MP17-05: Association of Urology Practice Implementation of Multiparametric MRI and Genomic Testing on Treatment Rates in Men With Newly diagnosed Prostate Cancer
Introduction: Prostate biomarkers, including multiparametric MRI (mpMRI) and/or tissue-based genomics, are increasingly used to inform decision-making by providing additional information about prostate cancer biology. Biomarkers applied uniformly, thereby ignoring significant competing physiologic risks (i.e., a high risk of noncancer mortality), may lead to treatment in a population with limited clinical benefit. We assessed the association of practice use of prostate biomarkers and treatment rates in men with newly diagnosed prostate cancer by noncancer mortality risk. Methods: We used a 20% national Medicare sample and performed a retrospective cohort study of men with newly diagnosed prostate cancer in 2015 through 2018. Practice-level use of mpMRI and genomics was characterized using the percentage of a practice’s patient panel using either biomarker and sorted into 3 groups—no use, use below the median (low use), use above the median (high use). We fit 3 separate multilevel models to evaluate the relationship between practice-level use of a biomarker (mpMRI, genomics, both) and use of localized treatment by noncancer mortality risk at 10 years. Results: Between 2015 and 2018, 1576 (61.5%) and 934 (36.4%) of 2,564 urology practices used mpMRI or genomics in at least 1 man with cancer. On multilevel regression analysis, high use of mpMRI by a practice was associated with a 6.0% (p < 0.05) higher treatment rate in the group with very high noncancer mortality risk, compared to no use. Adoption of genomics only impacted the moderate risk group, as high use of genomics was associated with a 4.5% (p < 0.05) reduction in the rate of treatment, compared to no testing. When evaluating both biomarkers (Figure), use of the two biomarkers was associated with significant increase in treatment rates in the group with the highest noncancer mortality risk (7.1% difference, p<0.05), compared with no testing. Conclusions: Compared to practices that do not use biomarker (genomics, mpMRI) testing after incident prostate cancer, practices that increasingly use biomarkers are more likely to treat patients with higher 10-year noncancer mortality risk. This association is mainly dominated by practice-level adoption of mpMRI. SOURCE OF Funding: This study is supported by funding from National Cancer Institute Advanced Training in Urologic Oncology T32 Grant No T32CA180984 (Faraj) and by funding from the National Cancer Institute RO1CA269367 (Hollenbeck, Shahinian).
