OBJECTIVE: Differentiating between cavernous hemangiomas of the liver and hepatic metastases on the basis of single-pass, contrast-enhanced CT is a significant and frequently encountered diagnostic challenge. Recognition of characteristic enhancement features of cavernous hemangiomas can aid in effectively distinguishing between these lesions. The purpose of this study was to determine sensitivity and specificity of dense, globular enhancement for differentiating cavernos hemangiomas and metastases during single-pass, contrast-enhanced CT. MATERIALS AND METHODS: CT appearance of 133 lesions in 91 patients with cavernous hemangiomas (44 patients) or metastases (47 patients) was retrospectively evaluated in a blinded review. CT examinations were performed with nonhelical technique following injection of 150 ml of contrast material. All patients with metastases had pathologic proof (n = 47). Patients with cavernous hemangiomas were clinically stable for at least 2 years after CT (n = 43) or had tissue proof (n = 1). All lesions were evaluated based on the following criteria: (1) Type of enhancement: globular, linear, diffuse and homogeneous, or diffuse and heterogeneous. (Globular enhancement was considered to be present when enhancing nodules less than 1 cm in diameter were seen within lesions.) (2) Continuity of enhancing tissue: continuous or noncontinuous. (Uninterrupted collections of contrast material within at least 50% of a lesion were considered continuous. Multiple, separate collections of contrast material were considered noncontinuous.) (3) Degree of enhancement: hypo-, iso-, or hyperdense relative to the aorta. (4) Distribution of enhancement: peripheral, central, or mixed. RESULTS: Seventy-six percent of cavernous hemangiomas had globular enhancement, compared to 10% of metastases (p < .001). Seventy-two percent of cavernous hemangiomas had enhancement isodense with the aorta, and 96% of metastases were hypodense (p < .001). Sixty-seven percent of cavernous hemangiomas had peripheral enhancement, compared to 38% of metastases (p < .001). The combined finding of globular, isodense enhancement was seen in 67% of cavernous hemangiomas and none of the metastases. Only 10% of cavernous hemangiomas had nonglobular, hypodense enhancement, compared with 90% of metastases. Combining all criteria, reviewers correctly classified 122 (92%) of the lesions. Presence of globular enhancement, isodense with the aorta, was 67% sensitive and 100% specific in differentiating cavernous hemangiomas from hepatic metastases. CONCLUSION: In most cases, differentiation of cavernous hemangiomas from hepatic metastases can confidently be made with single-pass, contrast-enhanced CT. Globular enhancement, isodense with the aorta, is 67% sensitive and 100% specific in differentiating cavernous hemangiomas and hepatic metastases.
OBJECTIVE: Differentiating between cavernous hemangiomas of the liver and hepatic metastases on the basis of single-pass, contrast-enhanced CT is a significant and frequently encountered diagnostic challenge. Recognition of characteristic enhancement features of cavernous hemangiomas can aid in effectively distinguishing between these lesions. The purpose of this study was to determine sensitivity and specificity of dense, globular enhancement for differentiating cavernos hemangiomas and metastases during single-pass, contrast-enhanced CT. MATERIALS AND METHODS: CT appearance of 133 lesions in 91 patients with cavernous hemangiomas (44 patients) or metastases (47 patients) was retrospectively evaluated in a blinded review. CT examinations were performed with nonhelical technique following injection of 150 ml of contrast material. All patients with metastases had pathologic proof (n = 47). Patients with cavernous hemangiomas were clinically stable for at least 2 years after CT (n = 43) or had tissue proof (n = 1). All lesions were evaluated based on the following criteria: (1) Type of enhancement: globular, linear, diffuse and homogeneous, or diffuse and heterogeneous. (Globular enhancement was considered to be present when enhancing nodules less than 1 cm in diameter were seen within lesions.) (2) Continuity of enhancing tissue: continuous or noncontinuous. (Uninterrupted collections of contrast material within at least 50% of a lesion were considered continuous. Multiple, separate collections of contrast material were considered noncontinuous.) (3) Degree of enhancement: hypo-, iso-, or hyperdense relative to the aorta. (4) Distribution of enhancement: peripheral, central, or mixed. RESULTS: Seventy-six percent of cavernous hemangiomas had globular enhancement, compared to 10% of metastases (p < .001). Seventy-two percent of cavernous hemangiomas had enhancement isodense with the aorta, and 96% of metastases were hypodense (p < .001). Sixty-seven percent of cavernous hemangiomas had peripheral enhancement, compared to 38% of metastases (p < .001). The combined finding of globular, isodense enhancement was seen in 67% of cavernous hemangiomas and none of the metastases. Only 10% of cavernous hemangiomas had nonglobular, hypodense enhancement, compared with 90% of metastases. Combining all criteria, reviewers correctly classified 122 (92%) of the lesions. Presence of globular enhancement, isodense with the aorta, was 67% sensitive and 100% specific in differentiating cavernous hemangiomas from hepatic metastases. CONCLUSION: In most cases, differentiation of cavernous hemangiomas from hepatic metastases can confidently be made with single-pass, contrast-enhanced CT. Globular enhancement, isodense with the aorta, is 67% sensitive and 100% specific in differentiating cavernous hemangiomas and hepatic metastases.