European Urology

4. Discussion

There is now general agreement that the performance of PSA as a diagnostic test for PCa leaves much to be desired. It has been shown that PSA's performance can be improved by adding other risk factors such as prostate volume and the outcome of DRE [12] x M.J. Roobol, E.W. Steyerberg, R. Kranse, et al. A risk-based strategy improves prostate-specific antigen–driven detection of prostate cancer. Eur Urol. 2010;57:79-85 Abstract, Full-text, PDF, Crossref. . Similar improvements in performance have also been noted for the same predictors in connection with PCA3 scoring 9 x I.L. Deras, S.M. Aubin, A. Blase, et al. PCA3: a molecular urine assay for predicting prostate biopsy outcome. J Urol. 2008;179:1587-1592 Crossref. , and 13 x M.P. Van Gils, D. Hessels, O. van Hooij, 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;13:939-943 Crossref. . Still, a direct head-to-head comparison of PSA and PCA3 has not been available until now and is necessary to understand the relationship between the two markers and to better understand the performance of PCA3 in a setting where almost all men of a large cohort were biopsied. Table 3 shows that a commonality of both markers is that there is no cut-off value at which sensitivity and specificity achieve a reasonable balance. In the setting of this prescreened population, a PSA cut-off of ≥2.0 ng/ml and the recommended score of 35 for PCA3 provide the best balance for the two parameters. However, a sensitivity of 57.4% for PSA ≥2.0 ng/ml and a sensitivity of 68% for a PCA3 score of 35 means that 42.6% and 32%, respectively, of all biopsy-detectable cancers are missed.

How then should PCA3 be applied, and which aspects are important in decision taking? As our data show, there is always a trade-off between sensitivity and specificity. If we wish to apply either one of the two tests with a higher sensitivity, then we have to accept a lower specificity, translating into a larger proportion of all men biopsied and a larger proportion of potentially unnecessary biopsies. Alternatively, if we decide to work with lower sensitivities and accept missing larger proportions of cancer, then we need to know that those missed cancers are not potentially life threatening.

A PCA3 score ≥35, as generally recommended, shows a PPV of 23.9% and misses 39 of the 122 cancers, 5 of which are considered to have serious characteristics. Still, 51.7% of all biopsies are saved. The fact that a PSA cut-off of 3.0 ng/ml has been used two or more times in this population obviously creates a disadvantage for PSA; however, we are dealing with a prescreened population, and we wish to imitate a situation that is common in most countries where PSA screening is prevalent. A recent inventory of tumour characteristics and treatment in both arms of the ERSPC shows that PCa cases detected in recent years in the control arm resemble those detected with screening [14] x Boevee S, Venderbos LDF, Ranniko A, et al. Change of treatment over time in both arms of the ERSPC. Eur J Cancer. In press. .

Table 6 provides a review of six studies 2 x D. Hessels, J.M.T. Klein Gunnewiek, I. van Oort, et al. DD3^PCA3-based molecular urine analysis for the diagnosis of prostate cancer. Eur Urol. 2003;44:8-16 discussion 15–6 Abstract, Full-text, PDF, Crossref. , 5 x M.P. Van Gils, E.B. Cornel, D. Hessels, et al. Molecular PCA3 diagnostics on prostatic fluid. Prostate. 2007;67:881-887 Crossref. , 7 x L.S. Marks, Y. Fradet, I.L. Deras, et al. PCA3 molecular urine assay for prostate cancer in men undergoing repeat biopsy. Urology. 2007;69:532-535 Crossref. , 9 x I.L. Deras, S.M. Aubin, A. Blase, et al. PCA3: a molecular urine assay for predicting prostate biopsy outcome. J Urol. 2008;179:1587-1592 Crossref. , 13 x M.P. Van Gils, D. Hessels, O. van Hooij, 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;13:939-943 Crossref. , and 15 x A. Haese, A. de la Taille, H. van Poppel, et al. Clinical utility of the PCA3 urine assay in European men scheduled for repeat biopsy. Eur Urol. 2008;54:1081-1088 Abstract, Full-text, PDF, Crossref. from the literature, of which two 2 x D. Hessels, J.M.T. Klein Gunnewiek, I. van Oort, et al. DD3^PCA3-based molecular urine analysis for the diagnosis of prostate cancer. Eur Urol. 2003;44:8-16 discussion 15–6 Abstract, Full-text, PDF, Crossref. , and 5 x M.P. Van Gils, E.B. Cornel, D. Hessels, et al. Molecular PCA3 diagnostics on prostatic fluid. Prostate. 2007;67:881-887 Crossref. report complete data on performance characteristics of PCA3. Sensitivity and specificity show a remarkable similarity in all studies, including the present one. The AUCs indicate an overall performance of a test and, contrary to sensitivity and specificity, do not have direct clinical relevance.

In considering these studies, the reader must realise that all except our study use variable PSA cut-off values with or without DRE as the primary diagnostic tool. This approach results in attribution or assignment bias because, contrary to our data, cancers that can be found in large proportions in the PSA ranges below the ones utilised in each one of these studies are, by definition, excluded. The present study, in which 95.6% of all men were biopsied (70% were biopsied for the first time) on the basis of PCA3 score, allows more conclusive evaluation of test performance.

Table 5 addresses the value of PCA3 in the PSA range ≤3.0 ng/ml, showing that 66.2% of these men have a PCA3 score ≥35, necessitating biopsy in about half of the population. The PPV of 21.6% compares very favourably with the PPV of DRE in men with PSA values of <4.0 ng/ml, reported to be 10% [16] x W.J. Catalona, D.S. Smith, D.K. Ornstein. 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. 1997;277:1452-1455 Crossref. .

The results obtained by authors who have studied the relationship between PCA3 scores and parameters of cancer aggressiveness differ by outcome. Although some studies 9 x I.L. Deras, S.M. Aubin, A. Blase, et al. PCA3: a molecular urine assay for predicting prostate biopsy outcome. J Urol. 2008;179:1587-1592 Crossref. , 17 x D. Hessels, M.P. van Gils, O. van Hooij, et al. Predicitve value of PCA3 in urinary sediments in determining clinico-pathological characteristics of prostate cancer. Prostate. 2010;70:10-16 Crossref. , and 18 x D. Schilling, T. de Reijke, B. Tombal, A. de la Taille, J. Hennenlotter, A. Stenzl. The prostate cancer gene 3 assay: indications for use in clinical practice. BJU Int. 2010;105:452-455 Crossref. have not established a positive relationship between PCA3 scores and parameters of more serious disease (T2 or higher or Gleason ≥6), other authors show such relationships 13 x M.P. Van Gils, D. Hessels, O. van Hooij, 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;13:939-943 Crossref. , 19 x E.J. Whitman, J. Groskopf, A. Ali, et al. PCA3 score before radical prostatectomy predicts extracapsular extension and tumor volume. J Urol. 2008;180:1975-1978 discussion 1978–9 , and 20 x M.P. Van Gils, D. Hessels, C.A. Hulsbergen-van de Kaa, et al. Detailed analysis of histopathological parameters in radical prostatectomy specimens and PCA3 urine test results. Prostate. 2008;68:1215-1222 Crossref. . Several authors including Nakanishi et al [21] x H. Nakanishi, J. Groskopf, H.A. Fritsche, et al. PCA3 molecular urine assay correlates with prostate cancer tumor volume: implication in selecting candidates for active surveillance. J Urol. 2008;179:1804-1809 discussion 1809–10 found a significant value for PCA3 in predicting small prostates with volumes <0.5 cm³; however, Deras et al [9] x I.L. Deras, S.M. Aubin, A. Blase, et al. PCA3: a molecular urine assay for predicting prostate biopsy outcome. J Urol. 2008;179:1587-1592 Crossref. could not confirm this finding. The reported and discussed data suggest that PCA3 may benefit from combination with other predictive factors. The limited literature on this issue is developing. Chun et al [22] x F.K. Chun, A. de la Taille, H. van Poppel, et al. Prostate cancer gene 3 (PCA3): development and internal validation of a novel biopsy nomogram. Eur Urol. 2009;56:659-668 Abstract, Full-text, PDF, Crossref. used a base model of clinical predictors, and PCA3 scores were used to construct a nomogram recommended for use in conjunction with PCA3 testing, leading to significant improvements of the AUC up to a maximum of 0.730. Ankerst et al [23] x D.P. Ankerst, J. Groskopf, J.R. Day, et al. Predicting prostate cancer risk through incorporation of prostate cancer gene 3. J Urol. 2008;180:1303-1308 discussion 1308 Crossref. added the PCA3 test to the National Cancer Institute risk calculator, which is based on the Prostate Cancer Prevention Trial [8] x I.M. Thompson, D.P. Ankerst, C. Chi, et al. Operating characteristics of prostate-specific antigen in men with an initial PSA level of 3.0 ng/ml or lower. JAMA. 2005;294:66-70 Crossref. , leading to a nonsignificant increase of the performance of this risk calculator. Adding the PCA3 test results to the ERSPC risk calculator showed an improvement of the AUC for predicting a positive biopsy of 0.69–0.71. A similar increase of the AUC was seen when adding a previously described kallikrein panel 24 x Roobol MJ, Hessels D, Vickers A, et al. Predicting prostate biopsy outcome in a screening setting. PSA, PCA3, a kallikrein panel, the risk calculator or a combination? ERSPC Rotterdam. Poster 974 presented at: European Association of Urology Annual Congress; April 16–20, 2010; Barcelona, Spain. , and 25 x A.J. Vickers, A.M. Cronin, M.J. Roobol, et al. A four-kallikrein panel predicts prostate cancer in men with recent screening: data from the European Randomized Study of Screening for Prostate Cancer, Rotterdam. Clin Cancer Res. 2010;16:3232-3239 Crossref. .

In the available data sets, it is remarkable that the highest usually reported PCA3 score of ≥100 has relatively low positive biopsy rates, such as the 54% reported by Deras et al [9] x I.L. Deras, S.M. Aubin, A. Blase, et al. PCA3: a molecular urine assay for predicting prostate biopsy outcome. J Urol. 2008;179:1587-1592 Crossref. and the 47% reported by Van Gils et al [13] x M.P. Van Gils, D. Hessels, O. van Hooij, 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;13:939-943 Crossref. . This finding is confirmed in an exaggerated fashion in our data set. As Table 4 shows, of 90 men with a PCA3 score ≥100, only 28 (31%) had PCa. The available information in the literature does not offer a conclusive explanation for this phenomenon. One explanation is that PCa is missed by biopsy. Reports on the number of biopsies taken are incomplete in the cited literature. To exclude this possibility, a follow-up study of our present protocol has been carried out in which repeat biopsy is offered to those men who had a PCA3 score ≥100 and a negative biopsy (M.J. Roobol, unpublished data, 2010).

Strengths and weaknesses of our study are that we included a large cohort of men, recruitment and testing took place under well-controlled circumstances, all tests were carried out in one laboratory, and the prescreened nature of our study population reflects the current clinical situation in most countries. Our study allows the establishment of performance characteristics of PCA3 because 95.6% of all men who were biopsied were also indicated by the PCA3 cut-off ≥10 and, for the first time, allows evaluation of PCA3 in men with low PSA values (<3.0 ng/ml). The fact that we are addressing a prescreened population could also be considered a weakness because data cannot be used to enhance our understanding of the performance characteristics of PCA3 in previously unscreened men. Due to prior screening, the number of first biopsies above the PSA cut-off of 3.0 ng/ml is low, and a sextant biopsy technique was applied.