P2Y Reaction Units Threshold for Implementing Modified Antiplatelet Preparation in Coil Embolization of Unruptured Aneurysms: A Prospective Validation Study.

Purpose To identify and apply an optimized P2Y reaction units (PRU) threshold for implementing modified antiplatelet preparation to prevent thromboembolic events in patients nonresponsive to clopidogrel (clopidogrel nonresponders) undergoing coil embolization of unruptured aneurysms and to evaluate the clinical validity. Materials and Methods The optimal PRU threshold for prediction of thromboembolic events was determined with the Youden index in post hoc analysis of a previous, prospectively enrolled cohort of 165 patients in whom the antiplatelet regimen was not modified. This optimal PRU threshold was used to define clopidogrel nonresponders in a prospective validation study of 244 patients. Standard preparation (aspirin, clopidogrel) was maintained for 126 patients responsive to clopidogrel (clopidogrel responders, 51.6%), and modified preparation (aspirin, prasugrel) was prescribed prior to embolization for 118 clopidogrel nonresponders (48.4%). Fifty-seven clopidogrel nonresponders from the previous cohort who did not receive the modified preparation were included as a historical control group. Thromboembolic and bleeding events were compared among groups by using logistic regression analysis. Results Post hoc analysis from the previous cohort yielded PRU of at least 220 as the optimal threshold for modified preparation selection. The thromboembolic event rate of the clopidogrel responders (one of 126 [0.8%]) was lower than that of the historical control group that received standard preparation (seven of 57 patients [12.3%]; adjusted risk difference [RD], -10.1%; 95% confidence interval [CI]: -18.5, -1.7; P = .015) and was similar to that of clopidogrel nonresponders who received modified preparation (one of 118 [0.8%]; adjusted RD, -0.5%; 95% CI: -3.1, 2.1; P = .001 for noninferiority; P = .699 for superiority). Bleeding event rates did not differ among groups (four of 126 clopidogrel responders [3.2%] vs four of 57 clopidogrel nonresponders that received standard preparation [7.0%] [adjusted RD, -4.5%; 95% CI: -11.1, 3.4; P = .228] vs five of 118 clopidogrel nonresponders that received modified preparation [4.2%] [adjusted RD, -0.6%; 95% CI: -5.8, 4.2; P = .813]). Conclusion Patients undergoing coil embolization of unruptured aneurysms, regardless of clopidogrel responsiveness, had low thromboembolic risk when using PRU of at least 220 as the threshold for implementing modified antiplatelet preparation with prasugrel. RSNA, 2016 Online supplemental material is available for this article.