Purpose: The Partin nomogram uses preoperative Gleason grade, serum prostate specific antigen and clinical stage to predict pathological outcome after radical prostatectomy. It was developed and validated in a select population of patients at 3 academic institutions. Although the nomogram is widely used, it has yet to be validated in a community based population. We assessed the performance of the nomogram in the Cancer of the Prostate Strategic Urological Research Endeavor, a nationwide, community based observational disease registry of men with prostate cancer.
Materials and methods: Included in the cohort were 1,162 men in Cancer of the Prostate Strategic Urological Research Endeavor who underwent radical prostatectomy. Using various probability thresholds the nomogram was used to predict outcomes in each patient. Using different probability thresholds receiver operating characteristics curves were then used to assess test performance.
Results: Of the men 860 (74%) had organ confined disease, 179 (15%) had established capsular penetration, 95 (8%) had seminal vesicle involvement and 37 (3%) had lymph node involvement. Calculated receiver operating characteristics curve area was 0.684 for predicting organ confined disease, 0.614 for predicting capsular penetration, 0.726 for predicting seminal vesicle involvement and 0.766 for predicting lymph node involvement. These values were lower than in previously published reports.
Conclusions: While the Partin nomogram performs adequately in a community based cohort of men who undergo radical prostatectomy for localized prostate cancer, it does not attain the success demonstrated in previous studies in select academic cohorts. This result was likely due to differences in the distribution of pathological outcomes in the community based cohort, in which more men had organ confined disease.