Angiography-Based C-Arm CT for the Assessment of Extrahepatic Shunting before Radioembolization

Abstract

To retrospectively assess the accuracy of angiography-based C-arm CT for the detection of extrahepatic shunting before SIRT. 30 patients (mean age: 64+/-12 years) with hypervascularized hepatic tumors underwent hepatic angiography, coil embolization of gastrointestinal collaterals and 99mTc-macroaggregated albumin (MAA) SPECT/CT before SIRT. Before MAA injection via a microcatheter from the intended treatment position, an angiography and angiography-based C-arm CT (XperCT, Philips Healthcare) were acquired. Angiographies and XperCT were performed from 48 microcatheter positions followed by MAA injections and MAA-SPECT/CT. MAA-SPECT/CT served as the reference standard for determining the accuracy of hepatic arteriography and C-arm CT for the detection of extrahepatic shunting. MAA-SPECT/CT revealed extrahepatic shunting in 5 patients (17%). Hepatic arteriography yielded a true negative in 22 (73%), a false negative in 5 (17%), and an unclear result in 3 patients (10%). C-arm CT yielded a true positive in 3 (10%), true negative in 24 (80%), false positive in 1 (3%), and false negative in 2 patients (7%). The specificity and the NPV of hepatic arteriography for the detection of extrahepatic shunting were 88% and 81%, respectively. For C-arm CT the sensitivity, specificity, PPV, NPV, and accuracy for the detection of extrahepatic shunting were 60%, 96%, 75%, 92%, and 90%, respectively. C-arm CT offers additional information to angiography when assessing SIRT patients for extrahepatic shunting. More accurate detection of extrahepatic shunting may optimize the workflow in SIRT preparations by avoiding unnecessary repeat angiographies.