Lifetime Health and Economic Outcomes of Biparametric MRI as Front-line Screening for Prostate Cancer Early Detection: A Decision Model Analysis
Ra'ad Al-Faouri, MD MMSc1, Boshen Jiao, PHD2, Vidit Sharma, MD3, Sumed Kaul, MS1, Aaron Fleishman, MPH1, Kevin Wymer, MD3, Stephen Boorjian, MD3, Aria Olumi, MD1, Ruth Etzioni, PHD4, Roman Gulati, MS4, Boris Gershman, MD1.
1Beth Israel Deaconess Medical Center, Boston, MA, USA, 2Harvard School of Public Health, Boston, MA, USA, 3Mayo Clinic, Rochester, MN, USA, 4Fred Hutchinson Cancer Research Center, Seattle, WA, USA.
Background: Screening strategies for most cancers employ imaging-based evaluation. However, contemporary screening paradigms for prostate cancer (PCa) are based on front-line biomarker testing with prostate-specific antigen (PSA), which may be followed by reflex multiparametric magnetic resonance imaging (mpMRI) for men with elevated PSA values. Early evidence has suggested that front-line screening with biparametric MRI (bpMRI) may be a viable imaging-based screening strategy. Herein, therefore we evaluated the cost-effectiveness of front-line bpMRI screening for prostate cancer compared to PSA-based approaches with and without mpMRI.
Methods: We conducted a decision analysis of PCa screening strategies by extending an existing microsimulation model of disease natural history, diagnosis, treatment, and survival. We simulated a contemporary population of US men aged 55 years with no prior screening or PCa diagnosis followed until age 100. Interventions included biennial screening up to age 69 using one of 9 strategies involving front-line PSA (with or without reflex mpMR) or front-line bpMRI (followed by MR-guided biopsy [MRGB] or MRGB plus systematic transrectal ultrasound-guided biopsy [TRUS]).
Results: Screening with front-line bpMRI incurs substantially higher numbers of biopsies and prostate cancer diagnoses (Figure 1). Although this strategy leads to fewer PCa deaths compared to front-line PSA-based strategies, it is associated with high rates of overdiagnosis without commensurate increases in lives saved or years of life saved. There was only moderate variation projected when using PI-RADS 3-5 instead of PI-RADS 4-5 for biopsy referral or when using MRGB instead of MRGB+TRUS. None of the strategies using front-line bpMRI were cost-effective compared to a strategy of front-line PSA. At the willingness-to-pay thresholds of $100,000 and $150,000 per quality-adjusted years of life saved, front-line PSA with mpMRI as a reflex test followed by MRGB+TRUS (for PI-RADS 3-5 lesions) produced the greatest NMB.
Conclusions: Compared to screening strategies based on front-line PSA, strategies based on front-line bpMRI are not cost-effective.
Figure 1: Projected clinical outcomes.bpMRI = biparametric magnetic resonance imaging; mpMRI = multiparametric magnetic resonance imaging; MRGB = MR-guided biopsy; NMB = net monetary benefit; PSA = prostate specific antigen; TRUS = transrectal ultrasound-guided biopsy; WTP = willingness to pay 
Back to 2023 Abstracts