Introduction:

In the past, almost all men with an elevated prostate-specific antigen (PSA) level would be recommended a transrectal ultrasound-guided biopsy (TRUS-Bx). The introduction of pre-biopsy multiparametric (mp)MRI and biomarkers as reflex tests to PSA has been shown to increase the detection of clinically significant cancer and reduce overdiagnosis. We sought to compare the cost-effectiveness of alternative pathways for prostate cancer detection in a US-based healthcare system.

Methods:

We set as the base case scenario a middle-age man with an elevated PSA (3-10 ng/mL) to construct a clinical decision tree and compare 5 different diagnostic pathways: (1) standard TRUS-Bx for all (reference), (2) mpMRI followed by systematic/targeted biopsy if positive mpMRI (PI-RADS score 3-5), (3) Biomarker (here: exemplified by 4Kscore) followed by standard TRUS-Bx if positive test (≥7.5% risk score), and two combined strategies with either; (4) mpMRI first or (5) biomarker first. The mean cost (USD) per patient and clinically significant cancer (csPCa) detected was calculated for a hypothetical cohort of 1,000 men, with an underlying prevalence of clinically significant prostate cancer (csPCa; Gleason grade group >2) of 30%. The following assumptions (sensitivity/specificity/cost) were made based on up-to-date evidence and Medicare data: mpMRI (0.9/0.37/$964); 4KScore biomarker (0.9/0.45/$500); TRUS-Bx (0.63/0.9/$3000); systematic/target biopsy (0.85/0.9/$3000). Analyses were conducted using TreeAge Pro 2019, R2. TreeAge Software, Williamstown, MA; software available at http://www.treeage.com.

Results:

For every 1,000 men, all strategies were superior to standard TRUS-Bx (189 csPCa detected) as they yielded more csPCa detected with fewer biopsies, except a Biomarker only strategy (170 csPCa detected), as shown in Table 1. A mixed pathway with mpMRI as the initial reflex test and a biomarker performed in negative cases yielded the highest detection of csPCa (247 csPCa) however also at the highest cost $4,188 and with the most biopsies (880) (other than the reference test of standard TRUS-Bx for all), as shown in Figure 1. An mpMRI only strategy was comparable in cost and pick-up rate to a ‘mixed biomarker-first’ pathway (in which a biomarker was used as the initial screening test, with positive cases receiving an mpMRI to allow targeting of the biopsy if possible, while negative biomarker cases were put on surveillance). The ‘mpMRI-only’ strategy detected 216 cases at a cost of $3,493, while a ‘mixed biomarker-first’ detected 194 cases at a cost of $3,289. However, the biomarker-first strategy required 59 fewer biopsies (714 mpMRI vs 655 mixed strategy with biomarker first). 

Conclusion:

Our analysis suggests that a diagnostic pathway in which patients with an elevated PSA have a reflex biomarker test, and if test-positive, undergo an mpMRI, to allow biopsies to be targeted if possible, yields more csPCa detected, with fewer biopsies and at a lower cost than a standard TRUS-Bx pathway. This pathway would allow the initial screening test with both PSA and the reflex blood biomarker to be performed in primary care, yielding further cost savings. Strategies involving an mpMRI as the first screening test yielded marginally more csPCa cancer detection, but at greater cost and with more biopsies performed.

Funding: N/A