[Comparison of two on-line risk calculators versus the detection of circulating prostate cells for the detection of high risk prostate cancer at first biopsy.]

OBJECTIVE: The limitations of total serum PSA values remains problematic; nomograms may improve the prediction of a positive prostate biopsy (PB). We compare in a prospective study of Chilean men with suspicion of prostate cancer due to an elevated total serum PSA and/or abnormal digital rectal examination, the use of two on-line nomograms with the detection of primary malignant circulating prostate cells (CPCs) to predict a positive PB for high risk prostate cancer. METHODS: Consecutive men with suspicion of prostate cancer underwent 12 core TRUS prostate biopsy. Age, total serum PSA and percent free PSA, family history, ethnic origin and prostate ultrasound results were registered. Risk assessment was performed using the online nomograms. The European Randomized Study of Screening for Prostate Cancer derived Prostate Risk Indicator (SWOP-PRI) and the North American Prostate Cancer Prevention Trail derived Prostate Risk Indicator (PCPT-CRC) were used to calculate risk of prostate cancer. Immediately before PB an 8 ml blood sample was taken to detect CPCs. Mononuclear cells were obtained by differential gel centrifugation and identified using double immunomarcation with anti-PSA and anti- P504S. Biopsies were classified as cancer/no-cancer, CPC detection test as negative/positive and the total number of cells/8ml registered. Areas under the curve (AUC) for total serum PSA, free percent, PSA, SWOP-PRI, PCPT-CRC and CPCs were calculated and compared. Diagnostic yields were calculated, including the number of possible biopsies that could be avoided and the number of clinically significant cancers that would be missed. RESULTS: 1,223 men aged 〉 55 years were analyzed, 467 (38.2%) had a biopsy positive for cancer of which 114/467 (24.45) complied with the criteria for active observation; 177/467 (36.8%) were Gleason 7 or higher. Discriminative power of detecting prostate cancer, showed areas under the curve of total PSA 0.559, SWOP nomogram 0.687, PCPTRC nomogram 0.716, free percent PSA 0.765 and CPC detection 0.844. CPC detection was superior to the other models (p〈0.0001). Using the recommended cutoff values, free percent PSA avoided 81% of biopsies missing 58% of significant cancers; for the other models the values were SWOP 75% and 56%; PCPTRC 61% and 62%, CPC detection 57% and 4% respectively. CONCLUSIONS: CPC detection was superior to the other models in predicting the presence of clinically significant prostate cancer at initial biopsy; potentially reduces the number of unnecessary biopsy while missing few significant cancers. Being a positive/negative test it avoids defining a cutoff value which may differ between populations. Multicenter studies to validate this method are warrented.