Characterization of hepatic lesions by perfusion-weighted MR imaging with an echoplanar sequence.

Abstract

The purpose of this study was to examine the usefulness of perfusion-weighted MR imaging with a single-shot gradient echoplanar sequence in characterizing hepatic tumors. Perfusion-weighted imaging was performed in 61 patients with 91 confirmed hepatic tumors (14 hemangiomas, 19 metastases, and 58 hepatocellular carcinomas). The perfusion-weighted imaging was started at the time of administration of 0.1 mmol/kg of gadolinium, and images were continuously obtained every 2 sec for 88 sec. Time-intensity curves for all the tumors were created for quantitative analysis. Patterns of enhancement were also evaluated. Changes in signal intensity that occurred throughout examination differed in three types of tumor. Transient signal intensity decreases in the perfusion phase significantly differed in three types of tumors (46% in hepatocellular carcinoma, 48% in hemangioma, and 15% in metastasis, p < .05 for hepatocellular carcinoma versus metastasis and for hemangioma versus metastasis). Signal intensity recovered rapidly for hepatocellular carcinoma and metastasis, whereas recovery was slower for hemangioma. Final signal intensity recovery was 94% in hepatocellular carcinoma, 91% in metastasis, and 54% in hemangioma compared with their initial signal intensities. (p < .05 for hepatocellular carcinoma versus hemangioma and for hemangioma versus metastasis.) The enhancement patterns also differed in three types of tumor. Perfusion-weighted imaging with a gradient echoplanar sequence provides real-time mapping of many points along the enhancement profile curves because of its excellent temporal resolution. Therefore, it accurately characterizes hepatic tumors based on their different negative-enhancement patterns.