Assessing the value of hepatic arterial phase CT imaging in patients with breast cancer.

Abstract

5 Background: Hypervascular liver metastases, classically seen in melanoma and other cancers, may be best seen on a hepatic arterial phase (HAP) CT and possibly missed on the standard portal venous phase (PVP). Breast cancer sometimes produces hypervascular liver metastases, but the incidence of this is not well established. Hence, some centers perform biphasic liver CT (HAP + PVP) as standard protocol in breast cancer patients, while others do not. Our center does not include HAP in this patient population; although, it is sometimes protocolled with HAP as per radiologist preference. We sought to determine if the detection for presence or absence of liver metastases is significantly affected by the addition of HAP in breast cancer. Methods: This retrospective study was conducted using a custom search on our Picture Archiving and Communication System for all female patients who received a biphasic liver CT (HAP + PVP) from Mar 2013 - Mar 2015. Inclusion criteria included known breast cancer, liver metastases described on CT report, and follow-up imaging to confirm the finding. A total of 25 CT studies met inclusion criteria. Results: 14/25 (56%) studies demonstrated typical non-hypervascular hepatic metastases. 11/25 (44%) demonstrated hypervascular hepatic lesions. Of these latter, 4/25 (16%) represented hypervascular metastases, while 7/25 (28%) pertained to indeterminate arterial-enhancing lesions called suspicious by the reporting radiologist. Further imaging including MRI, ultrasound, and follow-up CT were recommended in these indeterminate suspicious lesions, all of which confirmed benign entities. Of the 18 studies with true hepatic metastases, the presence of metastatic disease was detectable on the PVP in 18/18 (100%, P< .0001) regardless of the metastases’ hypervascular status. Conclusions: We found hypervascular metastases to be present in 22% (4/18; CI95 [3-41%]) of breast cancer patients with true liver metastases. However, the detection for the presence of hepatic metastases was not improved with the addition of HAP. The HAP instead led to detection of more indeterminate but ultimately benign entities of no clinical significance, thus resulting in unnecessary imaging and arguably greater patient anxiety.