Prediction of cancer relapse after prostatectomy: use the postoperative nomogram.

To the Editor: Shariat and colleagues (1) have identified several serum biomarkers associated with disease progression in prostate cancer. The addition of these biomarkers in a standard predictive model, the preoperative Kattan nomogram, greatly improved the prediction accuracy. This report is a major step in prostate cancer management because, despite increasing number of published biomarker studies, very few studies have led to efficient translational application. The objective was to determine the added value of a set of markers relative to established predictors of recurrence included in the preoperative Kattan nomogram: preoperative prostatespecific antigen, clinical stage, and Gleason score on biopsies. However, because the study focused on patients treated by radical prostatectomy, it could have been more pertinent to use the postoperative Kattan nomogram obtained after histopathologic examination of the surgical specimen, instead of the preoperative tool. In fact, this postoperative nomogramhas been not only validated several times since 1999 but also shown to provide more accurate prediction than the preoperative model (2, 3). The clinical stage and biopsy Gleason score included in the preoperative tool only approximate the real pathologic stage and Gleason score determined on surgical specimens, with frequent understaging and undergrading (4). Moreover, the postoperative model also includes an important predictor of disease recurrence, namely, the surgical margin status (5). Therefore, for an optimal translational application of their findings, Shariat and colleagues have now to provide evidence that the serum biomarkers also improved the accuracy of the postoperative prediction models.