Model demonstrated good discrimination for 15-year prostate cancer-specific mortality in validation cohorts
TUESDAY, March 12, 2019 (HealthDay News) — An individualized, multivariable model can predict survival in patients with nonmetastatic prostate cancer (PCa), according to a study published online March 12 in PLOS Medicine.
David R. Thurtle, B.M.B.S., from the University of Cambridge in the United Kingdom, and colleagues collated data for 10,089 men diagnosed with nonmetastatic PCa between 2000 and 2010 in Eastern England. Patients were followed for a median of 9.8 years, with 3,829 deaths (1,202 PCa-specific). Data were also included for 2,546 men diagnosed in Singapore who represented an external validation cohort. Using separate multivariable models, 15-year PCa-specific mortality (PCSM) and non-PCa mortality (NPCM) were examined. A model estimating individualized 10- and 15-year survival outcomes was developed combining age, prostate-specific antigen, histological grade, biopsy core involvement, stage, and primary treatment, which were all prognostic factors for PCSM, and age and comorbidity that were prognostic for NPCM.
The researchers found that the model demonstrated good discrimination in the U.K. and Singapore validation cohorts with a C-index of 0.84 and 0.84, respectively, for 15-year PCSM, which compared favorably with international risk-stratification criteria. For overall mortality, discrimination was maintained, with a C-index of 0.77 and 0.76, respectively.
“For the first time to our knowledge, this simultaneously presents individualized estimates of cancer-specific and overall survival outcomes and can model the impact of treatment on these outcomes,” the authors write.
Copyright © 2019 HealthDay. All rights reserved.