracy [19]. Based on a cohort of 522 men who underwent prostate biopsy with measurements of urinary PCA3, serum PSA, DRE, and biopsy history, this marker was incorporated into the risk assessment. External validation of the updated risk calculator was performed on the cohort of 443 European patients and compared with PCPT risk estimates. Incorporation of PCA3 improved the diagnostic accuracy of the PCPT risk calculator demonstrating that new markers for prostate cancer can be incorporated relatively easily into existing risk calculators to improve the ability to select men for prostate biopsy. In the future, additional biologic markers such as EPCA2, gene fusions, and metabolomics may be similarly incorporated into existing risk calculators. In summary, selection of men using a PSA cut-off, PSA density or PSA velocity threshold should really be something of the past as this selects men for biopsy based on the risk of any cancer. Currently, we should use continuous risk models to select men for biopsy. These include nomograms and risk calculators in which new markers can be incorporated as they are detected to improve their diagnostic accuracy. Using these methods, the risk of any biologically significant cancer can be better determined. In the future, individual risk assessment will be possible utilizing PSA at a young age and genetic testing in men who are candidates for biopsy. The hope then will be that the risk of a truly biologically significant cancer will be determined and that selective biopsy can be recommended. This assessment of risk should also lead to development of prostate cancer prevention strategies.
545
References [1] Thompson IM, Pauler DK, Goodman PJ, et al. Prevalence of prostate cancer among men with a prostate-specific antigen level ⱕ 4.0 ng per milliliter. N Engl J Med 2004;350:2239 – 46. [2] Carter HB, Pearson JD. PSA velocity for the diagnosis of early prostate cancer. A new concept Urol Clin North Am 1993;20:665–70. [3] Carter HB, Ferrucci L, Kettermann A, et al. Detection of life-threatening prostate cancer with prostate-specific antigen velocity during a window of curability. J Nat Cancer Inst 2006;98:1521–7. [4] Loeb S, Roehl KA, Helfand BT, et al. Can prostate specific antigen velocity thresholds decrease insignificant prostate cancer detection? J Urol 2010;183:112–7. [5] Smith DS, Catalona WJ. The nature of prostate cancer detected through prostate specific antigen based screening. J Urol 1994;152: 1732– 6. [6] Punglia RS, D’Amico AV, Catalona WJ, et al. Effect of verification bias on screening for prostate cancer by measurement of prostatespecific antigen. N Engl J Med 2003;349:335– 42. [7] Krumholtz JS, Carvalhal GF, Ramos CG, et al. Prostate-specific antigen cutoff of 2.6 ng/ml for prostate cancer screening is associated with favorable pathologic tumor features. Urol 2002;60:469 –73. [8] Antenor JV, Roehl KA, Eggener SE, et al. Preoperative PSA and progression-free survival after radical prostatectomy for stage T1c disease. Urol 2005;66:156 – 60. [9] Lucia MS, Darke AK, Goodman PJ, et al. Pathologic characteristics of cancers detected in the Prostate Cancer Prevention Trial: Implications for prostate cancer detection and chemoprevention. Cancer Prev Res 2008;1:167–73. [10] Makarov DV, Humphreys EB, Mangold LA, et al. Pathological outcomes and biochemical progression in men with T1c prostate cancer undergoing radical prostatectomy with prostate specific antigen 2.6 to 4.0 vs. 4.1 to 6.0 ng/ml. J Urol 2006;176:554 – 8. [11] Loeb S, Roehl KA, Antenor JV, et al. Baseline prostate-specific antigen compared with median prostate-specific antigen for age group as predictor of prostate cancer risk in men younger than 60 years. Urology 2006;67:316 –20. [12] Carter HB, Kettermann A, Ferrucci L, et al. Prostate-specific antigen velocity risk count assessment: A new concept for detection of lifethreatening prostate cancer during window of curability. Urology 2007;70:685–90. [13] Catalona WJ, Bartsch G, Rittenhouse HG, et al. Serum pro-prostate specific antigen preferentially detects aggressive prostate cancers in men with 2 to 4 ng/ml prostate specific antigen. J Urol 2004;171: 2239 – 44. [14] Le BV, Griffin CR, Loeb S, et al. [-2]Proenzyme prostate specific antigen is more accurate than total and free prostate specific antigen in differentiating prostate cancer from benign disease in a prospective prostate cancer screening study. J Urol 2010;183:1355–9. [15] Jansen FH, van Schaik RH, Kurstjens J, et al. Prostate-specific antigen (PSA) isoform p2PSA in combination with total PSA and free PSA improves diagnostic accuracy in prostate cancer detection. Eur Urol 2010 Feb 13 [Epub ahead of print]. [16] Thompson IM, Ankerst DP, Chi C, et al. Assessing prostate cancer risk: Results from the Prostate Cancer Prevention Trial. J Nat Cancer Inst 2006;98:529 –34. [17] Thompson IM, Chi C, Ankerst DP, et al. Effect of finasteride on the sensitivity of PSA for detecting prostate cancer. J Nat Cancer Inst 2006;98:1128 –33. [18] Eyre SJ, Ankerst DP, Wei JT, et al. Validation in a multiple urology practice cohort of the Prostate Cancer Prevention Trial calculator for predicting prostate cancer detection. J Urol 2009;182:2653– 8. [19] Ankerst DP, Groskopf J, Day JR, et al. Predicting prostate cancer risk through incorporation of prostate cancer gene 3. J Urol 2008;180: 1303– 8.