Partin Tables cannot accurately predict the pathological stage at radical prostatectomy

Area of Science:

• Urology
• Surgical Oncology
• Medical Statistics

Background:

• The Partin Tables are a critical tool for staging patients undergoing radical prostatectomy (RP).
• Predictions from the Partin Tables influence surgical decisions regarding nerve preservation and extent of resection.
• Accurate staging is essential for optimizing treatment strategies and patient outcomes.

Purpose of the Study:

• To externally validate the predictive accuracy of the Partin Tables for radical prostatectomy candidates.
• To assess the reliability of Partin Tables' predictions for extracapsular extension (ECE), seminal vesicle invasion (SVI), and lymph node invasion (LNI).
• To determine if current Partin Tables are sufficiently accurate for pre-operative decision-making.

Main Methods:

• External validation using a dataset of 3105 patients treated with RP at a European institution.
• Utilized standard validation metrics, including the area under the receiver operating characteristics curve (AUC).
• Evaluated predictions for ECE, SVI, and LNI using both 2001 and 2007 versions of the Partin Tables.

Main Results:

• The AUC values for the 2001 Partin Tables were 79.7% (ECE), 77.8% (SVI), and 73.0% (LNI).
• The AUC values for the 2007 Partin Tables were 79.8% (ECE), 80.5% (SVI), and 76.2% (LNI).
• A poor correlation was observed between the predicted probabilities and the actual observed rates for all endpoints.

Conclusions:

• The Partin Tables' predictions showed suboptimal accuracy for guiding critical surgical decisions in radical prostatectomy.
• The current format of the Partin Tables is not precise enough to influence pre-operative choices regarding nerve sparing or extent of resection.
• Further refinement of staging tools is necessary to improve surgical planning and patient management in radical prostatectomy.