Eur Urol Suppl 2006;5(2):239 865 PSA VELOCITY AND THE DETECTION OF CLINICALLY SIGNIFICANT CANCERS IN AN EARLY DETECTION SETTING (ERSPC, ROTTERDAM) Roobol M.J., Schröder F.H. Erasmus Medical Centre, Urology, Rotterdam, The Netherlands INTRODUCTION & OBJECTIVES: In contrast to men with prostatic disease or BPH only, patients with clinically detected prostate cancer (PC) exhibit an early linear phase followed by an exponential phase of PSA increase prior to diagnosis. The PSA increase in absolute terms per year is called PSA velocity (PSAV). In an early detection setting PSAV did not differentiate between the presence and absence of PC (Roobol et al.2004,2005). Here we assess the value of PSAV in differentiating less aggressive PC from PC with more unfavourable prognostic characteristics. MATERIAL & METHODS: We used data from 558 men screened for the third time within the ERSPC (section Rotterdam). The screening interval was 4-yrs. All men were biopsied because of a PSA >= 3.0 ng/ml. Men who were diagnosed with PC were subdivided into less aggressive (LA, defined as clinical stage < T2B and /or Gleason score < 7) and more aggressive (MA, clinical stage >= T2B and/or Gleason score >= 7). The third group consisted of men with a benign biopsy result. We compared PSAV values of the different groups and assessed number of cancers missed when using different PSAV cut-offs (assessed in the 2 nd screening interval) as an indication for biopsy. Using different PSAV cut-offs the number of biopsies done and PC cases detected were calculated. RESULTS: In 558 men biopsied 94 cancers were detected, 73 LA and 21 MA cases. Mean PSAV values in the first screening interval (8-4 yrs before biopsy) were res. 0.24, 0.19 and 0.23 ng/ml/yr for benign, LA PC and MA PC cases (p=0.487). Mean PSAV values in the second interval (4-0 yrs before biopsy) were res. 0.39,0.46 and 0.96 ng/ml/yr (p< 0.001). One MA case was left out of the analysis (PSAV of 26.4 ng/ml/yr). Using next to the PSA 3.0 ng/ml cut-off a PSAV cut-off of 0.20 ng/ml/yr would have resulted in 317 benign biopsies (32% less), 53 LA PC cases (28% less) and 21 MA PC cases (0% less). These numbers were res. –58%, -55% and –19% for a PSAV cut-off >= 0.40 ng/ml/yr and –82%, -83% and –38% for a PSAV cut-off >= 0.75 ng/ml/yr. CONCLUSIONS: A screening algorithm with the use of PSA and PSAV, results in a decrease in the number of biopsy sessions. The number of PC detected also decreases but this decrease is solely in the category PC defined as less aggressive when using a PSAV cut-off >= 0.20 ng/ ml/yr. A mean PSAV of 0.96 ng/ml/yr was significantly related to more aggressive PC cases. 866 CORRELATION OF PSA-VELOCITY WITH TUMOUR-STAGE AND GLEASON SCORE IN A SCREENING POPULATION Berger A.P. 1 , Deibl M. 2 , Klocker H. 1 , Steiner H. 1 , Bektic J. 1 , Pelzer A.E. 1 , Leonhartsberger N. 1 , Bartsch G. 1 , Horninger W. 1 1 Medical University Innsbruck, Dept. of Urology, Innsbruck, Austria, 2 Medical University Innsbruck, Dept. of Statistics, Innsbruck, Austria INTRODUCTION & OBJECTIVES: To correlate longitudinal PSA changes with tumour- stage (1997 TNM classification) and Gleason score in a screening population. MATERIAL & METHODS: Serial PSA measurements performed over 10 years were evaluated in 353 men who eventually developed prostate cancer. Patients were stratified into subgroups according to Gleason scores (biopsy; n = 353) and pathologic stage (available in those 269 patients undergoing radical retropubic prostatectomy). RESULTS: In men with cancer mean tPSA increased from 2.28 ng/ml at 10 years before diagnosis to 6.37 ng/ml at the time of positive biopsy (PSA-velocity: 0.409 ng/ml/year). PSA velocity for patients who presented with Gleason scores of up to 6 was significantly different (p < 0.05) compared to those with Gleason 7 and those with Gleason 8 or more (0.503, 0.667 and 1.21 ng/ml per year in the six years before diagnosis, respectively). However, no difference was observed when comparing patients with Gleason 4 to patients with Gleason 5 and 6 (p = 0.574 and 0.767, respectively). Also, no difference in PSA-velocity was found between patients with Gleason 7 compared to patients with Gleason 8 (p = 0.874). In patients with organ-confined disease the annual increase in PSA in the six years before diagnosis was significantly lower than in patients with pT3 or pT4 disease (0.465 and 0.718 ng/ml per year). No difference was found when comparing patients with pT2a and pT2b (p = 0.738), but significant difference was found when comparing pT2 with pT3 tumours. CONCLUSIONS: PSA-velocity is significantly associated with Gleason scores and pathologic stage. In a screening-population a cut-off value of 0.4 ng/ml/year might be helpful in selecting patients with potentially curable prostate cancer. Gleason Score PSA-velocity (mean) ng/ml/yr pathologic stage PSA-velocity (mean) ng/ml/yr 4 0.58 pT2a 0.44 5 0.48 pT2b 0.51 6 0.51 7 0.67 pT3a 0.68 8 0.69 pT3b 0.75 9 1.21 10 3.60 pT4 1.05 867 PROSTATE-SPECIFIC ANTIGEN VELOCITY AND PROSTATE-SPECIFIC ANTIGEN SLOPE COMPARED FOR PROSTATE CANCER DIAGNOSIS IN MEN WITH 3 OR MORE PSA MEASUREMENTS Benecchi L., Potenzoni M. Fidenza Hospital, Urology, Parma, Italy INTRODUCTION & OBJECTIVES: The rate of rise in PSA (PSA velocity) is greater for men with prostate cancer compared to those without the disease. The aim of this study is to compare different tools for the evaluation of a longitudinal series of PSA in the same patients, such as PSA velocity and PSA slope. MATERIAL & METHODS: This study was conducted on 312 male patients evaluated with transrectal ultrasound-guided biopsy of prostate with 6 or more cores. Patients with at least 3 consecutive PSA measurements (done in our centralised laboratory) in at least 18 months entered the study. PSA slope was estimated by the slope of the least squares regression line fit to PSA versus time in years and PSA velocity was calculated as the running average of the rate of change during at least 3 consecutive assays. RESULTS: Median age was 66 years (range 45 to 86). Overall 67 patients were affected by primary prostate cancer, 245 were controls without prostate cancer. PSA slope and PSA velocity were significantly higher in patients with prostate cancer than in controls. At the ROC analysis PSA slope evidenced better results than PSA velocity (AUC 0.743 for PSA slope; AUC 0.663 for PSA velocity; p = 0.037). At PSA slope (calculated with the least square fit) equal to zero, the sensitivity resulted as being 94% with a specificity of 38.8%. CONCLUSIONS: PSA slope calculated with 3 or more PSA assays permits longitudinal evaluation of PSA for prostate diagnosis. PSA slope improves both sensitivity and specificity in prostate cancer diagnosis compared with PSA velocity. 868 IS A PSA VELOCITY CUT-OFF OF 0.75 NG/ML/YEAR APPROPRIATE FOR ALL MEN? Connolly D. 1 , Black A. 2 , Murray L.J. 2 , Gavin A. 2 , Keane P.F. 1 1 Belfast City Hospital, Urology, Belfast, United Kingdom, 2 Queen’s University Belfast, Northern Ireland Cancer Registry, Belfast, United Kingdom INTRODUCTION & OBJECTIVES: PSA velocity (PSAV) has been used to differentiate prostate cancer from benign disease with a level of greater than 0.75ng/ml/year being used to recommend biopsy. There is however controversy as to the ideal cut-off. We investigated the association of PSAV with initial PSA to assess if a lower PSAV cut-off may be more appropriate in men with an initially low PSA. MATERIAL & METHODS: The Northern Ireland Cancer Registry (NICR) maintains an electronic register of all PSA tests performed in Northern Ireland, which is linked to the NICR database of incident cancers occurring within the region. Hospital discharge and histopathology data are used to identify men with benign histology (needle biopsy and TURP). Men with cancer or benign histology and an initial PSA less than 10ng/ml were identified. Those with at least three PSA tests before diagnosis, which were carried out over a minimum of 18 months were included. PSAV was calculated as the rate of change of PSA using linear regression analysis. RESULTS: 2247 men were included, with 1497 (66.6%) having benign histology and 750 (33.4%) cancer. of these, 32 were excluded as their PSA values suggested having received treatment for prostate cancer. The mean (10.34 vs. 0.94ng/ml/yr) and median (1.47 vs. 0.23ng/ml/yr) PSAV were significantly different (p<0.001) in the cancer and benign groups respectively. Median PSAV was significantly different in each initial PSA category. PSAV increased incrementally with initial PSA (see Table). Cancer Biopsy 0.75ng/ml/year cut-off Initial PSA (ng/ml) No. of men Median PSAV* No. of men Median PSAV* Sensitivity Specificity 0-1.99 88 0.37 393 0.05 0.45 0.9 2-3.99 175 1.21 406 0.28 0.67 0.75 4-5.99 164 1.47 340 0.37 0.70 0.64 6-7.99 166 1.61 227 0.49 0.77 0.64 8-9.99 125 2.31 131 0.54 0.77 0.55 All 718 1.47 1497 0.23 0.69 0.73 *Kruskal-Wallis test (p<0.001) CONCLUSIONS: PSAV is strongly associated with initial PSA. There are significant differences in PSAV in benign and malignant cases in each category of initial PSA. The single cut-off of 0.75ng/ml/year may not be appropriate for all men, particularly those with an initially high (>8ng/ml) or low (<2ng/ml) PSA.