ABSTRACT
IMPORTANCE: Prostate-specific antigen (PSA) screening for prostate cancer has resulted in a slight reduction in prostate cancer mortality but also a concomitant overdiagnosis of low-risk tumors. Prostate-specific antigen levels are affected by use of cholesterol-lowering statin drugs, but the association of statin use with PSA screening performance is unknown. OBJECTIVE: To investigate whether statin use was associated with outcomes of a randomized PSA-based prostate cancer screening intervention. DESIGN, SETTING, AND PARTICIPANTS: This post hoc subgroup analysis of a cohort from a population-based randomized clinical trial used data from the population-based Finnish Randomized Study of Prostate Cancer Screening, which randomized men to PSA screening or routine care from March 1, 1996, to December 31, 1999, with follow-up continuing until December 31, 2015. The population included all men aged 55 to 67 years at baseline and residing in the Tampere or Helsinki districts of Finland. Information on statin purchases from 1996 to 2009 was obtained from a national prescription registry. Eligible men were identified from the population registry of Finland. Prevalent prostate cancer cases at baseline were excluded. Data were analyzed from January 1, 2019 to March 31, 2021. INTERVENTIONS: Three invitations for PSA screening at 4-year intervals from 1996 to 2007 vs routine care. MAIN OUTCOMES AND MEASURES: Risk for prostate cancer overall, high-risk disease, and prostate cancer mortality in the screening group vs the control group as an intention-to-treat analysis. The analysis was stratified by statin use. RESULTS: The study comprised 78â¯606 men (median age, 59 years [range, 55-67 years]) with statin purchase data available. Although PSA screening was associated with increased prostate cancer incidence among statin nonusers (screening vs control, 11.2 vs 8.6 per 1000 person-years); rate ratio [RR], 1.31; 95% CI, 1.24-1.38), no similar increase in incidence was observed among statin users (6.9 vs 5.9 per 1000 person-years; RR, 1.02; 95% CI, 0.95-1.10; P < .001 for interaction). Incidence of low-risk (Gleason score 6) and localized tumors was lower among statin users, whereas detection of tumors with a Gleason score of 8 to 10 was similar. Screening was associated with a lower incidence of metastatic tumors regardless of statin use. CONCLUSION AND RELEVANCE: In this post hoc subgroup analysis of a cohort from a population-based randomized clinical trial, PSA screening among statin users was associated with a decreased incidence of advanced prostate cancer that was similar among statin nonusers, but with less increase in detection of low-grade localized tumors in statin users than in nonusers. These findings suggest that statin use does not materially compromise benefits of PSA-based screening.
Subject(s)
Hydroxymethylglutaryl-CoA Reductase Inhibitors , Prostatic Neoplasms , Aged , Early Detection of Cancer/methods , Humans , Hydroxymethylglutaryl-CoA Reductase Inhibitors/adverse effects , Male , Mass Screening , Middle Aged , Prostate-Specific Antigen , Prostatic Neoplasms/diagnosis , Prostatic Neoplasms/drug therapy , Prostatic Neoplasms/epidemiologyABSTRACT
BACKGROUND: The Finnish Randomized Study of Screening for Prostate Cancer (FinRSPC), the largest component of the European Randomized Study of Screening for Prostate Cancer (ERSPC), showed a smaller, nonsignificant reduction in prostate cancer-specific mortality by systematic prostate-specific antigen (PSA)-based screening compared with the overall ERSPC results. Nonsteroidal anti-inflammatory drugs (NSAIDs) reduce inflammation and also PSA elevations due to intraprostatic inflammation. OBJECTIVE: To explore whether NSAID usage modifies the effects of PSA-based screening on prostate cancer incidence and mortality. DESIGN, SETTING, AND PARTICIPANTS: A cohort of 78 165 men from the FinRSPC were linked to a comprehensive national prescription database to obtain information on NSAID reimbursements prior to screening. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: Prostate cancer risk and mortality were compared between the FinRSPC screening arm and the control arm among NSAID users and nonusers using an age-adjusted Cox regression model. RESULTS AND LIMITATIONS: Screening increased the detection of Gleason 6 (hazard ratio [HR] 1.59, 95% confidence interval [CI] 1.47-1.72 and HR 1.39, 95% CI 1.26-1.54) and localized prostate tumors (HR 1.25, 95% CI 1.18-1.32 and HR 1.11, 95% CI 1.03-1.20) more among baseline NSAID nonusers than among users, respectively (p for interaction <0.04 for both). This difference was observed in all three screening rounds. Detection of metastatic prostate cancer was similar in both NSAID users and nonusers. Screening decreased prostate cancer mortality among men using NSAIDs at FinRSPC randomization (HR 0.66, 95% CI 0.49-0.90) but not among nonusers (HR 0.95, 95% CI 0.81-1.12); p for interaction=0.04. CONCLUSIONS: Screening detected fewer well-differentiated localized tumors among NSAID users than among nonusers. This suggests that PSA screening may cause less overdiagnosis within this subgroup, whereas mortality benefit may be greater among NSAID users. PATIENT SUMMARY: Prostate cancer screening causes less overdiagnosis of well-differentiated localized prostate tumors among men who use nonsteroidal anti-inflammatory drugs.
