Prostate-specific antigen to predict outcome of external beam radiation for prostate cancer: Walter Reed Army Medical Center experience, 1988-1995.

Journal Article


To assess the ability of pretreatment and post-treatment prostate-specific antigen (PSA) measurements, clinical tumor stage, tumor grade, Gleason sum, race, age, and radiation dose to predict the recurrence of prostate cancer following external beam radiation therapy (XRT) since the introduction of PSA as a tumor marker at one tertiary care center.


The recurrence of prostate cancer among 371 evaluable patients of 389 patients treated with XRT at Walter Reed Army Medical Center was analyzed using Kaplan-Meier survival methodology and Cox multivariable regression models. Serologic (PSA) recurrence was determined using three consecutive rises in PSA after a nadir value. Clinical recurrence was defined as local recurrence (palpable or positive biopsy) and/or distant (radiographically evident) recurrence. Mean and median follow-up is 40.2 and 39.4 months, respectively (range 3.0 to 89.5), and minimum follow-up is 18 months for patients who were alive at the time of analysis. No patient received adjuvant hormonal therapy. Potential prognostic factors evaluated are pretreatment PSA, PSA nadir, age, race, clinical tumor stage, tumor grade, Gleason sum, and radiation dose.


Of the 371 evaluable patients, 125 had disease recurrence. The Kaplan-Meier 5-year disease-free survival (DFS) rates for significant pretreatment variables in univariate analyses are as follows: pretreatment PSA less than 4 (79%), 4.1 to 10 (67%), 10.1 to 20 (57%), 20.1 to 50 (27%), and more than 50 (0%); for clinical tumor Stage T1a-T1c (84%), T2a-T2c (51 %), and T3-T4 (29%); for tumor grade well (58%), moderate (58%), and poor (30%). Four-year DFS rates for Gleason sum are 2 to 4 (82%), 5 (72%), 6 (56%), and 7 to 10 (48%). In multivariable Cox regression analysis with backward elimination of nonsignificant variables, age, race, tumor grade, and radiation dose were eliminated, leaving pretreatment PSA, clinical tumor stage, and Gleason sum as significant prognostic factors. Analysis of a Cox model that included nadir PSA as a time-dependent variable showed it to be the strongest prognostic factor variable in the analysis.


XRT remains a suitable treatment modality for patients with pretreatment PSA less than 20.0, clinical tumor Stages T1-T2, and Gleason sum 2 to 6 prostate cancer. Patients achieving a nadir value less than 0.5 have more durable treatment outcomes.

Full Text

Duke Authors

Cited Authors

  • Preston, DM; Bauer, JJ; Connelly, RR; Sawyer, T; Halligan, J; Leifer, ES; McLeod, DG; Moul, JW

Published Date

  • January 1999

Published In

Volume / Issue

  • 53 / 1

Start / End Page

  • 131 - 138

PubMed ID

  • 9886602

Pubmed Central ID

  • 9886602

Electronic International Standard Serial Number (EISSN)

  • 1527-9995

International Standard Serial Number (ISSN)

  • 0090-4295

Digital Object Identifier (DOI)

  • 10.1016/s0090-4295(98)00464-6


  • eng