Prediction rule for characterization of hepatic lesions revealed on MR imaging: Estimation of malignancy

OBJECTIVE. Our aims were to establish factors that are most predictive of h epatic lesion malignancy and to formulate a prediction rule. MATERIALS AND METHODS. A cross-sectional study of 227 abdominal MR imag ing examinations revealed 85 lesions in 67 patients (29 men, 38 women: age ran ge, 29-78 years: mean age. 51.4 years) who were bring examined for primary malignancy (n = 42) or unknown lesion characterization (n = 25). All were r eferred for MR imaging after CT or sonography. Patient demographics (age, s ex. history of malignancy), lesion size and morphology, quantitative T2 cal culation. and pattern of enhancement on gadopentetate dimeglumine administr ation were evaluated for predictive ability. RESULTS. Thirty-two liver lesions were malignant (eight colon cancer. five breast cancer, four cervical cancer. three renal cancer, three lung cancer, and nine miscellaneous cancers), 53 were benign (37 hemangiomas, 15 cysts. and one focal nodular hyperplasia). Calculated T2 relaxation times (mean /- standard deviation [SD]) ware as follows. malignant tumors (91.72 +/- 21 .9 msec). hemangiomas (136.1 +/- 26.3 msec), cysts (284.1 +/- 38.2 msec) (p < 0.001). Logistic regression analysis indicated that lesion size and sex and age of patient were not significant independent predictors (p > 0.05). However. the combination of a history of malignancy, T2 value, and gadopent etate dimeglumine-enhancement pattern allowed generation of a prediction ru le with an area under the receiver operating characteristic curve of 0.95. The patient's weight, lesion morphology. and cell type of the primary malig nancy did not provide additional predictive information (p > 0.2). CONCLUSION. Wr recommend using the combination of T2 quantification and pat ient history of malignancy before deciding to administer gadopentetate dime glumine for optimal lesion characterization, especially for equivocal lesio ns with T2 values between 90 and 130 msec. These factors allowed the constr uction of a prediction rule for lesion characterization.