Summary Background The prostate-specific antigen (PSA) test is used to screen for prostate cancer but has a high false-positive rate that translates into unnecessary prostate biopsies and ...overdiagnosis of low-risk prostate cancers. We aimed to develop and validate a model to identify high-risk prostate cancer (with a Gleason score of at least 7) with better test characteristics than that provided by PSA screening alone. Methods The Stockholm 3 (STHLM3) study is a prospective, population-based, paired, screen-positive, diagnostic study of men without prostate cancer aged 50–69 years randomly invited by date of birth from the Swedish Population Register kept by the Swedish Tax Agency. Men with prostate cancer at enrolment were excluded from the study. The predefined STHLM3 model (a combination of plasma protein biomarkers PSA, free PSA, intact PSA, hK2, MSMB, MIC1, genetic polymorphisms 232 SNPs, and clinical variables age, family, history, previous prostate biopsy, prostate exam), and PSA concentration were both tested in all participants enrolled. The primary aim was to increase the specificity compared with PSA without decreasing the sensitivity to diagnose high-risk prostate cancer. The primary outcomes were number of detected high-risk cancers (sensitivity) and the number of performed prostate biopsies (specificity). The STHLM3 training cohort was used to train the STHLM3 model, which was prospectively tested in the STHLM3 validation cohort. Logistic regression was used to test for associations between biomarkers and clinical variables and prostate cancer with a Gleason score of at least 7. This study is registered with ISCRTN.com , number ISRCTN84445406. Findings The STHLM3 model performed significantly better than PSA alone for detection of cancers with a Gleason score of at least 7 (p<0·0001), the area under the curve was 0·56 (95% CI 0·55–0·60) with PSA alone and 0·74 (95% CI 0·72–0·75) with the STHLM3 model. All variables used in the STHLM3 model were significantly associated with prostate cancers with a Gleason score of at least 7 (p<0·05) in a multiple logistic regression model. At the same level of sensitivity as the PSA test using a cutoff of ≥3 ng/mL to diagnose high risk prostate cancer, use of the STHLM3 model could reduce the number of biopsies by 32% (95% CI 24–39) and could avoid 44% (35–54) of benign biopsies. Interpretation The STHLM3 model could reduce unnecessary biopsies without compromising the ability to diagnose prostate cancer with a Gleason score of at least 7, and could be a step towards personalised risk-based prostate cancer diagnostic programmes. Funding Stockholm County Council (Stockholms Läns Landsting).
Artificial intelligence (AI) as an independent reader of screening mammograms has shown promise, but there are few prospective studies. Our aim was to conduct a prospective clinical trial to examine ...how AI affects cancer detection and false positive findings in a real-world setting.
ScreenTrustCAD was a prospective, population-based, paired-reader, non-inferiority study done at the Capio Sankt Göran Hospital in Stockholm, Sweden. Consecutive women without breast implants aged 40–74 years participating in population-based screening in the geographical uptake area of the study hospital were included. The primary outcome was screen-detected breast cancer within 3 months of mammography, and the primary analysis was to assess non-inferiority (non-inferiority margin of 0·15 relative reduction in breast cancer diagnoses) of double reading by one radiologist plus AI compared with standard-of-care double reading by two radiologists. We also assessed single reading by AI alone and triple reading by two radiologists plus AI compared with standard-of-care double reading by two radiologists. This study is registered with ClinicalTrials.gov, NCT04778670.
From April 1, 2021, to June 9, 2022, 58 344 women aged 40–74 years underwent regular mammography screening, of whom 55 581 were included in the study. 269 (0·5%) women were diagnosed with screen-detected breast cancer based on an initial positive read: double reading by one radiologist plus AI was non-inferior for cancer detection compared with double reading by two radiologists (261 0·5% vs 250 0·4% detected cases; relative proportion 1·04 95% CI 1·00–1·09). Single reading by AI (246 0·4% vs 250 0·4% detected cases; relative proportion 0·98 0·93–1·04) and triple reading by two radiologists plus AI (269 0·5% vs 250 0·4% detected cases; relative proportion 1·08 1·04–1·11) were also non-inferior to double reading by two radiologists.
Replacing one radiologist with AI for independent reading of screening mammograms resulted in a 4% higher non-inferior cancer detection rate compared with radiologist double reading. Our study suggests that AI in the study setting has potential for controlled implementation, which would include risk management and real-world follow-up of performance.
Swedish Research Council, Swedish Cancer Society, Region Stockholm, and Lunit.