Evaluation of urinary prostate cancer antigen-3 (PCA3) and TMPRSS2-ERG

Transcription

1 Evaluation of urinary prostate cancer antigen-3 (PCA3) and TMPRSS2-ERG score changes when starting androgen-deprivation therapy with triptorelin 6-month formulation in patients with locally advanced and metastatic prostate cancer Luis Martínez-Piñeiro, Jack A. Schalken*, Patrick Cabri, Pascal Maisonobe, Alexandre de la Taille, on behalf of the Triptocare Study Group Urology Unit, Infanta Sofía University Hospital, Madrid, Spain, *Department of Urology, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands, Ipsen Pharma, Paris, and Department of Urology, CHU Mondor, Créteil, France Objective To assess prostate cancer antigen-3 (PCA3) and TMPRSS2-ERG scores in patients with advanced and metastatic prostate cancer at baseline and after 6 months of treatment with triptorelin 22.5 mg, and analyse these scores in patient-groups defined by different disease characteristics. Patients and Methods The Triptocare study was a prospective, open-label, multicentre, single-arm, Phase III study of triptorelin 22.5 mg in men with locally advanced or metastatic prostate cancer, who were naïve to androgen-deprivation therapy (ADT). The primary objective was to model the urinary PCA3 change at 6 months, according to baseline variables. Other outcome measures included urinary PCA3 and TMPRSS2-ERG scores and statuses, and serum testosterone and prostate-specific antigen (PSA) levels at baseline and at 1, 3 and 6 months after initiation of ADT. Safety was assessed by recording adverse events and changes in laboratory parameters. Results The intent-to-treat population comprised 322 patients; 39 (12.1%) had non-assessable PCA3 scores at baseline, and 109/322 (33.9%), 215/313 (68.7%) and 232/298 (77.9%) had non-assessable PCA3 scores at 1, 3 and 6 months, respectively. Baseline Gleason score was the only variable associated with non-assessability of PCA3 score at 6 months (P = 0.017) the hazard of having a non-assessable PCA3 score at 6 months was fold higher (95% confidence interval ) in patients with a Gleason score 8 vs those with a Gleason score 6. The median PCA3 scores at baseline were significantly higher in patients aged 65 years vs those aged <65 years and in patients with a serum PSA level <100 ng/ml vs those with serum PSA level of >200 ng/ml. The median PCA3 score was significantly lower in patients with metastasis than in patients with no metastasis or unknown metastasis status. TMPRSS2-ERG scores 35 were considered positive (n = 149 [51.6%]). Age, presence of metastasis, PSA level and Gleason score at baseline were not associated with a significant difference in the proportion of TMPRSS2-ERG-positive scores. The median serum PSA levels decreased from 45.5 ng/ml at baseline to 1.2 ng/ml after 6 months, and as expected, >90% of patients achieved castrate levels of testosterone (<50 ng/dl) at 1, 3, and 6 months during triptorelin treatment. The safety profile reported from this study is consistent with the known safety profile of triptorelin. Conclusion These data from the Triptocare study suggest that urinary PCA3 or TMPRSS2-ERG score are not reliable markers of cancer stage in advanced prostate cancer. Urinary PCA3 and TMPRSS2-ERG scores do not appear to be useful in assessing response to ADT in advanced prostate cancer, with most patients having non-assessable scores after 6 months of treatment. Keywords advanced prostate cancer, disease severity, GnRH agonist, PCA3, TMPRSS2-ERG BJU Int 2014; 114: wileyonlinelibrary.com BJU International 2013 BJU International doi: /bju Published by John Wiley & Sons Ltd.

2 PCA3 and TMPRSS2-ERG for assessing response to ADT in advanced prostate cancer Introduction PSA is not the ideal marker for determining the seriousness of prostate cancer [1,2], and therefore, new biomarkers such as the prostate cancer antigen-3 (PCA3) and gene fusions of the type 2 transmembrane serine protease (TMPRSS2) with v-erythroblastosis virus E26 oncogene homolog (ERG; TMPRSS2-ERG gene fusion) have been assessed. These newer urine biomarkers may have a role in determining the presence and aggressiveness of prostate cancer, but more evidence is needed. PCA3 mrna is over-expressed in malignant prostate tissue compared with benign prostate tissue and other normal tissue [3]. Assays are available to accurately measure PCA3 mrna and PSA mrna [4], and the PCA3 score derived from these measures has good sensitivity and specificity for predicting positive repeat prostate biopsy [5]. The use of these assays can reduce the number of unnecessary biopsies [6]. In patients with clinically localised prostate cancer, PCA3 score may also show correlation with tumour volume and may differentiate between low-volume/low-grade cancer and significant cancer [7 10], and is an independent risk factor for prostate cancer [7,11,12]. However, results have been conflicting and some authors found no correlation between tumour volume and PCA3 score [13 15]. Recently, it was suggested that while a low PCA3 score was a valuable predictor of pathologically confirmed, low-volume disease and insignificant prostate cancer, an elevated PCA3 score may be less useful for identifying advanced or aggressive prostate cancer [16]. The potential for PCA3 score as a prognostic marker in patients with metastatic prostate cancer is unknown. TMPRSS2-ERG is the most common gene fusion in prostate cancer occurring in 50% of all prostate cancer [17], and can be associated with aggressive or recurrent prostate cancer [17,18]. Therefore, the TMPRSS2-ERG score is also under investigation as a prognostic indicator in prostate cancer [17]. UsingacombinationofTMPRSS2-ERG and PCA3 scores may significantly improve the sensitivity of prostate cancer diagnosis [19 22]. In addition, the Canary/Early Detection Research Network (EDRN) Prostate Active Surveillance Study showed that TMPRSS2-ERG and PCA3 may help assess the risk of aggressive cancer in men undergoing active surveillance [23]. While a pilot study has assessed the impact of dutasteride on PCA3 score in men with BPH [24], there have been no studies assessing the impact of androgen-deprivation therapy (ADT) for prostate cancer on either PCA3 score or TMPRSS2-ERG score. The Triptocare study (ClinicalTrials.gov identifier: NCT ) was designed to assess PCA3 and TMPRSS2- ERG scores and their changes during the initiation of ADT with triptorelin 22.5 mg in endocrine therapy naïve patients with advanced prostate cancer (including patients with metastatic disease). This paper presents the baseline and follow-up data from the Triptocare study. Patients and Methods This was a prospective, open-label, multicentre, single-arm, Phase III study of triptorelin 22.5 mg in patients with advanced prostate cancer (ClinicalTrials.gov Identifier: NCT ). The study was conducted in compliance with the Declaration of Helsinki, Good Clinical Practice guidelines, all relevant Institutional Review Boards/independent Ethics Committees, and local regulatory requirements. All patients gave written informed consent before study entry. Adult men (aged 18 years) with histologically confirmed, locally advanced or metastatic prostate cancer who were naïve to ADT, but were considered candidates for hormonal therapy, were eligible for inclusion in the study. Other inclusion criteria included a performance status score of 2 accordingto WHO criteria and an estimated survival time of 12 months. Exclusion criteria were: vertebral metastases with a risk of spinal cord compression or urinary tract obstruction; surgical castration; previous or planned curative prostate cancer therapy (radiation therapy or surgery); previous hormone therapy; the use of an experimental treatment 30 days of study entry; diagnosis of any other cancer that is not stable or in remission 5 years of study entry; known hypersensitivity to triptorelin, GnRH, GnRH agonists or to the excipient; or inability to comply with the study instructions. Patients were recruited from 48 centres in Europe (Belgium, Denmark, France, Latvia, Lithuania, the Netherlands, Romania, Spain, and the UK) between November 2009 and April Interventions and Assessments After screening, all eligible patients received one i.m. dose of triptorelin 22.5 mg (6-month formulation). Treatment was discontinued if castrate levels of testosterone (<50 ng/dl) were not achieved at 1 month and 3 months (confirmed by a second measurement 2 weeks after the 3-month visit) after receiving the injection of triptorelin. Patients were assessed at baseline (information collected 7 14 days before administration of triptorelin), and 1, 3 and 6 months after receiving triptorelin. Physical examination including DRE and post-dre urine PSA, PCA3 and TMPRSS2-ERG mrna measurements were conducted at baseline, and at 1, 3 and 6 months to calculate the PCA3 and TMPRSS2-ERG scores. Samples were assessed for biomarker levels at a central laboratory. PCA3, TMPRSS2-ERG, and PSA mrnas were quantified in each specimen using the PROGENSA PCA3 assay or the second generation developmental TMPRSS2-ERG assay as described previously [4,21]. PCA3 scores were calculated as 1000 (PCA3 mrna BJU International 2013 BJU International 609

3 Martínez-Piñeiro et al. copies/psa mrna copies). TMPRSS2-ERG scores were calculated as (TMPRSS2-ERG mrna copies/psa mrna copies). Samples with a mean PSA mrna level of 7500 copies/ml were considered informative and the PCA3 and TMPRSS2-ERG scores were defined as assessable in such samples. Non-assessable PCA3 and TMPRSS2-ERG scores were those in which the urinary PSA mrna level was <7500 copies/ml. In addition, the proportion of samples with adequate PSA mrna ( 7500 copies/ml) but PCA3 mrna or TMPRSS2-ERG mrna below or equal to the limit of quantification was reported. A TMPRSS2-ERG score of 35 was defined as a positive score. This was based upon this threshold having the greatest utility, with an optimal balance between sensitivity (43%) and specificity (73%) for predicting significant cancer or high-grade cancer (data provided by Gen-Probe, San Diego, USA). Furthermore, this gene fusion is known to occur in 50% of men with prostate cancer [17], and this threshold was close to the median TMPRSS2-ERG score of 38 in this study population. Other assessments at baseline and at 1, 3 and 6 months were serum testosterone, serum PSA, and alkaline phosphatase measurements, and documentation of concomitant treatments and adverse events (AEs; graded according to version 3.0 of the National Cancer Institute Common Toxicity Criteria [NCI-CTC] [25]). Other baseline assessments included: patient characteristics (e.g. age, body mass index [BMI], body weight and vital signs); medical history and previous therapy; prostate cancer evaluation (by Gleason score and TNM staging [classification following standard local practice]; after completion of the study a retrospective centralised review of Gleason scores was conducted); WHO performance status; haematology assessment; and TRUS at the physicians discretion (249 patients [76.4%] had received TRUS in the 3 months before study entry). Vital signs, body weight and haematology were also measured at 6 months. Statistical Methods The primary analysis was conducted on the modified intent-to-treat (mitt) population, which consisted of all patients in the ITT population with an assessable PCA3 score at baseline. Secondary analyses were conducted on the ITT population, which consisted of all treated patients who had at least one baseline and one post baseline assessment of the primary biomarker (PCA3 score), or on the mitt population. For the primary analysis, the change in PCA3 at 6 months was modelled based on its non-assessability status using a survival (time-to-event) model and the following baseline variables: PCA3 score, age, PSA level, clinical stage, metastasis status, Gleason score, testosterone level, time since diagnosis, BMI, alkaline phosphatase level, race, region and TMPRSS2-ERG score status (i.e. positive or negative). The first step of this analysis was to perform univariate Cox analysis of the non-assessability of the PCA3 score at 6 months with each of these baseline variables. The variables that met the significance threshold of 0.20 were then entered into the Cox multivariate model. Secondary analyses included: modelling the change in PCA3 at 3 months using a similar model to the primary analysis, PCA3 and TMPRSS2-ERG scores and status at baseline, and months 1, 3 and 6; proportion of castrated patients (serum testosterone <50 ng/dl) at months 1, 3 and 6; PSA level at baseline and at months 1, 3 and 6; and safety profile assessed by AEs. The relationship between PCA3/TMPRSS2-ERG scores and measures of disease severity including Gleason score, PSA level and metastatic status at baseline were also assessed. In addition, correlations between baseline PCA3 scores and clinicopathological variables were assessed using Spearman s rank correlation (non-parametric), as well as univariate and multivariate regression analyses based on rank data. All secondary variables were analysed using descriptive summary statistics (n, mean, standard deviation (SD), median, 95% CI, minimum, maximum) for continuous variables, and frequency counts (and proportions and 95% CI) for binary/categorical variables. The sample size calculation was driven by the original primary efficacy objective (to model the PCA3 score change at 6 months, using a multivariate linear model and the following outcome variables at baseline: age, PSA, clinical stage (TNM), pathological stage (Gleason score) and TMPRSS2-ERG score). In all, 315 patients were considered sufficient to allow detection of an effect size of 0.4 given a maximal imbalance of between the two levels of any tested factor (assuming a power of 80% and a two-sided α level of 5%). Results In the Triptocare study, 339 patients were screened and 326 patients received treatment. The safety population comprised 325 patients (one patient was discontinued after 1 month with no post-baseline safety assessment as he was admitted to an asylum), and the ITT population comprised 322 patients (one patient was excluded as he had no recorded baseline PCA3 score, and two patients had no recorded post-baseline PCA3 score) (Fig. 1). Patient and disease characteristics at baseline are summarised in Table 1. The median time since histological diagnosis in the ITT population was 29 days. Of the 322 patients in the ITT population, 205 patients agreed to participate in the retrospective centralised review of Gleason scores, which resulted in confirmation of the Gleason score in 44.4% of patients, but 28.3% had their Gleason score upgraded and 27.3% had their Gleason score down-graded. 610 BJU International 2013 BJU International

4 PCA3 and TMPRSS2-ERG for assessing response to ADT in advanced prostate cancer Fig. 1 Patient progress through the trial. Assessed for eligibility (n=339) Enrolment Screening failures (n=13) Adverse event Consent withdrawn Does not meet entry criteria Patients receiving treatment (n=326) Discontinued (n=1) Admitted to asylum 1 Follow-up Safety population (n=325) Excluded (n=3) No baseline PCA3 score No post-baseline PCA3 score 1 2 ITT population (n=322) Analysis Retrospective analysis (n=205) mitt population (n=283) Excluded (n=39) PCA3 score non-assessable at baseline 1 Discontinued (n=27) Adverse event Lost to follow up Consent withdrawn Lack of efficacy Disease progression Other Completed study with no major deviation per-protocol population (n=256) PCA3 and TMPRSS2-ERG Scores At baseline, 283 patients had assessable PCA3 scores (mitt population) and 39 (12.1%) had non-assessable PCA3 scores due to PSA mrna levels below the level of detection. The proportion of patients with non-assessable PCA3 score increased with triptorelin treatment (Table 2). In the mitt population, the median PCA3 score was 48.0 (95% CI, ; n = 283) at baseline and 32.0 (95% CI, ; n = 61) 6 months after the single triptorelin injection. The region, PCA3 score, age, PSA level, metastasis status, Gleason score and testosterone level met the 0.20 threshold in the univariate analysis of the primary endpoint (non-assessability of PCA3 score at 6 months). The subsequent Cox multivariate analysis of these variables showed that Gleason score was the only variable associated with non-assessability of PCA3 score at 6 months (P = 0.017) the hazard of having a nonassessable PCA3 score at 6 months was times higher (95% CI, ) in patients with a Gleason score 8 vs those with a Gleason score 6. Patients with a baseline Gleason score 8 had a significantly higher hazard of having a non-assessable PCA3 score at 3 months (hazard ratio 1.804, 95% CI, , P = 0.030) vs those with a baseline Gleason 6. None of the other variables had a significant association with non-assessability of PCA3 score at 3 months. The median baseline PCA3 scores were significantly higher in patients aged 65 years than in those aged <65 years, significantly higher in patients with a serum PSA level of <100 ng/ml than in those with a serum PSA level of >200 ng/ml, and significantly lower in patients with metastasis than in patients with no metastasis or unknown metastasis (Table 3). Assessment of non-parametric correlations indicated weak but statistically significant correlations between baseline PCA3 score and age (ρ 0.18, P < 0.01) and serum PSA level (ρ 0.19, P < 0.01), but no correlation with Gleason score (ρ 0.11, P = 0.058). Multivariate regression analysis showed significant effects of age (P < 0.01) and metastasis status (P < 0.01), but not of Gleason score and PSA level. BJU International 2013 BJU International 611

5 Martínez-Piñeiro et al. Table 1 Patient and disease characteristics of patients with prostate cancer included in the Triptocare study (ITT population and mitt population). Characteristic Population, n (%) Total (ITT) (n = 322) Assessable PCA3 scores (mitt) (n = 283) Non-assessable PCA3 scores (n = 39) Age, years, n (%) (81.1) 228 (80.6) 33 (84.6) <65 61 (18.9) 55 (19.4) 6 (15.4) Median (95% CI) BMI, kg/m 2 * 26.1 ( ) 25.9 ( ) 27.4 ( ) Median (95% CI) testosterone level, ng/dl 329 ( ) 331 ( ) 318 ( ) N (%): PSA level, ng/ml : < (65.8) 184 (65.0) 28 (71.8) (12.4) 37 (13.1) 3 (7.7) > (21.4) 61 (21.6) 8 (20.5) Gleason score : 6 54 (16.8) 53 (18.7) 1 (2.6) 7 97 (30.1) 82 (29.0) 15 (38.5) (52.5) 146 (51.6) 23 (59.0) Metastases (any T, and N): M1 114 (35.4) 93 (32.9) 21 (53.8) M0 144 (44.7) 135 (47.7) 9 (23.1) Mx 64 (19.9) 55 (19.4) 9 (23.1) Primary tumour: T1 12 (3.7) 12 (4.2) 0 (0.0) T2 33 (10.2) 31 (11.0) 2 (5.1) T3 236 (73.3) 207 (73.1) 29 (74.4) T4 41 (12.7) 33 (11.7) 8 (20.5) *n = 311 in total population and 272 in population with assessable PCA3 scores; n = 321 in total population and 282 in population with assessable PCA3 scores; n = 320 in total population and 281 in population with assessable PCA3 scores. Table 2 Proportion of patients with non-assessable PCA3 scores or with PCA3 mrna below the limits of quantification during the Triptocare study (ITT population). PCA3 score assessability Baseline 1 Month 3 Months 6 Months n/n (%) n/n (%) n/n (%) n/n (%) Not assessable (PSA mrna <7500 copies/ml) 39/322 (12.1) 109/322 (33.9) 215/313 (68.7) 232/298 (77.9) PCA3 mrna BLQ* 15/322 (4.7) 40/322 (12.4) 31/313 (9.9) 27/298 (9.1) *But PSA mrna 7500 copies/ml; BLQ, below the limits of quantification. After 6 months, PCA3 scores remained significantly lower in patients with metastasis than in patients with no metastasis (only 17 and 33 patients, respectively, with assessable PCA3 scores), but did not differ significantly between age groups or PSA level groups. Of the 322 patients in the ITT population, 289 had assessable TMPRSS2-ERG scores at baseline; 149 (51.6%) had positive scores ( 35). The proportion of patients with non-assessable TMPRSS2-ERG score increased with triptorelin treatment (Table 4). At baseline, TMPRSS2-ERG scores were not significantly associated with any variable (Table 3). Efficacy Variables Baseline PSA and alkaline phosphatase levels varied according to different disease characteristics in a manner that would be expected. For example, the median alkaline phosphatase levels were significantly higher in patients with metastasis than in patients with no metastasis. The median PSA and testosterone levels decreased from baseline at all time-points (Table 5). As expected, >90% of patients achieved castrate levels of testosterone at 1, 3 and 6 months during triptorelin treatment. Safety In all, 11 serious AEs (SAEs) resulted in withdrawal from the study; all 11 were death that was not related to the study medication. AEs were considered possibly related to study medication in 107 (32.9%) patients (162 events). Five of these AEs were grade 3 (severe); one pain in extremity, one intentional overdose of paracetamol due to depression, and three hot flushes. The only treatment-related AEs that 612 BJU International 2013 BJU International

6 PCA3 and TMPRSS2-ERG for assessing response to ADT in advanced prostate cancer Table 3 Baseline PCA3 scores and TMPRSS2-ERG scores in patients with prostate cancer in the Triptocare study stratified by baseline age, PSA level, Gleason score and metastasis status. Variable PCA3 score* median TMPRSS2-ERG score % (95% CI) (95% CI), P <35 (n = 140) 35 (n = 149) Age, years ( ) 82.9 ( ) 77.9 ( ) < ( ) P < ( ) 22.1 ( ) Serum PSA level, ng/ml < ( ) P < ( ) 67.1 ( ) ( ) 14.3 ( ) 11.4 ( ) > ( ) P < ( ) 20.8 ( ) Gleason score ( ) 15.0 ( ) 22.1 ( ) ( ) 28.6 ( ) 29.5 ( ) ( ) 56.4 ( ) 47.0 ( ) Metastasis M ( ) P < ( ) 29.5 ( ) M ( ) 42.9 ( ) 51.0 ( ) Mx 65.0 ( ) P < ( ) 19.5 ( ) *In those with assessable scores, n = 283; indicates significant difference between marked groups (i.e. 95% CI do not overlap); P value for age <65 vs 65 years using Wilcoxon Mann Whitney test, based on patients with assessable baseline PCA3 score; P value for PSA level <100 vs >200 ng/ml using Wilcoxon Mann Whitney test, based on patients with assessable baseline PCA3 score, P value was non-significant for the comparison between PSA level <100 vs ng/ml; P value for PSA level vs >200 ng/ml using Wilcoxon Mann Whitney test, based on patients with assessable baseline PCA3 score; P value for M1 vs M0 using Wilcoxon Mann Whitney test, based on patients with assessable baseline PCA3 score, P value was non-significant for the comparison between M0 vs Mx; P value for M1 vs Mx using Wilcoxon Mann Whitney test, based on patients with assessable baseline PCA3 score. P value was non-significant for the comparisons between the Gleason score categories. Table 4 Proportion of patients with non-assessable TMPRSS2-ERG scores or with TMPRSS2-ERG mrna below the limits of quantification during the Triptocare study (ITT population). TMPRSS2-ERG score assessability Baseline 1 Month 3 Months 6 Months n/n (%) n/n (%) n/n (%) n/n (%) Not assessable (PSA mrna <7500 copies/ml) 33/322 (10.2) 117/322 (36.3) 213/313 (68.1) 241/298 (80.9) Positive score ( 35) 149/322 (46.3) 106/322 (32.9) 53/313 (16.9) 27/298 (9.1) Negative score (<35) 140/322 (43.5) 97/322 (30.1) 45/313 (14.4) 24/298 (8.1) Table 5 Serum testosterone level and serum PSA level in patients with prostate cancer in the Triptocare study. Baseline and follow-up data for the ITT population. Variable Baseline (n = 321) 1 Month (n = 320) 3 Months (n = 311) 6 Months (n = 296) Median (range) PSA level, ng/ml 45.5 ( ) 8.3 (0 581) 1.8 (0 969) 1.2 (0 1251) Median (95% CI) testosterone level, ng/dl 329 ( ) 15 (13 17) 10 (10 12) 12 (10 12) occurred in >3% of the study population were hot flushes (25.8% of patients). Discussion The Triptocare study showed that endocrine therapy naïve men with metastatic prostate cancer had significantly lower urinary PCA3 scores than men with no metastases. Furthermore, PCA3 scores were lower in younger patients and in patients with higher PSA levels. There was no obvious relationship between disease stage/severity and a positive urinary TMPRSS2-ERG score. Likewise, there was no obvious relationship in the kinetics of both these biomarkers with response to therapy with the GnRH agonist triptorelin. The only variable that showed a significant association with non-assessability of PCA3 score 6 months after starting triptorelin therapy was the baseline Gleason score. Interpretation of the finding with PCA3 scores at baseline is difficult. The median scores in the group with metastases (31.0) are similar to PCA3 scores reported for cancer-positive patients in biopsy cohorts: 33.7 [5], 35.1 [6], 38.0 [26]. The BJU International 2013 BJU International 613

7 Martínez-Piñeiro et al. reason for the elevated median PCA3 scores in the group with no metastasis in the present study may reflect the higher tumour volume in patients included in this study (e.g. median PSA level in the group without metastasis was 25 ng/ml in the present study and 7 ng/ml in the Haese et al. [5] study). However, PCA3 scores were also lower in patients with higher baseline serum PSA values (>200 ng/ml). A possible explanation for this finding is that the expression of PCA3 mrna may be reduced in higher grade tumours, resulting in lower PCA3 scores. An assessment of PCA3 expression in low- and high-grade prostate cancers among patients undergoing radical prostatectomy for histologically confirmed prostate cancer suggested lower expression of PCA3 in the higher grade cancers, but the difference was not statistically significant [27]. Furthermore, Partin et al. [28] showed that men with more advanced disease have higher grade, higher volume tumours that produce less PSA per gram of tumour, suggesting that PSA and PCA3 expression in higher grade cancers may be unpredictable and variable. The original objective of the present study was to model the PCA3 score change after a single administration of triptorelin. However, baseline data at interim analysis showed that there was a relatively high frequency of non-assessable PCA3 and TMPRSS2-ERG scores (>10%), and this led to the decision of modelling the non-assessability status of the PCA3 score instead of the score itself at 6 months as the primary endpoint. If the collection of samples is performed adequately, previous studies suggest that a non-assessable PCA3 score would occur in <2% of patients [4,5,29]. However, the frequency is similar to that in another published study [30], and measurement of PCA3 scores in this patient population is a new area of research and the frequency of non-assessable PCA3 scores is not well known. In addition, ADT clearly had an impact on the proportion of patients with non-assessable PCA3 score, to the point where only a small proportion of this population had assessable PCA3 scores after 6 months. A high Gleason score at baseline was the only factor found to significantly influence the likelihood of PCA3 score non-assessability at 6 months. As discussed above, PSA mrna expression in urine after DRE may be low (even if serum PSA is high) in high-grade tumours, and it is notable that only one of 39 (2.6%) patients with non-assessable PCA3 scores at baseline had a Gleason score 6. One limiting factor in the interpretation of the PCA3 scores cannot be ignored. The reference measures that are being used to assess disease characteristics (PSA level, etc.) are themselves inaccurate. Clearly, more studies are needed to assess the possible role of PCA3 score as a marker in advanced prostate cancer. For example, measurement of PCA3 and TMPRSS2-ERG in blood (rather than urine) may provide a better correlation with the presence of metastases. Both RNAs have been shown to be detectable in blood from men with advanced (but not localised) prostate cancer [31]. A role for TMPRSS2-ERG score in prognosis assessment has been suggested [21], but its significance in prognosis of advanced disease is not well understood. The Triptocare baseline data did not show a relationship between higher grade cancer or metastatic disease and positive TMPRSS2-ERG scores (using 35 as the threshold score). A threshold for positive TMPRSS2-ERG scores is not clearly defined in the literature, and the data that exists stems from populations with less advanced disease. Defining a TMPRSS2-ERG score was beyond the scope of the present analysis, but using a threshold score of 35 confirmed that 50% of patients in this population had the TMPRSS2-ERG gene fusion, and an exploratory analysis of the baseline data that used a higher threshold score of 100 yielded the same results as those reported here. The relationships between serum PSA levels and alkaline phosphatase levels with more aggressive or metastatic disease were as expected [32,33]. The reduction in PSA level and proportion of patients achieving castration with triptorelin 22.5 mg at 1, 3 and 6 months were also largely consistent with previous studies [34]. Similarly, the safety profile reported in the present study is consistent with that previously seen in clinical studies of triptorelin [34] and as summarised in the product labelling. The present study is the first to analyse urine biomarkers PCA3 and TMPRSS2-ERG in patients with advanced and metastatic prostate cancer, and to assess their changes during the initiation of ADT. These data from the Triptocare study suggest that urinary PCA3 or TMPRSS2-ERG score are not reliable markers of cancer stage in this patient population, and that these biomarkers measured in urine are not useful in assessing response to ADT in advanced prostate cancer. Urinary PCA3 appears to be a useful marker for localised prostate cancer detection and tumour volume prediction, but is not useful in advanced disease. Measurement of PCA3 and TMPRSS2-ERG in blood (rather than urine) may provide a better correlation with the presence of metastases, but is something that is still under investigation. Acknowledgments The Triptocare study was funded by Ipsen. The authors take full responsibility for the content of the paper but thank Martin Gilmour of ESP Bioscience, Crowthorne, UK (supported by Ipsen) for editorial assistance in the preparation of the manuscript. The authors thank all investigators in the Triptocare study: Jean Amiel (France), Olivier Bouchot (France), Greg Boustead (UK), Owen Cole (UK), Ioan Coman (Romania), Jean-Louis Davin (France), Jean de la Rosette (The Netherlands), M. T. De Goeij (The Netherlands), Jesper Eldrup (Denmark), Marc Fillet (Belgium), Ulla Geertsen (Denmark), Vincent Gnanapragasam (UK), Samuils Gordins (Latvia), Lucien Hoekx (Belgium), Alain Houlgatte (France), Ioan Ioiart (Romania), Kieran Jefferson (UK), Mindaugas Jievaltas 614 BJU International 2013 BJU International

8 PCA3 and TMPRSS2-ERG for assessing response to ADT in advanced prostate cancer (Lithuania), Giedrius Jocys (Lithuania), Palle Jörn Osther (Denmark), Masood A. Khan (UK), Howard Kynaston (UK), Thierry Lebret (France), John McKnight (UK), Kári J. Minkines (Denmark), Jean-Luc Moreau (France), Juan Morote (Spain), Joop Noordzig (The Netherlands), Cristina Oprean (Romania), Edgar Paez (UK), Gilles Pasticier (France), Raj Persad (UK), Dominique Prapotnich (France), Xavier Rebillard (France), Maria J. Ribal (Spain), Alfredo Rodriguez Antolin (Spain), Dominique Rossi (France), Alain Ruffion (France), Eduardo Solsona (Spain), Algimantas Sruogis (Lithuania), Bertrand Tombal (Belgium), Hendrik Van Poppel (Belgium), Francisco G. Veiga (Spain), Arnauld Villers (France), Egils Vjaters (Latvia), and Simon Williams (UK). Conflict of Interest Luis Martínez-Piñeiro and Alexandre de la Taille are principal investigators of the Triptocare study, which was funded by Ipsen. Luis Martínez-Piñeiro has been a consultant for Ipsen. Jack A. Schalken has been a consultant for Ipsen and Gen-Probe. Patrick Cabri and Pascal Maisonobe are paid employees of Ipsen. References 1 Denham JW, Steigler A, Wilcox C et al. Why are pretreatment prostate-specific antigen levels and biochemical recurrence poor predictors of prostate cancer survival? Cancer 2009; 115: Sartor O. Prostate specific antigen (PSA) may be a poor marker for progression of prostate cancer after treatment with a combination of chemo-hormonal therapies. JLaStateMedSoc2008; 160: Hessels D, Klein Gunnewiek JM, van Oort I et al. DD3(PCA3)-based molecular urine analysis for the diagnosis of prostate cancer. Eur Urol 2003; 44: 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: 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; 54: Crawford ED, Rove KO, Trabulsi EJ et al. Diagnostic performance of PCA3 to detect prostate cancer in men with increased prostate specific antigen: a prospective study of 1,962 cases. J Urol 2012; 188: Nakanishi H, Groskopf J, Fritsche HA et al. PCA3 molecular urine assay correlates with prostate cancer tumor volume: implication in selecting candidates for active surveillance. J Urol 2008; 179: Ploussard G, Durand X, Xylinas E et al. Prostate cancer antigen 3 score accurately predicts tumour volume and might help in selecting prostate cancer patients for active surveillance. Eur Urol 2011; 59: Whitman EJ, Groskopf J, Ali A et al. PCA3 score before radical prostatectomy predicts extracapsular extension and tumor volume. J Urol 2008; 180: van Poppel H, Haese A, Graefen M et al. The relationship between Prostate CAncer gene 3 (PCA3) and prostate cancer significance. BJU Int 2012; 109: Chun FK, de la Taille A, van Poppel H et al. Prostate cancer gene 3 (PCA3): development and internal validation of a novel biopsy nomogram. Eur Urol 2009; 56: Nilsson J, Skog J, Nordstrand A et al. Prostate cancer-derived urine exosomes: a novel approach to biomarkers for prostate cancer. Br J Cancer 2009; 100: van Gils MP, Hessels D, Hulsbergen-van de Kaa CA et al. Detailed analysis of histopathological parameters in radical prostatectomy specimens and PCA3 urine test results. Prostate 2008; 68: Hessels D, van Gils MP, van Hooij O et al. Predictive value of PCA3 in urinary sediments in determining clinico-pathological characteristics of prostate cancer. Prostate 2010; 70: Augustin H, Mayrhofer K, Pummer K, Mannweiler S. Relationship between prostate cancer gene 3 (PCA3) and characteristics of tumor aggressiveness. Prostate 2012; 73: Auprich M, Chun FK, Ward JF et al. Critical assessment of preoperative urinary prostate cancer antigen 3 on the accuracy of prostate cancer staging. Eur Urol 2011; 59: Tomlins SA, Bjartell A, Chinnaiyan AM et al. ETS gene fusions in prostate cancer: from discovery to daily clinical practice. Eur Urol 2009; 56: Rostad K, Hellwinkel OJ, Haukaas SA et al. TMPRSS2:ERG fusion transcripts in urine from prostate cancer patients correlate with a less favorable prognosis. APMIS 2009; 117: Hessels D, Smit FP, Verhaegh GW, Witjes JA, Cornel EB, Schalken JA. Detection of TMPRSS2-ERG fusion transcripts and prostate cancer antigen 3 in urinary sediments may improve diagnosis of prostate cancer. Clin Cancer Res 2007; 13: Salagierski M, Schalken JA. Molecular diagnosis of prostate cancer: PCA3 and TMPRSS2:ERG gene fusion. J Urol 2012; 187: Tomlins SA, Aubin SM, Siddiqui J et al. Urine TMPRSS2:ERG fusion transcript stratifies prostate cancer risk in men with elevated serum PSA. Sci Transl Med 2011; 3: 94ra72 22 Cornu JN, Cancel-Tassin G, Egrot C, Gaffory C, Haab F, Cussenot O. Urine TMPRSS2:ERG fusion transcript integrated with PCA3 score, genotyping, and biological features are correlated to the results of prostatic biopsies in men at risk of prostate cancer. Prostate 2013; 73: Lin DW, Newcomb LF, Brown EC et al. Urinary TMPRSS2:ERG and PCA3 in an active surveillance cohort: results from a baseline analysis in the Canary Prostate Active Surveillance Study. Clin Cancer Res 2013; 19: van Gils MP, Hessels D, Peelen WP, Vergunst H, Mulders PF, Schalken JA. Preliminary evaluation of the effect of dutasteride on PCA3 in post-dre urine sediments: a randomized, open-label, parallel-group pilot study. Prostate 2009; 69: National Cancer Institute. Common Terminology Criteria for Adverse Events v3.0 (CTCAE) Available at: protocoldevelopment/electronic_applications/docs/ctcaev3.pdf. Accessed November Deras IL, Aubin SM, Blase A et al. PCA3: a molecular urine assay for predicting prostate biopsy outcome. J Urol 2008; 179: Robert G, Jannink S, Smit F et al. Rational basis for the combination of PCA3 and TMPRSS2:ERG gene fusion for prostate cancer diagnosis. Prostate 2012; 73: Partin AW, Carter HB, Chan DW et al. Prostate specific antigen in the staging of localized prostate cancer: influence of tumor differentiation, tumor volume and benign hyperplasia. J Urol 1990; 143: Marks LS, Fradet Y, Deras IL et al. PCA3 molecular urine assay for prostate cancer in men undergoing repeat biopsy. Urology 2007; 69: Liss MA, Santos R, Osann K, Lau A, Ahlering TE, Ornstein DK. PCA3 molecular urine assay for prostate cancer: association with pathologic features and impact of collection protocols. World J Urol 2011; 29: Jost M, Day JR, Slaughter R et al. Molecular assays for the detection of prostate tumor derived nucleic acids in peripheral blood. Mole Can. 2010; 9: 174 BJU International 2013 BJU International 615

9 Martínez-Piñeiro et al. 32 Wolff JM, Ittel T, Borchers H, Brauers A, Jakse G. Efficacy of skeletal alkaline phosphatase and prostate-specific antigen in the diagnosis of bone metastasis in cancer of the prostate. Urol Int 1998; 61: Sonpavde G, Pond GR, Berry WR et al. Serum alkaline phosphatase changes predict survival independent of PSA changes in men with castration-resistant prostate cancer and bone metastasis receiving chemotherapy. Urol Oncol 2012; 30: Lundstrom EA, Rencken RK, van Wyk JH et al. Triptorelin 6-month formulation in the management of patients with locally advanced and metastatic prostate cancer: an open-label, non-comparative, multicentre, phase III study. Clin Drug Investig 2009; 29: Correspondence: Luis Martínez-Piñeiro, Servicio de Urología, Infanta Sofía University Hospital, Paseo de Europa 34, San Sebastián de los Reyes, Madrid, Spain. [email protected] Abbreviations: ADT, androgen-deprivation therapy; (S)AE, (serious) adverse events; BMI, body mass index; ERG, v-erythroblastosis virus E26 oncogene homolog; (m)itt, (modified) intent-to-treat; PCA3, prostate cancer antigen-3; TMPRSS2, type 2 transmembrane serine protease. 616 BJU International 2013 BJU International