An improved prognostic model for stage T1a and T1b prostate cancer by assessments of cancer extent Academic Article uri icon

Overview

MeSH Major

  • Adenocarcinoma
  • Prostatic Neoplasms
  • Transurethral Resection of Prostate

abstract

  • Treatment decisions on prostate cancer diagnosed by trans-urethral resection (TURP) of the prostate are difficult. The current TNM staging system for pT1 prostate cancer has not been re-evaluated for 25 years. Our objective was to optimise the predictive power of tumor extent measurements in TURP of the prostate specimens. A total of 914 patients diagnosed by TURP of the prostate between 1990 and 1996, managed conservatively were identified. The clinical end point was death from prostate cancer. Diagnostic serum prostate-specific antigen (PSA) and contemporary Gleason grading was available. Cancer extent was measured by the percentage of chips infiltrated by cancer. Death rates were compared by univariate and multivariate proportional hazards models, including baseline PSA and Gleason score. The percentage of positive chips was highly predictive of prostate cancer death when assessed as a continuous variable or as a grouped variable on the basis of and including the quintiles, quartiles, tertiles and median groups. In the univariate model, the most informative variable was a four group-split (≤10%, >10-25%, >25-75% and >75%); (HR=2.08, 95% CI=1.8-2.4, P<0.0001). The same was true in a multivariate model (ΔX(2) (1 d.f.)=15.0, P=0.0001). The current cutoff used by TNM (<=5%) was sub-optimal (ΔX(2) (1 d.f.)=4.8, P=0.023). The current TNM staging results in substantial loss of information. Staging by a four-group subdivision would substantially improve prognostication in patients with early stage disease and also may help to refine management decisions in patients who would do well with conservative treatments.

publication date

  • January 2011

Research

keywords

  • Academic Article

Identity

Language

  • eng

PubMed Central ID

  • PMC3853363

Digital Object Identifier (DOI)

  • 10.1038/modpathol.2010.182

PubMed ID

  • 20834240

Additional Document Info

start page

  • 58

end page

  • 63

volume

  • 24

number

  • 1