BACKGROUND: Accurate preoperative staging of prostate cancer (PCa) is important but current diagnostic methods cannot accurately determine extracapsular extension (ECE), resulting in the possible triage of patients towards a less appropriate arm of therapy. This has consequences to patient care and better methods of preoperatively determining ECE are required.
METHODS: We followed a biomarker development pathway and compared the preoperative serum expressions of VEGF-D, PEDF, IGF-I, IGFBP3, and CD14 in patients from the Irish Prostate Cancer Research Consortium (PCRC) with radical prostatectomy determined ECE against patients with nonECE.
RESULTS: The expression measurements of five proteins were fitted into a logistic regression model and backwards variable elimination methods were applied which resulted in a model with IGFBP3 and CD14 as the best combination biomarker panel. This panel was tested in an independent cohort of patients using an optimized multiplex electrochemiluminescence assay. Receiver operating characteristic curves were generated and the areas under the curve (AUC) were calculated as an estimation of prediction accuracy. The biomarker panel was validated with an AUC of 76.6%, and a sensitivity and specificity of 80% and 75% was obtained.
CONCLUSIONS: This is the first internally validated, preoperative serum biomarker panel that identifies ECE in patients with Gleason score 7 PCa with AUC 76.6%. The panel surpasses the routinely used diagnostic standards in accuracy and may help to improve preoperative cancer staging, better inform treatment options, and improve the referral patterns of patients with urgently treatable cancers towards more appropriate arms of therapy. Prostate © 2012 Wiley Periodicals, Inc.
Written by:
Oon SF, Fanning DM, Fan Y, Boyce S, Murphy TB, Fitzpatrick JM, Watson RW Are you the author?
UCD School of Medicine and Medical Science, Dublin, Ireland; UCD Conway Institute of Biomolecular and Biomedical Science, Dublin, Ireland. .
Reference: Prostate. 2012 Mar 13
doi: 10.1002/pros.22506.
PubMed Abstract
PMID: 22415934