Intrahepatic cholestasis related to vanishing bile duct syndrome in Hodgkin's disease.

Abstract

Jaundice is a not uncommon symptom in patients with Hodgkin's disease (HD) and generally reflects specific intrahepatic involvement, extrahepatic biliary obstruction by enlarged lymph nodes, primary infiltration of the common bile duct, or less commonly nonspecific causes such as viral hepatitis and drug toxicity (1-4). Intrahepatic cholestasis without infiltration by specific malignant cells has been reported in a small number of cases of HD (1-3,5,6). In these cases, the underlying mechanism of this complication was not yet well understood, although Hubscher et al. (5) demonstrated histological evidence of vanishing bile duct syndrome (VBDS) affecting small intraheptic bile ducts in three cases. This association between VBDS and HD is as yet unexplained (5). Before relating this syndrome to HD, it is essential to rule out other causes of VBDS. This syndrome may be developmental (syndromatic or nonsyndromatic paucity of interlobular bile ducts), immunological (primary biliary cirrhosis, graft-versus-host disease, primary sclerosing cholangitis), infective, vascular, or chemical in origin (7). In children it has been reported with histiocytosis X (8). VBDS is also a major histological finding of chronic rejection after liver transplantation (9). In this article we describe one additional patient (the first case in childhood) with progressive intrahepatic cholestasis related to VBDS. CASE REPORT A 3.5-year-old child was first seen with a 3-week history of submaxillary lymphadenopathy. He had no history of preexisting chronic liver disease or immune deficiency. Because of increasing volume, the adenopathy was biopsied and showed typical features of a mixed cellularity HD with Reed-Sternberg cells (grade II of Lukes-Rye classification). On admission his physical examination showed jaundice and enlarged cervical, submaxillary, and axillary lymphadenopathy, but no hepatosplenomegaly. His urine was dark and stool pale colored. The abdominal ultrasound, thoracoabdominal computed tomography scan, lymphangiogram, and bone marrow biopsy were normal. On admission biochemical values were consistent with cholestasis without any signs of hepatic failure (protrombin time 90%). Serological tests for hepatitis B and C, Epstein-Barr virus (EBV), cytomegalovirus (CMV), and HIV were negative; he had positive IgG antibodies for hepatitis A virus (HAV) but no IgM antibodies. The direct antiglobulin test was negative. There were no circulating immune complexes, and the level of serum α-fetoprotein was low. Wilson's disease, autoimmune hepatitis and α1-antitrypsin deficiency were ruled out. Chemotherapy with three cycles of VEBP (vincristine, VP-16, bleomycin, prednisone) followed by two cycles of ABVP (adriamycin, bleomycin, vinblastine, prednisone) and radiotherapy of the cervical areas were given for his HD. One month after his admission, because of persistent cholestasis (Fig. 1), the patient underwent a needle liver biopsy. The histopathologic examination showed signs of centrilobular cholestasis without any infiltration with HD. No evidence of virus incorporation (i.e., EBV and CMV) could be found in the liver. Paucity of interlobular bile ducts was noted; only 1 of the 10 portal tracts examined contained normal bile ducts. Cholestasis progressively increased without any signs of hepatic failure, and 2 months after the onset of the disease, a laparotomy was performed. The liver was green without nodules, and no intraabdominal lymphadenopathy was found. Perioperative cholangiography showed a patent biliary tract without any appearance of primary sclerosing cholangitis. It show the penetration of the dye into second- and third-order ducts without irregularity or parallelism as well as free flow to the duodenum. The histopathology of the biopsies performed on different segments of the liver did not show any HD infiltration; there was a centrilobular cholestasis with biliary stasis and thrombus in the portal area, a slight periportal fibrosis, and a marked ductopenia; only 1 of the 20 portal tracts examined had bile ducts (Fig. 2). Five months later, the lymphadenopathy had regressed, but cholestasis continued to increase and was associated with intermittent pruritus and a mild hepatomegaly. A third liver biopsy was performed and confirmed VBDS previously found with 5 normal bile ducts in 10 portal tracts examined, associated with periportal neoductular proliferation and moderate fibrosis. At the age of 4 years, splenomegaly appeared; 6 months later, endoscopic examination showed esophageal varices that were treated by sclerotherapy. A fourth liver biopsy was performed and showed a marked cholestasis and periportal fibrosis without hepatocyte lesion. VBDS persisted; only 1 of the 17 portal tracts examined had normal bile ducts. Because of deterioration of hepatic function and prolonged remission of HD, liver transplantation was considered, but the patient died while on the waiting list 6 months later with signs of hepatic failure. The parents refused postmortem examination. DISCUSSION When jaundice is present in HD, a satisfactory explanation for the hyperbilirubinemia cannot be found in every case even at autopsy (1-3). Perrera et al. (1) reviewed seven cases of unexplained hyperbilirubinemia in HD; five patients died 5-19 months after the onset of jaundice and the two remaining patients who responded to radiotherapy improved. Recently, Hubscher et al. (5), in addition to cholestasis of unknown etiology, found VBDS on subsequent review of the histological material of three patients. We report here one additional patient and the first case in childhood. Since jaundice developed before any drugs had been administered in our patient and the other causes of VBDS were ruled out, we can conclude that there is a direct relationship between HD and VBDS. The liver biopsy of the previously reported patients with HD and cholestasis of unknown cause showed predominantly centrilobular canalicular cholestasis with varying amounts of nonspecific inflammation. Only Juniper et al. (10) described that the bile ducts were not prominent and that a diminished number of ducts had been identified in the portal zone. It is possible that other cases of so-called idiopathic cholestasis in HD could have been associated with the same lesion, which had gone unnoticed. The patient discussed here had VBDS affecting small intrahepatic bile ducts on his histological specimen similar to that found by Hubscher et al. (5). In the present study, jaundice and cholestasis were present in our patient at the time of diagnosis. Although the patient responded to chemotherapy with additional radiotherapy, cholestasis progressed and he died from the liver disease. Of the 13 previously reported patients with idiopathic intrahepatic cholestasis, 9 had jaundice as a presenting symptom and 11 died with progressive damage (1,2,5-8). Lefkowitch et al. (11) described a patient with severe cholestatic hepatitis, who at autopsy was found to have widespread liver involvement with HD and bile ducts could not be identified, evidently obliterated by HD infiltrates. In our patient, as VBDS was discovered after the beginning of chemotherapy, prior hepatic infiltration cannot be ruled out; however, this seems unlikely because serial biopsies were taken at different times in the evolution of the disease, and particularly large-wedge biopsies were taken on different segments of the liver at the time of laparotomy. Various theories have been proposed to explain this unknown cholestasis in HD, including abnormal bile acid metabolism, toxemia, and accumulation in the body of a cholestatic hormone (12). The search for endocrine abnormalities, suspected with this unknown etiology of cholestasis in HD, was unsuccessful (12). A virus, probably responsible for the HD, has also suggested in the etiology of cholestasis (2). Recently, Waterson et al. (13) demonstrated that unexplained cholestatic jaundice might be seen before diagnosis of non-Hodgkin lymphoma, considered as a paraneoplastic phenomenon. This cholestasis of unknown etiology might therefore be of concern not only in HD, but also in other conditions. Cytostatic agents may cause hepatic damage, but we did not find any report demonstrating ductopenia in such cases (14,15). Moreover, in our patient, jaundice developed before any drugs were administered. Hubscher et al. (5) suggest that there is a release of toxic cytokines from lymphoma cells in HD. In liver transplant patients, although ischemia might be a cause of bile duct loss in rejection, current evidence suggests that direct lymphocytotoxic damage to bile ducts is the predominant mechanism. The recurrence of VBDS in some patients and slightly increased number of ductopenic rejection in some disease are in favor of an immunological disorder. These findings strongly suggest that adverse immune responses of the host tend to persist even if hepatic allografts are exchanged (9,16). From the data reported here, we can confirm, as in previous reported cases (5), that once the diagnosis of severe ductopenia has been confirmed histologically in HD, the condition is irreversible and is characterized by increasing clinical cholestasis progressing to hepatic failure. Our report confirms that VBDS is a possible mechanism for intrahepatic cholestasis in HD, even in young children. We think that VBDS should be included in the differential diagnosis of unexplained cholestasis in HD. In patients in whom lymphoma has been successfully eradicated, liver transplantation should be considered.FIG. 1: . Biochemical changes in the patient. Normal values: bilirubin (bili), 1-12 mg/L; aspartate aminotransferase (ASAT), 0-25 IU/L; γ-glutamyltranspeptidase (GGT), 5-25 IU/L; akaline phosphatase (AP), 150-400 IU/L. Times are expressed as months after first admission to hospital.FIG. 2: . Immunohistochemical staining for bile duct cytokeratins showing a portal space without any bile duct (A) and another portal space showing a ductal remnant (B; arrow). Immunoperoxidase; original magnification ×250.