False-positive screening results in the European randomized study of screening for prostate cancer

Abstract

Background

Screening for prostate cancer (PC) with prostate-specific antigen (PSA) has been shown to decrease mortality, but has adverse effects, such as false-positive (FP) screening results. We describe the frequency of FP results and assess their relation to subsequent screening attendance, test results and prostate cancer risk in a large randomized trial.

Materials and methods

We included data from five centres of the European Randomized Study of Screening for Prostate Cancer, altogether over 61,000 screened men. Men were screened with PSA test at a 2–7 year interval depending on the centre; PSA cut-off was 3.0–4.0 ng/ml. A positive screen with no histologically confirmed PC in biopsy within 1 year was defined as an FP result.

Results

Of the 61,604 men who were screened at least once, 17.8% had one or more FP result(s). Almost 20% of men who participated at all screening rounds had one or more FP result(s). More than half of the men with an FP result had another FP if screened again. Men with FP results had a fourfold risk of PC at subsequent screen (depending on the round, 10.0% versus 2.6–2.7% of men with negative screen, risk ratio 3.8–3.9). The PCs following an FP result were in 92.8% of cases localised and low-grade versus 90.4% following a screen-negative result.

Conclusions

Our results show that FP results are common adverse effects in PC screening, as they affect at least one in six screened men. False-positive men are more prone to be diagnosed with PC but are also likely to have consistently high PSA levels.

Introduction

The benefits of screening for prostate cancer (PC) with prostate-specific antigen (PSA) test are becoming clearer with evidence from the European Randomized Study of Screening for Prostate Cancer (ERSPC), which has shown a 20% relative decrease in mortality in the screening arm.¹ However, the adverse effects and cost-effectiveness of screening need to be thoroughly investigated before decisions regarding population-based screening can be made.

Serum PSA is an organ-specific marker that may be affected by any prostatic disease. Therefore, as all screening tests, it is not perfect in sensitivity or specificity. Undetected disease constitutes a false negative (FN) finding and a positive screening result in the absence of disease is a false positive (FP) result (Table 1). The challenge in PC screening is to define and predict the disease status based on the PSA test, as not all subjects can undergo the diagnostic test, the prostatic biopsy. Even the biopsy has uncertainties: first, the needle biopsy provides only a small sample of the prostate tissue potentially missing the cancer lesion and second, the presence of malignant tissue does not necessarily mean clinically significant PC (resulting in overdiagnosis). Overdiagnosis and overtreatment occur when PCs that would not have been diagnosed in the absence of screening are detected by screening and treated.2, 3

In addition, screening for PC with PSA has relatively high FP rate, or conversely, low specificity.⁴ Previously, the results from the Finnish component of the ERSPC trial have shown that 12.5% of the screened men (at 4-year interval) had an FP result at least once during three screening rounds.⁵ Similarly, 10.4% of men in the Prostate, Lung, Colorectal and Ovarian cancer screening trial (PLCO) had an FP result during four PSA tests and 3 years of follow-up.⁶

We present the proportion of FP results during three screening rounds in five centres of the ERSPC trial: Belgium, Finland, Italy, the Netherlands and Sweden – with more than 61,000 screened men. We also investigated subsequent screening compliance, PC risk and repeated FP result(s).