Purpose: To report our experience with helical CT evaluation of transient hepatic attenuation differences (THAD), and in particular of those not associated with focal lesions, in an attempt to provide an aetiopathogenetic picture that accounts for the morphology, evolution and density of THAD.
Materials and methods: Between January 1998 and January 2001 we observed THAD in 130/988 biphasic helical CT liver examinations performed in the arterial and portal dominant phase. THAD were associated to focal hepatic lesions in 87 patients; in 43 patients there was no such association. This second group of patients, composed of 23 males and 20 females ranging in age from 17 to 80 years (average = 58.8), was enrolled in the study. THAD were associated to: Budd-Chiari syndrome (9), portal venous thrombosis (10), liver cirrhosis (7), acute inflammation of an adjacent organ (4), dilatation of the entire biliary tree (3), hepatic stasis caused by heart failure (2) and constrictive pericarditis (1), fine-needle percutaneous biopsy (2), arterioportal shunting (2), parenchymal compression by fractured ribs (2) and by a strengthened phrenic pillar (1). THAD were evaluated according to extension, morphology and density. For each case at least 10 density measurements were performed by sampling regions of interest on the parenchyma with THAD and on the contralateral parenchyma. The results (mean and standard deviation) were compared to those relative to 30 healthy patients. 22/43 patients were followed up for 6#150;24 months by at least one US and helical CT examination. During CT, the direct appreciation of vascular thrombus during the portal dominant phase was also considered.
Results: We detected 18 localised and 25 diffuse THAD. The localised sectoral THAD (11), wedge-shaped with clear border sign, were associated to thrombosis of a portal branch (6), fine-needle percutaneous biopsy (2), arterioportal shunting (2) and partial Budd-Chiari syndrome (1). The localised non-sectoral THAD (7), with variable morphology and without the clear border sign, were associated to acute inflammation of an adjacent organ (4) and to parenchymal compression by the ribs or diaphragm (3). Diffuse THAD associated to Budd-Chiari syndrome (8) and to heart failure (3) showed mosaic enhancement of hepatic parenchyma (patchy pattern); those linked to portal trunk thrombosis (4) and cirrhosis (7) revealed predominant enhancement of external hepatic parenchyma (central-peripheral phenomenon); finally, those concurrent with dilatation of the entire biliary tree showed parenchymal enhancement close to the dilated bile ducts (peribiliary pattern). Follow-up (22/43) demonstrated complete THAD regression after removal (5/22) and less conspicuity of THAD after partial overcome of the stoppage (1/22). In 2/22 cases of arterioportal shunting no substantial changes were seen. The remaining 14/22 cases showed a gradual, slow tendency towards THAD regression with hypotrophy of the involved parenchyma and compensatory contralateral hypertrophy even in the case of endurance of the causative agents.
Conclusions: Based on our experience and the literature we suggest a classification for THAD unrelated to focal hepatic lesions. We recognise 4 causes: portal vein stoppage-obstruction, portal in-flow diversion, trauma and inflammation. When THAD is related to the first three causes pathogenesis is portal hypoperfusion. In the fourth group the mediators of the arterial phenomena are those of inflammation even though portal hypoperfusion might be involved as well. THAD identification makes the detection of vascular thrombi easier by comparison with their direct finding during the portal dominant phase. Finally, THAD are to be investigated for their potential utility in the detection and characterisation of several hepatic diseases. As a consequence, hepatic CT studies cannot ignore arterial dominant phase evaluation, even if no focal hepatic lesions are expected.