Venous Sclerosis and Fibrosis in Chronic Rejection of Liver Allografts 

Index


Yuichi Nakazawa, Neal Walker, Stephen Lynch, Andrew Clouston

Departments of Pathology, Gastroenterology and Hepatology, and Surgery, Princess Alexandra Hospital, Brisbane, Australia.


Fibrosis in liver allografts undergoing chronic rejection (CR) is a variable and poorly understood process. Suggested contributing causes include the prolonged cholestasis and ductular proliferation that can accompany CR; others have noted its occurrence in a proportion of allografts with prolonged ischemia. The latter does not explain why foam-cell arteriopathy sufficient to cause severe ductopenia has unpredictable and sometimes negligible changes on the hepatic parenchyma. Recently, Wanless has suggested that fibrosis in liver disease may be the result of venous occlusion. He proposes that occlusion of both the portal and central hepatic veins results in parenchymal extinction followed by fibrosis. We studied the temporal and spatial relationships of venous and biliary lesions in an attempt to clarify their potential contributions to graft fibrosis. Seventeen patients with explanted or postmortem grafts were available for study. Initially, venous lesions were assessed. Hepatic veins (HV, including terminal hepatic venules) and portal veins (PV) were segregated by size, and almost 3000 vessels analyzed for severity, prevalence and morphology of venous lesions. Then, fibrosis was scored in periportal and centrilobular regions and this was correlated with the presence of ductular proliferation and venous lesions. Three groups were found; group 1 (n=5) with no HV lesions, group 2 (n=5) with HV lesions only, and group 3 with lesions of both HV and PV. The earliest and most common lesion to develop, in 71% of grafts, was concentric intimal thickening of small HV (central veins) which mimicked that seen in veno-occlusive disease. The lesions were significantly more severe and more frequent grafts from group 3 compared with group 2. With increasing frequency and severity of small HV sclerosis there was a significant correlation with the development of eccentric or mixed eccentric/concentric fibrosis in medium/large veins. The morphology of the lesions in larger vessels suggested organized thrombosis as the cause. Scores for fibrosis varied. Centrilobular fibrosis was significantly greater in group 3 (with dual HV and PV lesions). Conversely, portal fibrosis scores were significantly higher in grafts with ductular proliferation and did not show any correlation with the venous lesions. Portal fibrosis was slowly progressive, but centrilobular fibrosis developed unpredictably and sometimes rapidly. This suggests a model of CR in which mild venoocclusive-like lesions develop commonly in terminal hepatic venules, probably due to immune-mediated damage. In only a proportion, with increased frequency and severity of the lesions, there is stasis and thrombosis in larger veins (PV and HV) resulting in parenchymal ischemia and fibrosis. This also helps to explain why ischemic changes in the acini may not be apparent despite significant foam-cell arteriopathy and bile duct loss, in those grafts with only mild venous sclerosis. The roles played by immunological mechanisms and arterial ischemia in this model remain to be determined.

Back to top of page

Go to Abstract Index
 

Please mail comments, corrections or suggestions to the TPIS administration at the UPMC.

  University of Pittsburgh.