Validation of 2001 Partin tables in Turkey: a multicenter study

Abstract

Objective: Although Partin tables were developed in United States to predict the stage of prostate cancer preoperatively, they are used by many clinics throughout the world assuming that these figures apply to their population as well. However the predictive value of current Partin tables, which was updated in 2001, has not been validated in most of the countries as well as in Turkey. Therefore, we evaluated the validity of 2001 Partin tables, for the ability to predict the pathological stage in Turkish patients.

Patients and methods: The clinical and pathological findings of 1043 patients who have had radical prostatectomy were assessed. Serum PSA values, clinical stage, biopsy Gleason score and the pathological features of the radical prostatectomy specimens were collected from each clinic and evaluated. The predictive value of Partin nomogram and pathological findings of prostatectomy specimens were compared and analyzed according to Receiver Operating Characteristics (ROC) analysis.

Results: Median age of the patients was 60 (45-74). In the presented study, percentage of patients with clinical stage T1c was 43%. Patients with Gleason score of 2-4 in biopsy constituted 23.4% of the study group. In the present study, the ratio of the patients with serum PSA higher than 10 ng/ml was 39.6%. Organ confined disease, seminal vesicle involvement, lymph node metastases ratios were 64.7%, 10.3%, 1.8% respectively. Area Under Curve (AUC) values for organ confined disease, seminal vesicle involvement and lymph node involvement were calculated as 0.665, 0.733 and 0.759 respectively.

Conclusion: It appears that Partin tables have a reasonable predictive value for the final pathological features like organ confined disease, seminal vesicle and lymph node involvement in Turkish patients. This multicenter study showed that current Partin tables could also be used in Turkish patients with comparable accuracy.