Prostate news article, March 2009
Summary
Prostate biopsy decisions are traditionally guided by digital rectal examination and measurement of serum total prostate specific antigen (tPSA). However, both techniques are subject to inherent weaknesses. Prostate cancer gene 3 (PCA3), a gene-based marker, specific for prostate cancer, supplements the predictive power of tPSA to improve diagnosis of disease. Inclusion of this new marker in the standard of care for men at risk for prostate cancer should be considered as it presents marked potential for better prostate biopsy decision making and for improving overall patient care.
Certainty - freedom from doubt; a total security from error.
What if clinicians were able to make diagnostic and treatment decisions in a state of complete certainty? Patients would no longer face the risks of unnecessary invasive techniques, with their own inherent risks, or of adverse events related to avoidable treatment. In diagnosis of prostate cancer (pCA), the urologist would make prostate biopsy decisions based on non-invasive or minimally-invasive indicators indicating the presence of the disease. No patient would undergo biopsy who did not have prostate cancer. No patient would risk the anxiety and co-morbidities of biopsy without good cause. Further, what if those non-invasive or minimally-invasive indicators quantified tumour aggressiveness? The urologist could then determine the urgency of biopsy and treatment in light of patient desires and life expectancy based on other health status indicators.
Unfortunately, today’s urologists and their patients enjoy no such security in diagnosis and management of the most prevalent disease that they manage.
On the other hand, biopsy decisions have not been uninformed. In addition to clinical variables, demographics, and the presence or absence of other risk factors, digital rectal examination (DRE) of the prostate and measurement of serum total prostate-specific antigen (tPSA) have traditionally been used to assist in biopsy decision-making. Use of these techniques in combination has improved the detection and treatment of pCA over the last few decades. But, the improvement has been limited due to intrinsic weaknesses in both methods. DRE is subjective and displays marginal predictive value[1-3], while PSA is subject to various inherent flaws primarily driven by non-specificity for pCA. Thus, poor survival in prostate cancer results from a current lack of specific, highly predictive methods for early detection and for differentiation of aggressive and indolent cancers.
Recently, a non-invasive urinary test for prostate cancer gene 3 (PCA3) has been developed. PCA3 is an emerging gene-based marker that is highly specific for pCA. This review examines the current diagnostic dilemmas, the weaknesses of traditional testing, and the potential of PCA3 to complement existing diagnostic methodologies.
The dilemma
Let us illustrate the current clinical dilemma with an example: a fifty-five year-old Caucasian man sees his urologist for an initial appointment following referral from primary care. He reports a family history of pCA in second-degree relatives, he complains of frequent urination, and his latest annual physical indicated a serum prostate specific antigen (tPSA) concentration of 2.7 ng/ml. The urologist performs a digital rectal examination (DRE), which reveals no suspicious nodule, and subsequently confirms the tPSA value. Should the urologist recommend prostate biopsy at this time? Despite more than two decades of research on PSA, clinical experience, and establishment of institutional procedures, urologists still cannot make this decision with great confidence.
This example is neither unrealistic nor uncommon[4]. While patients present with more obvious risk for prostate cancer, still others present with more apparently benign indications who are subsequently found to harbour pCA (the prostate cancer rate is ~20% in the range of 2.5-4 ng/ml tPSA, inclusive of indolent and clinically relevant cancers[5]. Even taking a more compelling example wherein the hypothetical patient presents with tPSA between 4 and 10 ng/ml, there is still no certainty that prostate cancer will be detected upon biopsy. In fact, up to 60% of men with tPSA in this range will have a negative biopsy.
PSA: contemporary use and inherent flaws
Beginning at approximately age 50[6], men with a life expectancy of at least ten years undergo annual measurement of serum tPSA. If tPSA results are elevated in comparison to previous results but no other symptoms indicate a risk for pCA, DRE or tPSA testing may be repeated at appropriate intervals to observe and confirm any trends. If tPSA continues to increase or subsequent DRE results are suspicious, the clinician may attempt to rule out various benign conditions using imaging techniques, cystoscopy, and determination of percent free PSA (%fPSA). If the results of these analyses indicate a sufficient risk for pCA, biopsy will be recommended. Of course, the definition of “sufficient risk” will depend on the physician, their interpretation of the data, and established institutional procedures.
Characterizing risk based solely on serum tPSA findings presents inherent difficulties. PSA is specific for prostate tissue but not for prostate cancer. Elevated values of serum tPSA are observed in multiple benign conditions involving enlargement of the prostate [7-11] including prostatic hypertrophy (BPH)[7], and acute prostatitis[8]. Conversely, high body mass index (BMI) erroneously lowers tPSA values as a result of hemodilution[12]. Thus, the interpretation of tPSA values is prone to error arising from non-specific sources.
Furthermore, serum tPSA values have been demonstrated to be poor indicators of pCA aggressiveness regardless of the cutpoint chosen[13]. Because PSA does not correlate well with aggressiveness, there is a trend in clinical practice toward overdiagnosis and consequent overtreatment of prostate cancer[14].
In light of these inherent weaknesses of serum tPSA testing, it is clear that complementary indicators are needed to better inform the decisions to biopsy or perform radical treatments. Preferably, these new indicators would be insensitive to the non-specific factors that affect serum tPSA results. Furthermore, emerging indicators would ideally correlate with tumour aggressiveness and provide information independent of and complementary to serum tPSA.
Identification of PCA3 and early studies
PCA3, also referred to as PCA3DD3 or DD3PCA3 in the literature, was first identified in 1999 via differential display analysis in healthy, hyperplastic, and cancerous tissue from patients undergoing radical prostatectomy[15]. PCA3 exhibited a high level of expression in prostate tumours and was apparently absent in benign tissue. Further analyses demonstrated low but quantifiable expression in benign prostate tissue but undetectable levels of expression in normal tissues from all major organs. Additionally, no expression could be detected in tumours from breast, cervix, endometrium, ovary, or testis, and cancer cell lines from bladder, breast, kidney, and ovaries.
Subsequent studies using quantitative research tests for PCA3 demonstrated a median 66-fold upregulation in prostate malignancies and a high sensitivity and specificity for detection of pCA[16]. The PROGENSA™ PCA3 test was developed soon thereafter. The test employs Transcription-Mediated Amplification (TMA™) technology to quantify PCA3 and PSA messenger RNA (mRNA) in urine samples collected following DRE. The DRE is required to release prostate cells into the urine and the quantification of PSA mRNA is required to normalize for the total mRNA present in a sample (PSA mRNA levels in prostate cells released into urine are completely unrelated to PSA protein levels in blood and are essentially unchanged in pCA[17]). Thus, the method measures both PCA3 mRNA and PSA mRNA and the results are represented as a ratio of the two mRNAs referred to as the “PCA3 Score.” Similar to other gene-based tests, the PCA3 Assay is comparable in cost and complexity. Samples must be sent to an accredited laboratory experienced in performing molecular testing and PCA3 Scores are reported back to the urologist.
Sample collection and specimen stability are robust. Informative rates (percentage of urine samples yielding accurately quantifiable mRNAs for assay) have been demonstrated to exceed 99%[18] and the assays demonstrate good reproducibility with intra- and inter-assay coefficients of variation of <13% and <12%, respectively and total variation of <20% for the PCA3 Score[19].
Clinical Application of the PROGENSA™ PCA3 Assay
Utility of PCA3 testing
Determination of a PCA3 Score may be useful in several clinical scenarios. First, the Score may be used to increase confidence in an initial biopsy decision where serum tPSA results are uncertain (2.5-10 ng/ml). Second, PCA3 testing may be used to increase confidence in a re-biopsy decision wherein DRE and serum tPSA results are suspicious and/or family history and other factors indicate increased risk for pCA. Lastly, when biopsy results are positive but tumour aggressiveness is unknown, PCA3 may be useful in comparing risks and benefits of radical prostatectomy versus active surveillance management. Thus, the availability of a PCA3 Score alone or in combination with existing methods may better guide biopsy decision-making than current methods and may also be useful as an indicator of clinical stage and disease significance.
PCA3 clinical performance in comparison to serum tPSA
Comparative research studies have consistently demonstrated better prostate cancer predictive value of PCA3 versus serum tPSA. Marks and colleagues[20] studied 233 men with a prior negative prostate biopsy but with evidence of serum tPSA persistently above 2.5 ng/ml. Applying receiver-operating characteristic (ROC) curve analysis to PCA3 and serum tPSA results obtained after a re-biopsy of these men yielded significantly higher area under the curve (AUC) for PCA3 versus serum tPSA (Table 1).
In a separate study, Groskopf, et al.[21] compared PCA3 and tPSA in 70 men who received prostate biopsy based on pre-existing risk factors in comparison with 52 apparently healthy men with no known risk factors. At a PCA3 Score cutpoint of 50, sensitivity was 69% and specificity was 79%. For serum tPSA at the established cutpoint of 2.5 ng/ml and with sensitivity held constant at 69%, specificity for tPSA was 60%.
The foregoing results have been recently confirmed using a time-resolved fluorescence-based variant of the PCA3 test by van Gils, et al.[22] In their multicenter study of 583 men with serum tPSA between 3 and 15 ng/ml, the AUC for prediction of positive biopsy was higher for PCA3 than for serum tPSA testing (Table 1). Also, a correlation of increasing PCA3 Score with increasing probability of positive repeat biopsy was demonstrated [22].
Associations with prostate volume and tumour volume/tumour aggressiveness
An association of a marker with prostate volume regardless of the presence or absence of prostate cancer is an undesirable characteristic as it indicates non-specificity of the marker. Whereas serum tPSA shows such associations, PCA3 has not in an initial study by Deras, et al.[23] In this study, the associations of both markers with prostate volume were evaluated in a cohort of 570 men scheduled for initial or repeat prostate biopsy. Serum tPSA values increased demonstrably and significantly (p<0.0001) with increasing prostate volume whereas PCA3 scores were unaffected (p=0.54).
Conversely, associations of a marker with tumour aggressiveness are obviously desirable. Nakanishi and colleagues[24] assessed PCA3 correlation with tumour volume and tumour aggressiveness via analysis of pre-prostatectomy urine collections in 96 men with biopsy-confirmed pCA. PCA3 Score increased linearly and significantly with increasing tumour volume (R=0.27, p=0.008), and a significant difference (p=0.007) was demonstrated when comparing PCA3 scores for individuals with low volume and low grade tumours to PCA3 Scores in the subpopulation with “significant cancers” (individuals with combinations of high dominant tumour volumes and Gleason scores).
Most recently, a study by Haese et al.[18] has supported the findings of Nakanishi, Deras and Marks. In this multicenter, multinational European study, 463 men with one or more previous negative biopsies were re-biopsied following DRE and urine collections. The re-biopsy yielded 128 cancers (28%). Detected cancers were classified as indolent if they were stage T1c, had PSA density <0.15 ng/ml, Gleason score at biopsy ≤6, and had ≤33% positive cores (a standardized minimum of 10 peripheral zone cores were acquired). Median PCA3 Scores were higher in significant cancer versus indolent (42.1 v. 21.4, p=0.0059). Additionally, PCA3 Scores were higher for patients with biopsy Gleason score ≥7 v. <7 (p=0.0401) and for patients with stage T2 v. T1c cancer (p=0.005). Furthermore, the PCA3 Score was not affected by age and number of previous negative biopsies. PCA3 Score was also insensitive to total prostate volume whereas serum tPSA increased with volume. Lastly, PCA3 Score corresponded with increasing probability of a positive repeat biopsy.
PCA3 performance versus serum total PSA range
It is desirable for any new indicator of pCA to maintain its predictive value across the entire range of population serum tPSA values. To address the performance of PCA3 in this regard, Deras, et al.[23] categorised their study cohort into patients with serum tPSA <4 ng/ml, 4-10 ng/ml, and >10 ng/ml and determined sensitivity and specificity of PCA3 for detection of positive biopsy within each category. Across all categories, sensitivity was 54% at a specificity of 74%. These parameters varied <10% across serum tPSA range categories. The Haese study[18] confirmed these findings. PCA3 Scores did not vary significantly (p=0.7282) across the same categories of serum tPSA concentration. Across all categories, sensitivity was 47% at a specificity of 72%, results which are in excellent agreement with Deras[23].
Complementarity to serum total PSA
If PSA and PCA3 are destined to be complementary in clinical management and treatment decisions, there must be evidence that the utility of both markers is enhanced when data are analyzed as covarying predictors of biopsy outcomes. In the Deras study[23], urinary PCA3 and serum tPSA were evaluated in univariate and multivariate logistic regression models. Predictive probability relative to biopsy outcome was determined. For tPSA alone, PCA3 alone, and the combination of tPSA + PCA3, the areas under the curve from ROC were 0.547, 0.686, and 0.752, respectively (Figure 1). The increase in AUC observed in the multivariate model was strongly significant (p=0.0002), demonstrating the complementarity of the methods.
Furthermore, the incorporation of PCA3 information into the Prostate Cancer Prevention Trial (PCPT) risk calculator demonstrated additional benefit[25]. Although it was not possible to measure directly urinary PCA3 Scores in the PCPT study population, statistical methods to incorporate data from external populations were developed [25]. PCA3, serum tPSA, and DRE data from a cohort of 521 men undergoing prostate biopsy were included in the original PCPT risk calculator, serum tPSA by itself, urinary PCA3 by itself, and an updated PCPT risk calculator incorporating PCA3. AUC for the PCPT calculator incorporating PCA3 was 0.703, which was statistically superior to the PCPT calculator without PCA3 (AUC=0.618, p<0.025). Both the updated PCPT calculator and PCA3 by itself were also superior to serum tPSA alone. In addition, sensitivity, positive predictive value, and negative predictive value were all maximized for the PCPT risks with PCA3 incorporated.
The recent Haese study[18] again supports these earlier findings. In a multivariate logistic regression model for prediction of pCA at repeat biopsy, PCA3 score was an independent predictor (p=0.006) of outcome following adjustment for age, serum tPSA, %fPSA, DRE, and prostate volume. Inclusion of PCA3 in the base model containing the other terms improved accuracy of the model by 4.2%, which was significant at p<0.001. PCA3 Score was also the most informative univariate predictor in this study.
Avoidance of unnecessary biopsies
The effectiveness of PSA screening on prostate cancer mortality is still a matter for debate, however an increasing number of men undergo PSA testing annually. For example, data from the National Cancer Institute indicates that an additional 1.8 million U.S. men age 40-69 and 1.2 million men over 70 would have an abnormal PSA value if the threshold for PSA was decreased from 4.0 ng/ml to 2.5 ng/ml [26]. This results in an additional 3.0 million men with an elevated PSA that would be candidates for biopsy. Such trends demonstrate a need for other cancer specific markers to improve diagnostic and treatment decision-making.
The Haese study[18] provides some insights regarding identification of patients for whom biopsy is unwarranted. At a PCA3 Score cut-off of 20, repeat biopsies would have been reduced by 44% while missing only 9% of cancers. This finding suggests the potential of PCA3 to reduce the incidence of overdiagnosis and further studies are ongoing.
Conclusion
Over the past two decades, serum tPSA, together with other indicators, has guided biopsy decisions for urologists. While research during this period has characterized the strengths of tPSA testing, many weaknesses have also been revealed. The discovery and clinical evaluation of PCA3 has demonstrated that the marker supplements tPSA in diagnosis and is insensitive to the non-specific factors that can affect circulating tPSA levels. The addition of PCA3 to the urologist’s diagnostic toolset will not result in a state of certainty; however, diagnostic sensitivity, specificity, and predictive value are incrementally improved by its inclusion. In turn, biopsy and management decisions may be better informed. This has the potential to improve the overall level of patient care.
References
- Schroder FH, van der Maas P, Beemsterboer P, et al. Evaluation of the digital rectal examination as a screening test for prostate cancer. Rotterdam section of the European Randomized Study for Prostate Cancer. J Natl Cancer Inst 1998;90:1817-23.
- Issa MM, Zasada W, Ward K, et al. The value of digital rectal examination as a predictor of prostate cancer diagnosis among United States Veterans referred for prostate biopsy. Cancer Detect Prev 2006;30:269-75.
- Richie JP, Catalona WJ, Ahmann FR, et al. Effect of patient age on early detection of prostate cancer with serum prostate-specific antigen and digital rectal examination. Urology 1993;42:365-74.
- 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.
- Canby-Hagino E, Hernandez J, Brand TC, et al. Prostate cancer risk with positive family history, normal prostate examination findings, and PSA less than 4.0 ng/ml. Urology 2007;70:748-52.
- National Cancer Institute, NIH. Available at: https://www.cancer.gov/cancertopics/factsheet/Detection/PSA. Accessed July 25, 2008.
- Nadler RB, Humphrey PA, Smith DS, et al. Effect of inflammation and benign prostatic hyperplasia on elevated serum prostate specific antigen levels. J Urol 1995;154:407-13.
- Sindhwani P, Wilson CM. Prostatitis and serum prostate-specific antigen. Curr Urol Rep 2005;6:307-12.
- Bhanot S, Gopalakrishnan R, Oliver RT. Post-biopsy rise in serum PSA. A potential tool for the dynamic evaluation of prostate cancer/prostatic intraepithelial neoplasia (PIN). Cancer Biol Ther 2003;2:67-70.
- Kiran PR. Markedly raised serum prostate specific antigen levels. Prostatic infarction rather than malignancy? Aust Fam Physician 2001;30:458-60.
- Oremek GM, Seiffert UB. Physical activity releases prostate-specific antigen (PSA) form the prostate gland into blood and increases serum PSA concentrations. Clin Chem 1996;42:691-5.
- Banez LL, Hamilton RJ, Partin AW, et al. Obesity-related plasma hemodilution and PSA concentration among men with prostate cancer. JAMA 2007;298:2275-80.
- Fall K, Garmo H, Andren O, et al. Prostate specific antigen level as a predictor of lethal prostate cancer. J Natl Cancer Inst 2007;99:526-32.
- Zappa M, Ciatto S, Bonardi R, et al. Overdiagnosis of prostate carcinoma by screening: an estimate based on the results of the Florence Screening Pilot Study. Ann Oncol 1998;9:1297-300.
- BussemakersMJG, Van Bokhoven A, Verhaegh GW, et al. DD3: A new prostate-specific gene, highly over-expressed in prostate cancer. Cancer Res 1999;59:5975-9.
- Hessels D, Klein Gunnewiek MT, van Oort I, et al. DD3PCA3-based molecular urine analysis for the diagnosis of prostate cancer. Eur Urol 2003;44:8-16.
- Haese A, de la Taille A, van Poppel H, et al. Clinical utility of the PCA3 urine assay in European men scheduled for repeat biopsy. Eur Urol 2008 (in press).
- Sokoll L, Ellis W, Lange P, et al. A multicenter evaluation of the PCA3 molecular urine test: preanalytical effects, analytical performance, and diagnostic accuracy. Clin Chim Acta 2008;389:1-6.
- Marks LS, Fradet Y, Deras IL, et al. PCA3 molecular urine assay for prostate cancer in men undergoing repeat biopsy. Urology 2007;69:532-5.
- Groskopf J, Aubin SM, Deras IL, et al. APTIMA PCA3 molecular urine test: development of a method to aid in the diagnosis of prostate cancer. Clin Chem 2006;52:1089-95.
- van Gils MPMQ, Hessels D, van Hooij O, et al. The time-resolved fluorescence-based PCA3 test on urinary sediments after digital rectal examination; a Dutch multicenter validation of the diagnostic performance. Clin Cancer Res 2007;13:939-43.
- Deras IL, Aubin SMJ, Blase A, et al. PCA3: a molecular urine assay for predicting prostate biopsy outcome. J Urol 2008;179:1587-92.
- Nakanishi H, Groskopf J, Fritsche HA, et al. PCA3 molecular urine assay correlates with prostate cancer tumour volume: implication in selecting candidates for active surveillance. J Urol 2008;179:1804-10.
- Parekh DJ, Ankerst DP, Groskopf J, et al. Predicting prostate cancer risk through incorporation of a molecular biomarker, PCA3. American Urological Association annual meeting, 2008.
- Welch, GH, Schwartz, LM, Woloshin, S. Prostate-Specific Antigen Levels in the United States: Implications of Various Definitions for Abnormal. JNCI Journal of the National Cancer Institute 2005 97(15):1132-1137.
Table 1: Summary of serum tPSA and PCA3 comparative studies
Figure legend
Figure 1. ROC curve analysis of serum tPSA, PCA3 Score, and a multivariate logistic regression model for prediction of positive initial biopsy. The multivariate model included logarithmic transformations of PCA3 Score and tPSA in addition to prostate volume and DRE result. (Reproduced from Deras, et al.[24])
