Multivariate models to predict clinically important outcomes at prostatectomy for patients with organ-confined disease and needle biopsy Gleason scores of 6 or less

David S. Dimarco, Michael L. Blute, Horst Zincke, John C. Cheville, Darren L. Riehle, Christine M. Lohse, V. Shane Pankratz, Thomas J. Sebo

Research output: Contribution to journalArticle

9 Scopus citations


The objective of this study was to determine the clinical and biopsy features associated with outcomes at radical retropubic prostatectomy (RRP) in patients with clinically organ-confined prostate cancers and biopsy Gleason scores (GS) of 6 or less. We reviewed 274 biopsies with GS 6 or less cancers from patients with clinically organ-confined disease between 1995 and 1998 to determine statistically significant predictors for the following outcomes at RRP: tumor volume, small (<0.5 cc), confined (pT2) tumors with RRP GS of 6 or less (potentially "insignificant" tumors), and extraprostatic extension (EPE). Clinical and pathologic features evaluated included age, serum prostate specific antigen (PSA), clinical stage, percent biopsy cores and surface area positive for cancer (tumor extent), perineural invasion, MIB-I proliferation, and DNA ploidy by digital image analysis (DIA). Multivariate analyses showed that biopsy tumor extent (median percent surface area positive 3.3%; P < 0.001 and median biopsy cores positive 28.6%; P = 0.001) and PSA (median 5.5 ng/mL; P = 0.009) predicted tumor volume (median 1.4 cc). Biopsy tumor extent (P = 0.002), PSA (P = 0.002), and percent S-phase nuclei (P = 0.050) predicted potentially "insignificant" tumors at RRP (n = 76, 28%). Percent surface area positive for cancer (P = 0.003) predicted EPE (n = 22, 8%). DNA ploidy (n = 211, 79% diploid) and MIB-I proliferation (median 1.4%) did not add information to predict these RRP outcomes. Biopsy tumor extent and serum PSA were significantly associated with tumor volume. Biopsy tumor extent, serum PSA, and percent S-phase nuclei by DIA were predictive of potentially insignificant tumors. Patients with clinically confined disease, <5% biopsy surface area positive for cancer, <20% biopsy cores positive for cancer, and GS 6 or less, had a 48% chance of having a potentially insignificant tumor at diagnosis if the serum PSA was <10 ng/mL. Percent surface area predicted EPE at RRP. DNA ploidy and MIB-I proliferation by DIA did not provide additional information.

Original languageEnglish (US)
Pages (from-to)439-446
Number of pages8
JournalUrologic Oncology: Seminars and Original Investigations
Issue number6
StatePublished - Jan 1 2003



  • Clinically confined
  • DNA ploidy
  • Digital image analysis
  • Gleason score
  • Insignificant tumor
  • MIB-I
  • Needle biopsy
  • Prostate carcinoma

ASJC Scopus subject areas

  • Oncology
  • Urology

Cite this