Addition of gadolinium chelates to heavily T2-weighted MR imaging: Limitedrole in differentiating hepatic hemangiomas from metastases

OBJECTIVE. The purpose of this study was to determine whether the addition of gadolinium-enhanced imaging to heavily T2-weighted MR imaging of the liv er is valuable in differentiating hemangiomas from metastases. The T2 relax ation time was also included in our analysis. SUBJECTS AND METHODS. Fifty-one patients with 52 proven liver lesions (24 h emangiomas and 28 metastases) larger than 1 cm underwent MR imaging at 1.5 T with T2-weighted spin-echo (TR/TE, 3000/80, 160) and gadolinium chelate-e nhanced dynamic T1-weighted gradient-recalled echo (80/2.6, 80) pulse seque nces. Images were reviewed by observers who were unaware of the patients' c linical history; first, only T2-weighted images were reviewed and then T2-w eighted plus dynamic images were reviewed together. The T2 relaxation times were calculated:for each lesion. Diagnostic accuracy by each method was co mpared using receiver operating characteristic analysis. RESULTS. Mean T2 relaxation times were 76 +/- 26 msec for metastases and 13 3 +/- 25 msec for hemangiomas. The addition of dynamic scanning to the T2-w eighted sequence made a statistically significant difference for only one o bserver (p = 0.03). However, it did not make a statistically significant co ntribution for either observer when compared with the T2 relaxation time. A lthough addition of the dynamic images resulted in correct diagnosis of six lesions, three lesions were misdiagnosed after having been correctly chara cterized on the T2-weighted images alone. CONCLUSION. When optimized T2-weighted images are obtained and the T2 relax ation time is calculated, routine use of gadolinium enhancement for differe ntiation of hemangiomas from metastases is unnecessary although dynamic sca nning is valuable in selected cases.