Do patients who undergo multiparametric MRI for prostate cancer benefit from additional staging imaging? Results from a statewide collaborative.

ABSTRACT

OBJECTIVES: Prostate cancer (CaP) staging traditionally includes computed tomography (CT) and technetium-99m bone scintigraphy (BS) for assessment of lymph node (LN) and bone metastases, respectively. In recent years, multiparametric magnetic resonance imaging (mpMRI) has been used in diagnostic assessment of CaP. We sought to compare the accuracy of mpMRI to CT and BS for pretreatment staging. MATERIALS AND METHODS: Using the Michigan Urological Surgery Improvement Collaborative registry, we identified men undergoing pretreatment mpMRI in addition to CT and/or BS in 2012 to 2018. Imaging reports were classified as positive, negative, or equivocal for detection of LN and bone metastases. A best value comparator (BVC) was used to adjudicate metastatic status in the absence of pathologic data. mpMRI accuracy was calculated using pessimistic (equivocal=positive) and optimistic (equivocal = negative) interpretations. We compared the diagnostic performance of mpMRI, CT, and BS in detecting metastases. RESULTS: In total, 364 men underwent CT and mpMRI, and 646 underwent BS and mpMRI. Based on the BVC, 52 men (14%) harbored LN metastases and 38 (5.9%) harbored bone metastases. Sensitivity of mpMRI for LN metastases was significantly higher than CT (65-73% vs 38%, P < 0.005), and specificity of mpMRI and CT were 97% to 99% and 99% (P = 0.2-0.4), respectively. For bone metastases, BS sensitivity was 68% as compared to 42% to 71% (P = 0.02-0.83) for mpMRI. Specificity for bone metastases was 95% to 99% across all modalities. CONCLUSIONS: Using statewide data, mpMRI appears superior to CT and comparable to BS for detection of LN and bone metastases, respectively. Pretreatment mpMRI may obviate the need for additional staging imaging.