Repeat Prostate Cancer Screening using Blood-based Risk Prediction or Prostate-specific Antigen in the Era of Magnetic Resonance Imaging-guided Biopsies : A Secondary Analysis of the STHLM3-MRI Randomized Clinical Trial

Abstract

Background and objective: The use of blood-based risk prediction tools has been proposed to improve prostate cancer screening, but data on repeated screening are lacking. Our aim was to compare outcomes using the blood tests prostate-specific antigen (PSA) and Stockholm3 for repeat prostate cancer screening.

Methods: In the population-based screening-by-invitation STHLM3-MRI trial, men aged 50-74 yr were invited to participate in screening. At 2-3 yr after the initial round, men with PSA ≥1.5 ng/ml at trial inclusion who were randomized to magnetic resonance imaging (MRI)-enhanced screening and were not diagnosed with prostate cancer after the initial round were invited for repeat screening involving analysis of PSA and Stockholm3. Biparametric 1.5-T MRI was performed in cases with PSA ≥3 ng/ml or Stockholm3 ≥0.11. Men with Prostate Imaging0Reporting and Data System ≥3 lesions were referred for targeted plus systematic biopsies. The primary outcome was Gleason ≥7 cancer. Secondary outcomes included the number of MRI scans and biopsy procedures, and detection of Gleason 6 and Gleason ≥4 + 3 cancer. Outcomes were compared using the relative positive fractions (RPF).

Key findings and limitations: Of 7609 men from the initial screening round, 2078 were eligible for repeat screening and 1500 (72%) participated. For detection of Gleason ≥7 prostate cancer, the area under the receiver operating characteristic curve was 0.765 (95% confidence interval [CI] 0.725-0.805) for Stockholm3 and 0.651 (95% CI 0.601-0.701) for PSA. Stockholm3 ≥0.15 was associated with 41% fewer MRI scans in comparison to PSA ≥3 ng/ml (RPF 0.59, 95%CI 0.54-0.64), while the detection of GS ≥4 + 3 cancers was similar (RPF 1.00, 95% CI 0.78-1.29). Stockholm3 ≥0.15 detected fewer Gleason ≥7 (RPF 0.75, 95% CI 0.59-0.95) and Gleason 6 (RPF 0.73, 95% CI 0.46-1.16) cancers. Stockholm3 ≥0.11 was associated with no decrease in the number of MRI scans, but an increase of the number of cancer cases detected. Limitations include the lack of long-term outcomes.

Conclusions and clinical implications: Use of the Stockholm3 test for repeated prostate cancer screening could reduce the need for MRI while maintaining detection rates for high-risk cancer.

Patient summary: In this study, we invited men to a second round of prostate cancer screening. We found that use of a new blood test called Stockholm3 can make screening programs more efficient by using fewer resources while still detecting aggressive cancers.

Keywords: Biomarker; Magnetic resonance imaging; Prostate cancer; Screening.