The association between tumour density and prostate cancer recurrence following radical prostatectomy.

PURPOSE Tumour density (TD) may be an independent prognostic factor in men with prostate cancer. The purpose of this study was to evaluate the association between prostate cancer TD and recurrence following radical prostatectomy. MATERIALS AND METHODS Between 1995 and 2007, 645 patients from The Ottawa Hospital or Memorial Sloan-Kettering Cancer Center who had cancer and prostate volumes measured from radical prostatectomy specimens. Tumour density was defined as the relative tumour to prostate volume (tumour volume/prostate volume) and recurrence was defined as a prostate-specific antigen (PSA) >0.2 ng/mL and rising, or postoperative use of radiation or hormonal therapy. Associations between TD and recurrence are adjusted for preoperative PSA, prostatectomy Gleason sum, tumour stage and margin status. RESULTS Median follow-up was 40.8 months. Tumour density was associated with preoperative PSA, Gleason sum, tumour stage and surgical margin status (all p < 0.0001). As a continuous variable, TD predicted recurrence-free survival (adjusted HR 1.34 per 10% increase in TD; p = 0.04). As a categorical variable, the group of patients with a TD of >10% had a 2.7 times greater hazard of recurrence compared to patients with a TD <5% (95%CI 1.41, 5.19; p = 0.003). Despite the independent association between TD and recurrence, the clinical value of TD remains in question as the discriminative performance (area under the curve) of predictive models only improved from 0.865 to 0.876. CONCLUSIONS Prostate cancer TD is associated with known prognostic factors and is also independently predictive of recurrence following radical prostatectomy.