Background: Hepatic granuloma (HG) is a well defined histopathological finding with an heterogenous clinical presentation. Diagnosis of a specific clinical entity is not possible every time. Descriptive studies may shed light on the various etiologies also common and distinctive findings among these patients.
Methods: We reviewed the results of the liver biopsies of 592 patients. Characteristics of the patients with HG were extracted from the hospital charts. Laboratory studies included biochemical tests, hepatitis C virus (HCV) antibody, Brucella agglutination tests, tuberculin skin test. According to the diagnostic clues further tests (thoracic computed tomography (CT), ultrasonography, organ biopsy in addition to liver, antimitochondrial antibody, hepatitis B surface (HBs) antigen, venereal disease research laboratory (VDRL)) were performed.
Results: HG was found in 13 of the 592 patients (2.2%). Primary biliary cirrhosis (three cases) was the most frequent cause followed by sarcoidosis, miliary tuberculosis and BCGitis (Bacillus Calmette Guerin) (two cases each). Two patients with HG could not be diagnosed. Only three patients had remarkable physical examination findings. Alkaline phosphatase and gamma-glutamyl transpeptidase were the most frequently elevated enzymes. Abdominal ultrasonography provided no specific diagnostic clue in any patient. Localization of the HGs was portal in 6 patients, parenchymal in 5 patients and both portal and parenchymal in 2 patients. Three exitus were due to BCGitis, miliary tuberculosis and fungal infection.
Conclusions: Tuberculosis is still among the most common etiologic factors. BCGitis has a fulminant rather than an indolent course. Abdominal ultrasonography could be used to rule out obstructive jaundice rather than to reach a specific diagnosis. Involvement of portal area by HG in most of the cases might cause obstruction of the biliary canaliculi and elevation of the cholestatic enzymes. Follow up of the difficult cases may be the best approach since the presence of HG was not proved as a bad prognostic factor for any disease.