Physicians currently use many variables to estimate the probability of prostate cancer recurrence following surgical removal of the whole or part of the prostate gland. In a study appearing online in advance of publication in the July print issue of the Journal of Clinical Investigation, a team of researchers led by Michael Donovan at Aureon Laboratories report their development of a novel model for the improved prediction of prostate cancer recurrence. The authors examined many clinical variables in more than 300 patients and used multidisciplinary methodology they term "systems pathology" in order to develop their predictive model. They applied their predictive model to a separate cohort of patients and showed that the model can successfully predict prostate cancer recurrence (defined by rising serum levels of prostate-specific antigen) with 82% predictive accuracy, 96% sensitivity, and 72% specificity. This model has broad applications for the diagnosis, treatment management, and prediction of recurrence of prostate cancer.
TITLE: Improved prediction of prostate cancer recurrence through systems pathology
AUTHOR CONTACT:
Michael J. Donovan
Aureon Laboratories Inc., Yonkers, New York, USA.
###
JCI table of contents: June 7, 2007
Contact: Brooke Grindlinger
Journal of Clinical Investigation