[Gamna-Gandy bodies in cirrhosis: a meaningless finding?].

PURPOSE: To determine the presence of Gamna-Gandy bodies (GGB) on MRI in patients with liver cirrhosis.

PATIENTS AND METHODS: A total of 117 consecutive patients with cirrhosis followed-up by MRI were retrospectively reviewed. Two groups were defined: the first group included patients with GGB on MRI (n=15), the other group included patients without GGB (n=102). Both characteristics of groups were reviewed using standard cirrhosis criteria evaluation: sex, age, etiology of cirrhosis, Child-Pugh score, presence of esophageal varices, splenomegaly, ascitis, recanalization of the periumbilical veins, and presence of hepatic encephalopathy. Fisher's exact test and student t-test were used to compare both groups.

RESULTS: GGB were more frequently observed in patients with splenomegaly (P=0.035). Hemochromatosis was the only etiology for cirrhosis statistically correlated to the presence of GGB (P=0.006) in our series. No other statistically significant association was noted between GGB and other characteristics of our cirrhotic patients.

CONCLUSION: Easily identified on all MRI pulse sequences, GGB do not correlate with the severity of cirrhosis. However, they are strongly correlated with the presence of splenomegaly and may be the result of segmental splenic hypertension. They are frequent in patients with hemochromatosis.