RESUMO
PURPOSE: To analyze the variable MRI features and clinical significance of intracranial cavernous realformation. MATERIALS AND METHODS: Forty patients(mean age 35.4) with cavernous malformation were evaluated by MRI. Eleven patients were surgically confirmed. Cavernous malformations were divided into four categories on the basis of the MR imaging characteristics, especially on T2-weighted image. Type I lesion was defined as an extralesional subacute hemorrhage outside the low signal rim, type II as an intralesional hemorrhage surrounded by low signal rim, type III lesion as an intralesional thrombosis with variegated central core surrounded by low signal rim, and type IV lesion as a focal old hemorrhagic core with small low signal intensity. Type IV was further divided into IVa and IVb, whether the lesion has small iso- or hypersignal central core (IVa) or not (IVb). Follow-up MRI was evaluated in 12 patients who were managed conservatively. Follow-up intervals ranged from 2 weeks to 29 months (mean 6months). RESULTS: Total 80 lesions were detected in 40 patients. Multiple lesions were noted in 10 patients. The topography of the cavernous malformations was supratentorial in 75% and infratentorial in 23%. There were 10 lesions in type I, 15 in type 11, 21 in type III, 14 in type IVa, and 20 in type IVb. Type I lesions mainly showed mass effect and edema. Type III lesions showed minimal contrast enhancements in 7 lesions on delayed images. Type II lesions showed the characteristics of both type I and type III lesions. On follow up images, decrease in size in 5, change of type in 7, rebleeding in 2 and no change in 12 lesions were demonstrated. Hemorrhage, edema and mass effect were combined in the cases of rebleeding. On follow-up study, the estimated risk of bleeding was 32.3%/person-year and 13.7%/lesion-year. CONCLUSION: Cavernous realformations show as variable appearance, on MR imaging suggesting variable stages of evolution. The MR morphologic classification and evaluation of secondary findings are helpful to predict natural course and possibility of rebleeding of the lesion.