Shifting risk-stratified early prostate cancer detection to a primary healthcare setting.

Abstract

To evaluate the feasibility of multivariable risk stratification for early prostate cancer (PCa) detection in a primary healthcare diagnostic facility with regard to its effects on the referral rate and subsequent PCa diagnoses compared to a PSA threshold of 3.0 ng/mL as the current referral indicator. In 2014, the Erasmus MC Cancer Institute and the primary healthcare diagnostic facility STAR-SHL (located in Rotterdam city centre) initiated this observational study, in which general practitioners (GPs) could refer men who wished to undergo PCa screening to STAR-SHL for consultation by specially trained personnel. Referral recommendations to secondary healthcare were based on the outcome of application of the Rotterdam Prostate Cancer Risk Calculator (RPCRC) and were compared to the current Dutch GPs' PSA referral threshold of 3.0 ng/mL. For data collection on PCa diagnoses, the study cohort was linked to the Dutch nationwide pathology databank (PALGA). Between January 2014 and February 2021, 507 men were referred for consultation and in 495 men prostate-specific antigen (PSA) was tested. The median (interquartile range) follow-up from consultation to PALGA linkage was 43 (25-65)months. In total, 279 men (56%) had a PSA level ≥3.0 ng/mL, of whom 68% (95% confidence interval [95% CI] 63-74) were considered at low risk according to the RPCRC. Within 1year after consultation, one of these men (0.52%; 95% CI 0.092-2.9) was diagnosed with clinically significant (cs)PCa (i.e., International Society of Urological Pathology Grade Group ≥2). Thereafter, another four (2.1%; 95% CI 0.82-5.3) low-risk men were diagnosed with csPCa. Of the high-risk men who were biopsied within 1year after consultation (n = 61), 77% (95% CI 65-86) were diagnosed with PCa and 49% (95% CI 37-61) with csPCa. In a primary healthcare diagnostic facility, the RPCRC could reduce up to 68% of referrals to secondary healthcare, as compared to a PSA referral threshold of 3.0 ng/mL. Deploying the RPCRC in this setting resulted in a high csPCa detection rate in those men biopsied. This strategy can be considered safe since the observational data showed low proportions of csPCa among men atlow risk.