Correspondence *Division of Internal Medicine, 1500 East Medical Center Drive, 3116 Taubman Center, Box 0368, University of Michigan Health System, Ann Arbor, MI 48109-0368, USA

Email adebened@med.umich.edu

The case

A 39-year-old man with no past medical history presented to the hospital as a direct admission by his gastroenterologist for evaluation of a 2-month history of abdominal pain, jaundice, non-bloody diarrhea, weakness, and weight loss. He denied having similar symptoms in the past, a history of jaundice, blood transfusions, alcohol abuse, intravenous drug abuse, high-risk sexual behavior, exposure to known toxins, or a family history of liver disease. He admitted taking ginseng, bee pollen, and other unspecified over-the-counter stimulants to stay awake while driving cross-country for his job. His physical examination was remarkable for marked cachexia, mild tenderness on palpation of all four abdominal quadrants, and jaundice. No fever, lymphadenopathy, or hepatosplenomegaly was present. Initial laboratory studies revealed elevated aminotransferase and bilirubin levels, and prolonged prothrombin and partial thromboplastin time as well as an elevated International Normalized Ratio (INR; Table 1). Additional laboratory tests revealed normal concentrations of ceruloplasmin (95.4 mg/dl) and fetoprotein (2.0 ng/ml), as well as negative serologies for HAV, HBV, HCV, cytomegalovirus, Clostridium difficile, and Entamoeba histolytica. Antimitochondrial and antinuclear antibody test results were also negative. Evaluation for immune system deficiency, including tests for HIV and human T-cell lymphotrophic virus I and II, was negative. Peripheral blood flow cytometry detected a mixture of B and T cells without a monotypic cell population.

Table 1 Laboratory values for the case patient at diagnosis and 3 months afterwards.

Full tableFigures & Tables indexDownload Power Point slide (74K)

Abdominal ultrasound revealed normal hepatic and portal Doppler readouts. The results of endoscopic retrograde cholangiopancreatography and magnetic resonance cholangiopancreatography were negative. Liver core biopsy revealed moderate centrilobular cholestasis with intrahepatocytic and intracanalicular bile deposits, as well as a loss of true interlobular bile ducts in 70–80% of portal tracts, which was consistent with a diagnosis of vanishing bile duct syndrome (VBDS; Figure 1). No malignant cells were observed in the liver tissue, and flow cytometry of the cells from the liver biopsy showed that there were no monoclonal variants of lymphocytes suggestive of a hematological malignancy.

Figure 1 Vanishing bile duct syndrome.

Liver core biopsy demonstrates a paucity of bile ducts within a portal tract. Intrahepatocytic and intracanalicular bile deposits were also observed, but those images are not shown.

Full figure and legend (27K)Figures & Tables indexDownload Power Point slide (70K)

A CT scan of the thorax, abdomen, and pelvis revealed a distended gallbladder without evidence of intrahepatic biliary ductal dilatation and without evidence of lymphadenopathy or malignancy. The patient underwent two colonoscopies during which multiple biopsies were taken to investigate the cause of his diarrhea. The gross images seen during the colonoscopies were consistent with a diagnosis of IBD; however, the images did not clearly distinguish between a diagnosis of ulcerative colitis or Crohn's disease. The biopsies were remarkable for terminal ileitis and active colitis with ulceration (Figure 2), but the histological features of the biopsies were also inconclusive. Anti-Saccharomyces cerevisiae antibodies (ASCA) IgA, ASCA IgG, and anti-OmpC IgA were positive; and perinuclear antineutrophil cytoplasmic antibodies (p-ANCA) were negative. These serologies were suggestive of Crohn's disease;¹ however, because a colectomy was not performed and the gross and histological images were inconclusive, a definitive diagnosis could not be made. A diagnosis of IBD, type unclassified (IBDU) was therefore made, in accordance with The Montreal Working Party recommendation.²

Figure 2 IBD, type unclassified.

(A) Terminal ileitis. (B) Active colitis with ulceration.

Full figure and legend (44K)Figures & Tables indexDownload Power Point slide (88K)

No clear etiology for the patient's VBDS or IBDU was established during the patient's initial evaluation. The patient's jaundice and diarrhea improved with budesonide 9 mg orally for 3 days (Figure 3). In addition, the patient's INR responded to treatment with vitamin K. The patient was discharged in a stable condition 6 days after his initial presentation. He was instructed to continue taking budesonide 9 mg daily.

Figure 3 The budesonide effect.

A short course of budesonide therapy decreased the patient's total bilirubin levels and improved his diarrhea during his initial hospitalization. When budesonide therapy was discontinued during his initial hospitalization, his total bilirubin level rose. The patient was subsequently maintained on budesonide therapy until the initiation of chemotherapy and intravenous steroid infusions. To convert values from mg/dl to m/l multiply by a conversion factor of 17.1.

Full figure and legend (10K)Figures & Tables indexDownload Power Point slide (52K)

Several months later, the patient presented to the emergency department with increased weakness, fatigue, worsening jaundice, non-bloody diarrhea, decreased appetite, weight loss, and muscle wasting. Laboratory investigations were performed and the test results compared with the patient's initial laboratory test results (Table 1). The comparison was most notable for an increase in conjugated bilirubin concentration, International Normalized Ratio, and white blood cell count. A repeat CT scan of the thorax, abdomen, and pelvis revealed the presence of multiple nodal masses within the bilateral axilla and anterior mediastinum. Marked intra-abdominal lymphadenopathy was also observed (Figure 4). Biopsy of a left axillary lymph node revealed numerous large Reed–Sternberg cells in a mixed inflammatory background with large bands of sclerosis consistent with a diagnosis of nodular sclerosing Hodgkin's lymphoma. Flow cytometry of cells from the axillary lymph node biopsy revealed a predominantly T-cell population with no aberrant antigen expression. The patient was discharged 5 days after his diagnosis of Hodgkin's lymphoma with plans to initiate immediate outpatient chemotherapy.

Figure 4 Delayed lymphadenopathy.

(A) The initial abdominal non-contrast enhanced CT scan demonstrated no clinically significant lymphadenopathy. (B) Three months later, a repeat abdominal contrast enhanced CT scan demonstrated clinically significant intra-abdominal lymphadenopathy (arrows).

Full figure and legend (30K)Figures & Tables indexDownload Power Point slide (74K)

The decision was made to avoid the standard chemotherapy agents for the treatment of Hodgkin's lymphoma (doxorubicin, bleomycin, vinblastine, and dacarbazine) because doxorubicin and vinblastine are metabolized by the liver and therefore could not be safely administered in the setting of the patient's severe liver dysfunction.³ Instead, nitrogen mustard and dexamethasone were administered because of their more favorable safety profiles in the setting of liver dysfunction—especially nitrogen mustard, which is not hepatically metabolized. The patient, for a total of 5 months, received nitrogen mustard 11 mg and dexamethasone 20 mg intravenous infusions on the first and eighth days of each month, as well as varying doses of oral dexamethasone daily (4–20 mg). The patient's budesonide was discontinued before the first infusion of dexamethasone. The patient's jaundice improved after cycles one and three of chemotherapy and steroid treatment, but cycles two, four, and five failed to produce a decrease in his total bilirubin levels—instead an increase was observed. The patient's jaundice and diarrhea did eventually stabilize, but he unexpectedly died at home of unknown causes before his chemotherapy regimen was completed. An autopsy was requested, but the patient's family declined.

Discussion of diagnosis

Damage to the hepatobiliary system in patients with Hodgkin's lymphoma is usually caused by direct invasion by malignant cells or by extensive, obstructive lymphadenopathy. VBDS, which is characterized by a paucity of intrahepatic bile ducts, represents another mechanism by which Hodgkin's disease can affect the hepatobiliary system. In 1993, Hubscher et al.⁴ were the first to describe three cases of VBDS secondary to Hodgkin's lymphoma. Since then, to our knowledge, there have been 14 cases reported in the English literature that describe VBDS secondary to Hodgkin's lymphoma. There were two main histological findings of VBDS secondary to Hodgkin's lymphoma that Hubscher et al. showed: approximately 80% of portal tracts lacked interlobular bile ducts, and there was significant intrahepatic cholestasis. These two findings were present in our patient and have also been observed in other reported cases of VBDS secondary to Hodgkin's lymphoma. Lymphoma was a concern at the time of our patient's initial presentation, but it was not possible to make a tissue diagnosis because there was no pathologic lymphadenopathy. A delay in the development of lymphadenopathy in Hodgkin's lymphoma presenting as VBDS was previously noted by Crosbie et al.⁵

There are two categories of pathophysiological mechanisms that could explain how Hodgkin's lymphoma causes VBDS. The first possible mechanism is direct bile duct damage resulting from microscopic lymphoma cell infiltration of bile ducts and hepatic sinusoids.^{6, 7} This mechanism was considered by Hubscher et al. in their initial report, but not preferred, as they thought a different mechanism would better account for the severe cholestasis that was present in patients who had very small tumor burdens. These authors therefore proposed the second possible mechanism: paraneoplastic bile duct destruction caused by the release of toxic cytokines from lymphoma cells. The latter can better account for the disparity between the minor load of lymphoma and severe cholestasis seen in cases of VBDS caused by Hodgkin's lymphoma.^{4, 5, 8, 9}

With regard to the relationship between Hodgkin's lymphoma and IBDU, it is well reported that the presence of IBD is not associated with an increased risk of lymphoma, and only a small increased risk is connected to the use of immunomodulators as a treatment for IBD.^{10, 11} A few authors have described lymphoma presenting as IBD.^{12, 13}

Unique to this case is our patient's concomitant presentation of VBDS and IBDU before his diagnosis of Hodgkin's lymphoma. The onset of the patient's jaundice and diarrhea coincided; this temporal relationship argues in favor of a paraneoplastic process affecting both intrahepatic biliary and colonic epithelium. In this case, the jaundice and diarrhea responded to steroid treatment, initially budesonide and later dexamethasone, which implies an inflammatory etiology. This particular presentation could also be caused by T-cell dysfunction associated with the lymphoma, leading to cytotoxic effects at the biliary level and marked anergy in the colon.

Differential diagnosis

In this case, jaundice caused by VBDS and diarrhea caused by IBDU were the clinical manifestations of what was probably a paraneoplastic process associated with Hodgkin's lymphoma. There are, however, several other relevant etiologies that were considered during diagnosis. In regard to VBDS, our initial evaluation focused on etiologies that cause aminotransferase abnormalities with predominantly a cholestatic pattern, including viral hepatitis, primary biliary cirrhosis, primary sclerosing cholangitis, and biliary obstruction. The patient's cholangiographic evaluation was negative, however, and the liver biopsy did not suggest primary sclerosing cholangitis, as the patient had no observed fibrosis of the surrounding bile duct tissue in an onion skin-like pattern, duct atrophy, or replacement of ducts with rounded or ovoid scars.¹⁴

We deliberated on the possible etiologies that primarily present as hepatocellular injury, which include viral hepatitis, autoimmune hepatitis, alcoholic hepatitis, hemochromatosis, -1 antitrypsin deficiency, and Wilson's disease. These were ruled out on the basis of our patient's clinical presentation as well as laboratory and radiological evaluations. Diseases that cause infiltrative patterns of liver dysfunction such as HIV, cytomegalovirus, and hepatocellular carcinoma were also considered. In addition, the possibility that the patient had drug-induced cholestasis caused by over-the-counter stimulants was considered. To our knowledge, however, there had been only one case study evaluating cholestasis after the use of a product containing ginseng: in this patient, unlike our patient, there was no evidence of ductopenia.¹⁵ In regard to our patient's diarrhea, he was determined to have IBDU on the basis of evaluation of gross colonoscopic images, histological biopsies, serologies, and a negative infectious workup.

Treatment and management

As explained above, standard treatment for Hodgkin's lymphoma—doxorubicin, bleomycin, vinblastine and dacarbazine—was not used in this case because of the patient's severe liver dysfunction. Instead, he received nitrogen mustard and dexamethasone. He had both intrahepatic cholestasis and a paucity of bile ducts on liver biopsy. The presence of ductopenia is considered a poor prognostic indicator when treating Hodgkin's disease that presents as VBDS.¹⁶ Hodgkin's disease that presents with only intrahepatic cholestasis has a more favorable outcome and might be amenable to single-modality radiation therapy alone.¹⁶

Conclusion

This case highlights the challenges of diagnosing patients who present with concomitant pathologies that are not usually associated with each other. Hodgkin's lymphoma should be considered in any patient who presents with ductopenia, intrahepatic cholestasis, VBDS, and/or IBDU, despite the absence of pathologic lymphadenopathy, elevated white blood cell count, or suggestive flow cytometry results. Early lymph node biopsy should be considered if no other diagnosis is made. Alternatively, patients could be closely followed for the development of lymphadenopathy and/or white blood cell count elevation, at which point lymph node biopsy might be warranted.

References

1. Reese GE et al. (2006) Diagnostic precision of anti-Saccharomyces cerevisiae antibodies and perinuclear antineutrophil cytoplasmic antibodies in inflammatory bowel disease. Am J Gastroenterol 101: 2410–2422 | Article | PubMed | ChemPort |
2. Satsangi J et al. (2006) The Montreal classification of inflammatory bowel disease: controversies, consensus, and implications. Gut 55: 749–753 | Article | PubMed | ChemPort |
3. Koren G et al. (1992) The effects of impaired liver function on the elimination of antineoplastic agents. Ann Pharmacother 26: 363–371 | PubMed | ChemPort |
4. Hubscher SG et al. (1993) Vanishing bile duct syndrome: a possible mechanism for intrahepatic cholestasis in Hodgkin's lymphoma. Hepatology 17: 70–77 | PubMed | ChemPort |
5. Crosbie OM et al. (1997) Resolution of paraneoplastic bile duct paucity following successful treatment of Hodgkin's disease. Hepatology 26: 5–8 | Article | PubMed | ChemPort |
6. Cavalli G et al. (1979) Changes in the small biliary passages in the hepatic localization of Hodgkin's disease. Virchows Arch A Pathol Anat Histol 384: 295–306 | Article | PubMed | ChemPort |
7. Lefkowitch JH et al. (1985) Hepatic Hodgkin's disease simulating cholestatic hepatitis with liver failure. Arch Pathol Lab Med 109: 424–426 | PubMed | ChemPort |
8. Gottrand F et al. (1997) Intrahepatic cholestasis related to vanishing bile duct syndrome in Hodgkin's disease. J Pediatr Gastroenterol Nutr 24: 430–433 | Article | PubMed | ChemPort |
9. Yusuf MA et al. (2000) Jaundice caused by vanishing bile duct syndrome in a child with Hodgkin lymphoma. J Pediatr Hematol Oncol 22: 154–157 | Article | PubMed | ChemPort |
10. Loftus EV Jr et al. (2000) Risk of lymphoma in inflammatory bowel disease. Am J Gastroenterol 95: 2308–2312 | Article | PubMed |
11. Lewis JD et al. (2001) Inflammatory bowel disease is not associated with an increased risk of lymphoma. Gastroenterology 121: 1080–1087 | Article | PubMed | ChemPort |
12. Ramasamy KA et al. (2002) A case of primary non-Hodgkin's lymphoma of the transverse colon presenting as inflammatory bowel disease. Eur J Gastroenterol Hepatol 14: 1401–1403 | Article | PubMed |
13. Vincenzi B et al. (2001) Unusual presentation of Hodgkin's disease mimicking inflammatory bowel disease. Leuk Lymphoma 42: 521–526 | PubMed | ChemPort |
14. Scheuer PJ (1998) Ludwig Symposium on biliary disorders—part II. Pathologic features and evolution of primary biliary cirrhosis and primary sclerosing cholangitis. Mayo Clin Proc 73: 179–183 | PubMed | ChemPort |
15. Hamid S et al. (1997) Protracted cholestatic hepatitis after the use of prostata. Ann Intern Med 127: 169–170 | PubMed | ChemPort |
16. Barta SK et al. (2006) Idiopathic cholestasis as a paraneoplastic phenomenon in Hodgkin's lymphoma. Clin Lymphoma Myeloma 7: 77–82 | PubMed |

Contact the journal about this article

Subject areas under which this article appears: Gall bladder and biliary tract | Inflammatory bowel disease | Cancer