Modelling study shows less-intense screening is okay for men with low baseline PSA levels

Référence

Heijnsdijk EAM, Gulati R, Tsodikov A, et al. Lifetime benefits and harms of prostate-specific antigen–based risk-stratified screening for prostate cancer. J Natl Cancer Inst 2020;112(10):1013-1020.

Sommaire

Screening for prostate cancer has been controversial for several reasons, some of which are directly related to balancing the potential benefits against the harms of screening. These researchers used 2 independently developed computer simulation models to evaluate the trade-offs of less frequent screening in men aged 45 or 50 years with a baseline PSA below 1.0 ng/mL and the cessation of screening altogether if the PSA is below 1.0 ng/mL in men 60 years of age. The idea of less frequent screening means fewer direct harms of pursuing false positive results and reducing overdiagnosis, but at the risk of (potentially) more cancers missed and more cancer deaths. The models used virtual populations in which individuals can develop prostate cancer and their cancers can progress. The authors also compared model performance with real-world data from the Malmo Preventive Project. The models started with virtual men 45 or 50 years of age and followed them up until age 85. The models evaluated several strategies: screening every 2 years until age 69; lengthening the screening interval to 8 years for a baseline PSA of less than 1.0 ng/mL but decreasing again to 2 years if a subsequent PSA is greater than 1.0 ng/mL; stopping screening when PSA is less than 1.0 ng/mL at age 60 years and older; or a combination of the last 2 strategies. Regardless of strategy, the models used PSA levels exceeding 4.0 ng/mL to trigger further investigation. The investigators found that the models reasonably replicated the PSA distributions seen in the real-world Malmo data. Using a base model, screening 10,000 men biennially from ages 45 to 69 would require more than 100,000 screens; one model would yield 277 overdiagnoses and the other would yield 348 overdiagnoses. Additionally, they would save 110 to 160 lives and gain 921 to 1312 life-years relative to no screening. Using the risk-based screening, the models predicted the number of screenings would nearly be cut in half (47% reduction) and the overdiagnoses reduced by 2.1% or 0.9%, respectively. However, the models also saw a small reduction in lives saved (3.8% and 3.1%, respectively). Additionally, stopping screening if the PSA was less than 1.0 ng/mL at 60 years of age or older decreased overdiagnoses (24.0% and 5.5% relative reductions, respectively) and lives saved (13.1% and 5.0% relative reductions, respectively).