Lifetime Health and Economic Outcomes of Biparametric Magnetic Resonance Imaging as First-Line Screening for Prostate Cancer : A Decision Model Analysis

Abstract

Background: Contemporary prostate cancer (PCa) screening uses first-line prostate-specific antigen (PSA) testing, possibly followed by multiparametric magnetic resonance imaging (mpMRI) for men with elevated PSA levels. First-line biparametric MRI (bpMRI) screening has been proposed as an alternative.

Objective: To evaluate the comparative effectiveness and cost-effectiveness of first-line bpMRI versus PSA-based screening.

Design: Decision analysis using a microsimulation model.

Data sources: Surveillance, Epidemiology, and End Results database; randomized trials.

Target population: U.S. men aged 55 years with no prior screening or PCa diagnosis.

Time horizon: Lifetime.

Perspective: U.S. health care system.

Intervention: Biennial screening to age 69 years using first-line PSA testing (test-positive threshold, 4 µg/L) with or without second-line mpMRI or first-line bpMRI (test-positive threshold, PI-RADS [Prostate Imaging Reporting and Data System] 3 to 5 or 4 to 5), followed by biopsy guided by MRI or MRI plus transrectal ultrasonography.

Outcome measures: Screening tests, biopsies, diagnoses, overdiagnoses, treatments, PCa deaths, quality-adjusted and unadjusted life-years saved, and costs.

Results of base-case analysis: For 1000 men, first-line bpMRI versus first-line PSA testing prevented 2 to 3 PCa deaths and added 10 to 30 life-years (4 to 11 days per person) but increased the number of biopsies by 1506 to 4174 and the number of overdiagnoses by 38 to 124 depending on the biopsy imaging scheme. At conventional cost-effectiveness thresholds, first-line PSA testing with mpMRI followed by either biopsy approach for PI-RADS 4 to 5 produced the greatest net monetary benefits.

Results of sensitivity analysis: First-line PSA testing remained more cost-effective even if bpMRI was free, all men with low-risk PCa underwent surveillance, or screening was quadrennial.

Limitation: Performance of first-line bpMRI was based on second-line mpMRI data.

Conclusion: Decision analysis suggests that comparative effectiveness and cost-effectiveness of PCa screening are driven by false-positive results and overdiagnoses, favoring first-line PSA testing with mpMRI over first-line bpMRI.

Primary funding source: National Cancer Institute.