Introduction:

Prostate biparametric magnetic resonance imaging (bpMRI) has similar cancer detection rates to multiparametric MRI (mpMRI) without necessitating venipunctures or gadolinium contrast for dynamic contrast-enhanced imaging (DCE). DCE is utilized in mpMRI to differentiate between PIRADS-3 and PIRADS-4 lesions in the peripheral zone (PZ). However, if PZ PIRADS-3 lesions are biopsied similarly to PIRADS-4 or -5 lesions, characteristics of DCE would be inconsequential and bpMRI alone could guide management. Additionally, some patients are biopsy-keen and would favor biopsy for any PIRADS-3 lesion, further limiting the utility of mpMRI over bpMRI. Shared decision (SD) making between the patient and their urologist regarding biopsy preferences and physician’s suspicion of clinically significant cancer (CSC) may help identify those patients for which bpMRI is sufficient to guide management. This approach could save patients the discomfort of venipuncture, risk of contrast reactions, and cut down on health-care associated costs. We investigated the effect of bpMRI with a SD pathway on patient experience, cost, and MRI throughput.

Methods:

In an IRB approved retrospective study, we reviewed our institutional database of 287 treatment-naïve men undergoing mpMRI. We used decision trees to create three possible MRI work up scenarios. Scenario 1 (S1) was the current standard of care, mpMRI. Scenario 2 (S2) was initial bpMRI with decision to observe PIRADS-3, but biopsy PIRADS-4 and higher. Thus all patients with PZ PIRADS-3 were called back for re-imaging with DCE. Scenario 3 (S3) was initial bpMRI with SD protocol: of the patients with PZ PIRADS-3, only the subset determined to directly benefit from DCE were called back for re-imaging. We compared outcomes of system cost, patient cost, MRI throughput, and patient experience for each scenario. System cost was calculated as MRI and radiologist fee, obtained from Centers for Medicare and Medicaid Services. Patient cost was estimated as a half-day productivity loss, obtained from the US Department of Labor Statistics. MRI throughout was calculated using acquisition and room times from our facility. Patient unpleasant experiences were estimated by utilizing rates of adverse events for venipuncture and contrast administration described in the literature and rates of callbacks when initial bpMRI was insufficient in our cohort.

Results:

Of 287 men, 50 (17.4%) had PZ PIRADS-3 lesions: DCE was negative for 36 (12 not biopsied) and positive with upgrade to PZ PIRADS-4 for 14 (11 biopsied). Combined, 23 of 287 men (8.0%) could have had management altered by DCE result. In S2, all patients with PZ PIRADS-3 (50) were called back to complete MRI work up with DCE. In S3, using the SD model, of the patients with PZ PIRADS-3, only the subset determined to directly benefit from DCE (23) were called back. When comparing outcomes for S2 and S3 with those for S1 (standard of care), system cost was reduced by both: S2 [-$32,646.75 (20%)] and S3 [-$36,366 (22.4%)]. Per patient cost was increased by both: S2 [+$18 (17%)] and S3 [+$8 (8%)]. Time to complete MRI workup was reduced by both: S2 [-1.3 days (8.0%)] and S3 [-4.2 days (25.9%)]. Number of unpleasant patient experiences was increased by S2 [+59.1 (13%)], but decreased by S3 [-25.2 (48.2%)]. Using S3 as a control due to its lowest number of unpleasant experiences, we calculated number needed to harm (NNH) for S1 to be 5.9 patients and S2 to be 58.8 patients. 

Conclusion:

Our results suggest that for treatment-naïve men, using bpMRI as the initial prostate imaging protocol combined with a SD on biopsy preferences and physician’s suspicion of CSC reduces patient unpleasant experiences, minimizes system cost, and maximizes MRI throughput, while having only a minor increase on per patient cost.

Funding: N/A