Prediction of prostate cancer in extended-field biopsies of the prostate

Winkler, M. H., Kulinskaya, E. and Gillatt, D. G. (2004) Prediction of prostate cancer in extended-field biopsies of the prostate. BJU International, 93 (4). pp. 516-521. ISSN 1464-4096

Full text not available from this repository. (Request a copy)

Abstract

OBJECTIVES: To assess the prediction of prostate cancer using extended-field prostatic biopsies (8–11 cores), as such biopsy protocols are recommended to increase the detection of prostate cancer, and as fewer cancers are missed this should improve the prediction of biopsy outcome from the patients’ history, transrectal ultrasonography (TRUS) and serum markers. PATIENTS AND METHODS: In all, 260 patients were prospectively evaluated and 206 with a total prostate-specific antigen (PSA) level of < 20 ng/mL were included. All patients were evaluated for age, family history, lower urinary tract symptoms (LUTS), medication for LUTS, previous prostate biopsy, the presence of cysts, a digital rectal examination, calcifications or hypoechoic lesions on TRUS, total and transitional zone volume, total PSA (tPSA), PSA density (tPSAD), total PSA transition zone density (tPSATZD), complexed PSA (cPSA), cPSA density (cPSAD), cPSA transitional zone density (cPSATZD), free/total (f/t)PSA ratio and free/complexed PSA ratio (f/cPSA). Logistic regression was used to predict the outcome; 80% of the patients were used to generate the models and 20% to test the prediction. RESULTS: Two models were constructed; the most accurate contained family history, cPSA, cPSAD, cPSATZD, f/cPSA, PSAD and tPSATZD (sensitivity 91%, specificity 70%). A workable and concise model contained tPSATZD, cPSATZD and f/cPSA, and had a sensitivity of 93% and a specificity of 60%. The best single predictor was tPSATZD with a sensitivity of 92% and a specificity of 55%. Using regression models can produce considerable gains in specificity. This would allow unnecessary prostate biopsies to be avoided for a third of patients compared with tPSA alone. CONCLUSIONS: The present analysis for PSA indices appeared to be slightly more accurate than those in previously published studies. Most of this improvement in diagnostic accuracy was ascribed to the use of an extended-field biopsy protocol. Prostate cancer in a first-degree relative was the only variable that contributed significantly to the regression model. tPSATZD was the best volume-adjusted PSA index. The f/tPSA appeared to be the best test with no volume adjustment, followed by f/cPSA and cPSA. Although the models are cumbersome and expensive for use in general urological practice they could be used to optimize biopsy strategies on the basis of predicted cancer probabilities in screening studies. The cost of the models may compare favourably with tPSA because of the high specificity that can be achieved.

Item Type:	Article
Uncontrolled Keywords:	sdg 3 - good health and well-being ,/dk/atira/pure/sustainabledevelopmentgoals/good_health_and_well_being
Faculty \ School:	Faculty of Science > School of Computing Sciences
UEA Research Groups:	Faculty of Science > Research Groups > Data Science and Statistics Faculty of Medicine and Health Sciences > Research Centres > Business and Local Government Data Research Centre (former - to 2023) Faculty of Science > Research Groups > Norwich Epidemiology Centre Faculty of Medicine and Health Sciences > Research Groups > Norwich Epidemiology Centre
Depositing User:	Vishal Gautam
Date Deposited:	21 Jul 2011 18:26
Last Modified:	17 Apr 2023 23:44
URI:	https://ueaeprints.uea.ac.uk/id/eprint/23600
DOI:	10.1111/j.1464-410X.2003.04670.x

Actions (login required)

View Item