The optimal timing for discontinuing screening of prostate cancer (PCa) in elderly men is currently not known and remains debated.
Objective
To assess prostate cancer–specific mortality (PCSM) in elderly men who previously underwent prostate-specific antigen (PSA)-based screening and to identify those who may benefit from continued screening.
Design, setting, and participants
A total of 7052 men, who participated in the screening arm of the Rotterdam section of the European Randomized Study of Screening for Prostate Cancer and were aged 70–74 yr at their last screening visit after undergoing a maximum of three screening rounds without being diagnosed with PCa, were included.
Outcome measurements and statistical analysis
The cumulative incidence of PCSM by the age of 85 yr was assessed. Additionally, a competing risk regression was performed to assess the potential predictors of PCSM.
Results and limitations
The median follow-up was 16 yr. The cumulative incidence of PCSM by the age of 85 yr was 0.54% (95% confidence interval [CI]: 0.40–0.70) in all men, 0.11% (95% CI: 0.05–0.27) in men with PSA <2 ng/ml, 0.85% (95% CI: 0.47–1.5) in men with PSA 2–3 ng/ml, and 6.8% (95% CI: 3.1–15) in men with PSA ≥6.5 ng/ml and no previous benign biopsy. PSA (subdistribution hazard ratio [sHR]: 2.0; 95% CI: 1.7–2.3), previous benign prostate biopsy (sHR: 0.41; 95% CI: 0.23–0.72), and hypertension (sHR: 0.48; 95% CI: 0.25–0.91) were significantly associated with PCSM.
Conclusions
Men aged 70–74 yr who have previously undergone PSA-based screening without receiving a PCa diagnosis have a very low risk of dying from PCa by the age of 85 yr. These data suggest that screening may be discontinued in men with PSA <3.0 ng/ml or previous benign prostate biopsies. Those with higher PSA levels and no prior biopsies may consider continued screening if life expectancy exceeds 10 yr.
Recently, the 16-yr follow-up of the population-based European randomised study of screening for PCa (ERSPC) demonstrated that protocolled screening mostly based on PSA testing led to a 20%-reduction of the relative risk for prostate cancer (PCa) mortality (absolute difference of rate between screening and control groups: 0.18%) (PMID: 30824296).
In total, one PCa death could be averted by inviting 570 males for screening, while 18 men had to be diagnosed with PCa. In most centres, 4-yr intervals until the stopping age varying between 67 and 78 yr have been used. Practically, the question arises if some individuals at high risk for PCa mortality with a life expectancy of at least 10 yr. may still benefit from continued screening and possibly curative treatment, while others could safely terminate their early detection work-up and not be subjected to overdiagnosis and associated potential harms.
De Vos and collaborators assessed 7,052 men who participated in the Dutch screening arm of ERSPC, were aged 70-74 yr at the last screening visit and underwent a maximum of three screening rounds without being diagnosed with PCa, hence enabling for a follow-up of at least 15 yr (PMID: 37919190).
The primary outcome, cumulative incidence of PCa-specific mortality (PCSM) by the age of 85 yr, was 0.54% in the total cohort with no impact observed by the number of previous screens. While cumulative incidence of PCSM was only 0.11% in males with PSA <2 ng/ml, men with PSA ≥6.5 ng/ml and no previous benign biopsy exhibited a distinctly higher PCSM rate of 6.8%. Of note, cumulative incidence of other-cause mortality was 52% by the age of 85 yr. In a multivariate analysis, PSA, previous benign biopsy and hypertension were all independent predictors of PCSM.
One major limitation of this study is the fact that sextant biopsies have been used in the Rotterdam screening protocol. In contrast, contemporary practice of applying MR-fusion biopsies (target ± 12-core systematic) would further reduce the risk of undersampling, potentially affecting the study results and conclusions by decreasing PCSM. Furthermore, no data on PSA testing or opportunistic screening after discontinuing the screening study are available. Nevertheless, these findings underscore the need for customised well-grounded use of ongoing screening in elderly males in order to achieve a favourable harm-benefit ratio.
All-in-all, their risk of dying from PCa amounted to one tenth from the hazard of other-cause mortality in elderly men. Particularly for those stratified by low PSA levels at the last follow-up and the history of benign biopsy, truncation of further screening appears to be justified. For those at a higher risk for PCSM a diligent shared decision-making incorporating current performance status, comorbidities, and patient preferences, using risk calculators estimating individual life expectancy (PMID: 28199761), should be carried out. Further research in this area including more precise tools of geriatric prognostic assessment is warranted.