BACKGROUND: Genome-wide association studies have identified single nucleotide polymorphisms (SNPs) associated with higher risk of prostate cancer (PCa).
This study aimed to evaluate whether published SNPs improve the performance of a clinical risk-calculator in predicting prostate biopsy result.
METHODS: Three hundred forty-six patients with a previous prostate biopsy (191 positive, 155 negative) were enrolled. After literature search, nine SNPs were selected for their statistically significant association with increased PCa risk. Allelic odds ratios were computed and a new logistic regression model was built integrating the clinical risk score (i.e., prior biopsy results, PSA level, prostate volume, transrectal ultrasound, and digital rectal examination) and a multilocus genetic risk score (MGRS). Areas under the receiver operating characteristic (ROC) curves (AUC) of the clinical score alone versus the integrated clinic-genetic model were compared. The added value of the MGRS was assessed using the Integrated Discrimination Improvement (IDI) and Net Reclassification Improvement (NRI) statistics.
RESULTS: Predictive performance of the integrated clinico-genetic model (AUC = 0.781) was slightly higher than predictive performance of the clinical score alone (AUC = 0.770). The prediction of PCa was significantly improved with an IDI of 0.015 (P-value = 0.035) and a continuous NRI of 0.403 (P-value < 0.001).
CONCLUSIONS: The predictive performance of the clinical model was only slightly improved by adding MGRS questioning the real clinical added value with regards to the cost of genetic testing and performance of current inexpensive clinical risk-calculators.
Written by:
Butoescu V, Ambroise J, Stainier A, Dekairelle AF, Gala JL, Tombal B. Are you the author?
Service d'Urologie, Institut de Recherche Expérimentale et Clinique (IREC), Cliniques universitaires Saint Luc, Université catholique de Louvain, Brussels, Belgium.
Reference: Prostate. 2013 Nov 22. Epub ahead of print.
doi: 10.1002/pros.22757
PubMed Abstract
PMID: 24265090
