The use of prostate-specific antigen testing in the detection of localized prostate cancer: Current opinion and urological practice in the United Kingdom

Background: The prostate-specific antigen (PSA) test and itsinterpretation plays a crucial role in the detection of earlylocalized prostate cancer. However, inaccuracy of the test,inability to predict the aggressiveness of the disease and thelack of evidence about the comparative effectiveness of treatmentshave led to major dilemmas in considering whether to employthe PSA test and which cut-off points to use in interpretingits results. The aim of this study was to evaluate current urologicalpractice in the UK regarding the use of PSA testing. Methods:A postal questionnaire survey of all consultant urologist membersof the British Association of Urological Surgeons was conducted.Statistical analysis included proportional odds regression modelsto examine factors associated with urologists' preferences fordifferent definitions of ‘normal’ PSA cut-off levels.Results: The survey response rate was 60%. The majority of consultanturologists applied the PSA test routinely. There was a highlevel of agreement amongst UK urologists on normal PSA cut-offpoints (<4.0 ng/ml) for asymptomatic men under 60 years ofage. There was very wide variation in the definition of normalPSA cut-offs for older (60 years) asymptomatic men. A preferencefor lower cut-off points, leading to investigation with ultrasoundand biopsy, was significantly associated with larger urologydepartment size, the presence of a prostate cancer subspecialistin the department and relatively short length of specializationin urology. Conclusions: Prostate cancer screening and earlydetection practices and reported incidence rates of the diseaseare likely to be influenced by variation in urologists' interpretationsof PSA. Despite increasing evidence in favour of lower PSA cut-offlevels, particularly for younger men (<60 years), urologistsin the UK are divided over their interpretation. Men, particularlyover age 60 years, have varying chances of further investigationfollowing PSA testing. Any trial of prostate cancer screeningor treatment should take this potential variation into account.Standard protocols for PSA interpretation should be implemented.