Aims Lymph node metastasis (N1) is an adverse prognostic factor for men with clinically localised prostate cancer (PCa), but the prediction of N1 disease remains difficult. Urokinase plasminogen activator receptor (uPAR) plays an important role in angiogenesis and tumorigenesis. We analysed whether plasma levels of the soluble uPAR forms uPAR(I-III), uPAR(II-III) and uPAR(I) were associated with the risk of N1 disease in men with clinically localised PCa.

Methods The present study includes all men (n=518) who underwent radical prostatectomy (RP) for clinically localised PCa, 29 of whom had N1 disease. Soluble uPAR forms were measured using three time-resolved fluorescence immunoassays. The prognostic value of the different uPAR forms together with clinicopathological parameters for N1 disease were analysed using logistic regression, receiver operating characteristic (ROC) regression analysis and quantified using the areas under the ROC curve (AUC).

Results All soluble uPAR levels were significantly (p=0.03) higher in patients with N1 disease compared with patients with N0/x disease. ROC curves including clinical tumour stage, biopsy Gleason score, prostate-specific antigen and percent positive biopsies had an AUC of 87.7% for prediction of N1 disease. With the addition of uPAR(I) to the model, the AUC increased to 88.4%.

Conclusions Addition of uPAR(I) level to known diagnostic parameters did not increase the prediction of N1 disease following RP in men with clinically localised PCa. Our results indicate that the plasma levels at diagnosis of the different uPAR forms do not hold important predictive or prognostic information in men with clinically localised PCa.