Search

Although nonalcoholic steatohepatitis (NASH) has generally been considered a benign condition, the increasing prevalence and severity of obesity has heightened concerns about the frequency with which NASH progresses to end-stage liver disease. The aim of this study is to determine the frequency, clinical features, and posttransplantation history of decompensated liver disease secondary to NASH. The frequency of NASH as a cause of end-stage liver disease was prospectively determined in patients evaluated for liver transplantation. NASH was considered to be the primary cause of liver disease in patients who had histological evidence of steatohepatitis and in whom chronic liver diseases other than NASH were excluded. Posttransplantation histological characteristics were also determined in patients with NASH and compared with those of patients with pretransplantation diagnoses of cholestatic liver diseases, alcoholic disease, and hepatitis C. Of 1,207 patients evaluated for liver transplantation during the study period, 31 patients (2.6%) had NASH as the primary cause of liver disease. In the same period, 546 liver transplantations were performed, 16 of which (2.9%) were for end-stage disease secondary to NASH. Posttransplantation steatosis was seen in 60% of transplant recipients with NASH versus 5% of those with cholestatic disease, 15% of those with alcoholic disease, and 15% of those with hepatitis C. Steatohepatitis recurred in 33% of transplant recipients with NASH, with progression to cirrhosis in 12.5%. NASH can progress to end-stage liver disease in a minority of affected patients and was the primary cause of liver disease in 2.9% of patients evaluated for liver transplantation at our center. Recurrence of steatosis and NASH is frequent and can be severe after liver transplantation.

The cell of origin and direction of differentiation of the clear cell tumor of the lung (the so-called sugar tumor) remains enigmatic. Recognition of HMB-45 immunoreactivity and identification of melanosomes have suggested a relationship to angiomyolipoma of kidney or liver and lymphangiomyoma. This has given rise to the concept that clear cell tumors are neoplasms of so-called perivascular epithelioid cells--PEComas. Herein we report the existence of four similar tumors occurring in extrapulmonary sites, one of which had malignant features. The three benign tumors occurred in females ages 9, 20, and 40 years; two were located in the rectum and one in the vulva. The malignant tumor occurred in the inter-atrial cardiac septum of a 29-year-old man. Common histologic features were a richly vascular organoid architecture, tumor cells with clear to pale eosinophilic cytoplasm, abundant glycogen, and immunoreactivity for HMB 45, but not S100, multiple keratin, neuroendocrine, or muscle markers. Benign tumors demonstrated low mitotic activity, no necrosis, and good circumscription; the malignant tumor showed considerable mitotic activity, necrosis, and an infiltrative growth pattern. Ultrastructurally, glycogen was present, mitochondria were abundant, and membrane-bound lamellated bodies consistent with premelanosomes were present in two cases, and equivocal in one. Because these tumors have light microscopic, immunohistochemical, and electron microscopic features similar to pulmonary sugar tumors, we propose the name primary extrapulmonary sugar tumor (PEST) for them. Although most PEST's are probably benign, malignant forms appear to exist. These cases further support the concept of a family of systemic HMB-45 positive tumors that include sugar tumors, angiomyolipoma of kidney or liver, lymphangiomyomas, and clear-cell myomelanocytic tumors of the falciform ligament/ligamentum teres.

Background & Aims: The epidemiology of primary biliary cirrhosis (PBC) has not been studied systematically in the United States. We report the incidence and prevalence of this condition in the general population. We also examined the validity of the Mayo natural history model for PBC among these unselected patients from the community., Methods: The Rochester Epidemiology Project entails a computerized index of diagnoses from the health care encounters of residents of Olmsted County, Minnesota. For potential cases identified using this database, the complete (inpatient and outpatient) medical records were reviewed to verify the diagnosis and extract information necessary for the application of the Mayo model. We estimated the incidence and prevalence of PBC in this population and compared the actual survival of patients with PBC in the community with the survival predicted for PBC patients by the Mayo natural history model., Results: The age-adjusted (to 1990 U.S. whites) incidence of PBC per 100,000 person-years for years 1975-1995 was 4.5 (95% confidence interval [CI], 3.1-5.9) for women, 0.7 (95% CI, 0.1-1.3) for men, and 2.7 (95% CI, 1.9-3.5) overall. The age- and sex-adjusted prevalence per 100,000 persons as of 1995 was 65.4 (95% CI, 43.0-87.9) for women, 12.1 (95% CI, 1.1-23.1) for men, and 40.2 (95% CI, 27.2-53.1) overall. The Mayo natural history model accurately predicted the actual survival of these patients., Conclusions: This first description of the epidemiology of PBC in the United States indicates that its incidence and prevalence in this country are among the highest reported. Outcomes among these unselected patients from a community population further validated the Mayo natural history model of PBC.

After orthotopic liver transplantation (OLT), patients with chronic hepatitis C virus (HCV) infection show nearly universal persistence of viremia and reinfection of the liver, but identifying the point at which the liver is reinfected morphologically can be difficult. One tool that may potentially be useful to detect reinfection is reverse transcriptase-polymerase chain reaction (RT-PCR), which has proven to be highly sensitive for detecting HCV RNA in formalin-fixed paraffin-embedded liver tissue. Our purpose was to gain insight into the time frame of HCV reinfection by assaying for HCV RNA in serial posttransplant liver biopsy specimens. Our study population consisted of 14 patients who underwent liver transplantation for hepatitis C and had confirmed HCV RNA in pretransplant serum, absence of HCV RNA in donor livers, and available consecutive posttransplant liver allograft specimens. We performed RT-PCR for HCV RNA in serial posttransplant liver biopsy specimens, beginning at 1 week until at least one biopsy from each tested positive. HCV RNA was detected in liver tissue by RT-PCR in 1-week post-OLT liver samples in 6 of 14 (42.8%) patients, the earliest being 5 days post-OLT. Eventually, each of the remaining eight samples became RT-PCR positive as well; the first detections occurred in these at 3 weeks (three cases), 4 weeks (three cases), 48 weeks (one case), and 144 weeks (one case). Histologic identification of hepatitis C recurrence was relatively insensitive in relation to these molecular data. These data suggest that (1) HCV RNA reinfection is nearly universal after liver transplantation in patients with chronic hepatitis C infection, (2) molecular reinfection by HCV occurs at a variable interval post-OLT, with the majority of allograft livers reinfected as early as 1 week, and (3) morphologic features of hepatitis C are usually appreciable at the time of "molecular" recurrence.

Chronic hepatic allograft rejection is characterized by the histological findings of ductopenia and a decreased number of hepatic arteries in portal tracts in the presence of foam cell (obliterative) arteriopathy. Recent studies have extended the histological spectrum of chronic rejection to include the presence of biliary epithelial atrophy or pyknosis involving the majority of small ducts present in the liver biopsy specimen. Overall, the incidence of chronic rejection in adults appears to be decreasing and is currently approximately 4%. However, the incidence of chronic rejection in pediatric liver transplant recipients has been more stable and ranges from 8% to 12% in most studies. Clinical risk factors associated with chronic rejection include: underlying liver disease, HLA donor-recipient matching, positive lymphocytotoxic cross-match, cytomegalovirus infection, recipient age, donor-recipient ethnic origin, male donor into female recipient, number of acute rejection episodes, histological severity of acute rejection episodes, low cyclosporine trough levels, and retransplantation for chronic rejection. Chronic rejection, once diagnosed, frequently leads to graft failure; however, a number of reports indicated 20% to 30% of the patients with this diagnosis may respond to additional immunosuppressive therapy or even resolve spontaneously receiving baseline immunosuppression. Newer immunosuppressive agents, such as tacrolimus and mycophenolate, may successfully reverse chronic rejection, particularly when it is diagnosed in its early histological stages.

Hereditary hemochromatosis (HHC) is the most common inherited single gene disorder in people of northern European descent. Hereditary hemochromatosis is characterized by increased intestinal absorption of iron leading to its deposition into multiple organs. The classic description of HHC is bronze diabetes in a patient with cirrhosis. Hereditary hemochromatosis is increasingly being diagnosed at an earlier, less symptomatic stage. Diagnosis is based on an elevated fasting early morning transferrin saturation. Treatment is by phlebotomy, which, if initiated before the development of cirrhosis or diabetes, is associated with a normal life expectancy. Recently, a gene associated with HHC was discovered and named HFE. Two point mutations of this gene have been referred to as C282Y and H63D. Several US and European studies have found that 60% to 93% of patients with suspected HHC are homozygous for C282Y. Positive results of HFE gene testing may eliminate the need for a liver biopsy in selected cases. The greatest utility of HFE gene testing will likely be in screening family members of an identified proband and in helping to resolve ambiguous cases.

Hepatic allograft rejection has been divided into humoral (or hyperacute), acute (or cellular), and chronic (or ductopenic) forms. Humoral rejection is extremely uncommon in the liver and is not graded. Acute rejection will occur in approximately 50% of liver allografts, is more common in the first few weeks posttransplantation, and is defined by Snover's triad of portal hepatitis, endothelialitis (or endotheliitis), and lymphocytic cholangitis. This form of rejection is generally reversible, either spontaneously or with additional immunosuppressive therapy, and can be reliably graded using a system with categories of mild, moderate, and severe rejection, associated with 37%, 48%, and 75% unfavorable shortterm and 1%, 12%, and 14% unfavorable long-term outcomes, respectively. Chronic rejection is characterized histologically by progressive duct loss and a lipid-rich vasculopathy that can be difficult to diagnose in early phases. Chronic rejection typically occurs several months to a year posttransplantation, although exceptions exist.

Hepatic lymphangiomatosis is a rare disorder characterized by cystic dilatation of the lymphatic vessels in the hepatic parenchyma. It can occur in the liver alone, in the liver and spleen, or in multiple organs. Clinically, diagnosis can be difficult because of the rarity and protean manifestations of this disorder. We describe a 53-year-old woman with hepatic lymphangiomatosis in whom polycystic liver disease had been previously diagnosed. In addition, we review 12 cases of hepatic, splenic, and hepatosplenic lymphangiomatosis with or without systemic lymphangiomatosis and discuss the differential diagnosis.

Currently, one of the major indications for liver transplantation is infection with hepatitis C virus (HCV). Many studies have suggested that recurrent infection with HCV is universal after transplantation. Fastidious techniques, such as reverse transcriptase-polymerase chain reaction (RT-PCR), have proved to be highly sensitive for detecting HCV RNA in serum and in fresh-frozen and formalin-fixed paraffin-embedded (FFPE) liver tissue. In this study, we wanted to determine whether the identification of HCV RNA in liver tissue by RT-PCR might reflect the detection of circulating HCV RNA in blood within the tissue, rather than implying true tissue infection. We performed RT-PCR for HCV RNA in FFPE liver biopsy specimens taken from 14 donor allografts shortly before and immediately after implantation into recipients. The recipients were known to have HCV RNA in serum and explanted liver tissue, as determined by RT-PCR. We were unable to detect HCV RNA in any of the study samples, either before or after transplantation. In a related study, qualitative and quantitative HCV RNA analyses were performed by RT-PCR and branched DNA (bDNA) amplification, respectively, on serum samples collected pretransplantation and immediately posttransplantation from 10 other patients who underwent transplantation for hepatitis C. HCV RNA was detected in all serum samples before and after transplantation by RT-PCR; however, the bDNA assay detected HCV RNA in only 6 of 10 samples pre-orthotopic liver transplantation (OLT) and in none of the immediately post-OLT samples. In our system, despite the RT-PCR detection of HCV RNA in serum before and after the transplantation, HCV RNA is not detectable in the peripheral blood that accompanies formalin-fixed liver tissue. This implies that RT-PCR detection of HCV RNA in tissue reflects true liver infection, rather than contamination by HCV RNA in accompanying peripheral blood.

A case of hepatoid adenocarcinoma of the stomach is reported, and the literature is reviewed. The stomach is one of the most common sites in which hepatoid adenocarcinomas have been detected. Of the 59 cases reviewed from the literature (including the current case), a 2:1 male predominance was noted, and the serum alpha-fetoprotein level was almost always increased. All patients were adults, and most had evidence of metastases at the time of resection. Prognosis seems less favorable than that associated with the more common intestinal type of adenocarcinoma of the stomach. Hepatoid adenocarcinomas typically show periodic acid-Schiff-positive, diastase-resistant intracytoplasmic globules, which are demonstrated to be positive with antibodies to alpha-fetoprotein. The tumor cells resemble liver cells, and rare cases, including our own, have evidence of bile production. In our case, messenger RNA for albumin, unique to liver cells, was demonstrated by in situ hybridization of the tumor cells.

Background & Aims: The effects of parenteral growth hormone, glutamine supplementation, and a high carbohydrate-low fat (HCLF) diet on gut adaptation in short-bowel syndrome are unclear. The aim of this study was to compare effects of this treatment regimen and placebo in patients with short-bowel syndrome., Methods: A randomized, 6-week, double-blind, placebo-controlled, crossover study in 8 patients with short-bowel syndrome (average small bowel length, 71 cm; mean duration, 12.9 years) was performed. Active treatment was growth hormone (0.14 mg.kg-1.day-1), oral glutamine (0.63 g.kg-1.day-1), and the HCLF diet for 21 days. The weight, basal metabolic rate, nutrient and electrolyte balance, serum insulin-like growth factor I levels, D-xylose absorption, morphology and DNA proliferation of small intestinal mucosa, and gastrointestinal transit were evaluated. Treatments were compared by paired t test., Results: Active treatment transiently increased body weight, significantly but modestly increased the absorption of sodium and potassium, and decreased gastric emptying. The assimilation of macronutrients, stool volumes, and morphometry of small bowel mucosa were not statistically different in the two treatment arms., Conclusions: Although treatment with growth hormone, glutamine, and HCLF diet for 3 weeks resulted in modest improvements in electrolyte absorption and delayed gastric emptying, there were no improvements in small bowel morphology, stool losses, or macronutrient absorption.

Objective: To determine whether diffuse flat dysplasia, akin to that seen with chronic ulcerative colitis, occurs as a mucosal field defect accompanying small bowel adenocarcinoma in patients with celiac sprue., Design: A pathologic investigation of archival tissue was undertaken., Material and Methods: From the tissue archives at Mayo Clinic Rochester for the period from January 1978 to January 1993, resected sprue-associated adenocarcinomas with adequate adjacent mucosa for study were identified. On the basis of multiple strip biopsy specimens obtained at the tumor margins and within 4 cm distal and proximal to the tumor, histologic maps were created., Results: Among 94 patients with primary adenocarcinoma of the small bowel, 8 had a concurrent diagnosis of celiac sprue. Of these eight cases, six surgically resected specimens (five duodenal and one jejunal adenocarcinoma) were adequate for study. A thin rim of benign dysplasia was demonstrated at the tumor margin in three of the six specimens; however, contiguous fields of flat dysplasia were not present in any of these cases., Conclusion: These data support focal dysplasia (for example, adenoma) rather than sheets of flat dysplasia as the premalignant lesion in sprue-associated small bowel adenocarcinoma. Blind endoscopic biopsies of small bowel mucosa are not justified for neoplasia surveillance in patients with celiac sprue.

Angiotropic large cell lymphoma is a rare, aggressive type of malignant lymphoma that primarily involves intravascular spaces and most often has clinical manifestations in the skin and central nervous system. Virtually any organ can be affected, however, including the lymph nodes and spleen. Peripheral blood involvement is usually not detectable morphologically. Conventional lymphoma in association with this entity has also been described. Herein we present a case of angiotropic lymphoma of B-cell lineage that affected the liver and skin. Reanalysis of a lymph node specimen that had been excised 3 years previously demonstrated a follicular small cleaved cell lymphoma. To our knowledge, this is the first reported case of possible evolution of follicular lymphoma to large cell lymphoma of angiotropic type.

Background & Aims: Intraductal papillary-mucinous tumor (IPMT) of the pancreatic ducts is increasingly recognized. This study investigated if clinical, imaging, or, histological features predicated outcome, formulated a treatment algorithm, and clarified relationships among IPMT, mucinous cystic neoplasms of the pancreas (MCN), and chronic pancreatitis., Methods: The medical records, radiographs, and pathological specimens of 15 patients with IPMT (dilated main pancreatic duct or branch ducts with mucin overproduction) who were evaluated between October 1983 and January 1994 were reviewed., Results: One patient had hepatic metastases. Fourteen underwent an operation (6 distal pancreatectomy, 4 total pancreatectomy, and 4 pancreaticoduodenectomy); all had dysplastic intraductal epithelium and chronic pancreatitis, whereas 3 had invasive adenocarcinoma. After a median of 25 months, 10 patients were alive; 3 of 4 with malignant and 2 of 11 with benign IPMT died (P < 0.05). Patients with or without carcinoma had similar clinical and radiographic features. A clinical diagnosis of chronic pancreatitis had been made in 9 patients with benign IMPT and in none with malignant IPMT (P < 0.05)., Conclusions: IPMT is a dysplastic and likely precancerous lesion that is frequently diagnosed as chronic pancreatitis and is separate from MCN. Because it is not possible to distinguish noninvasive from invasive IPMT preoperatively, complete surgical excision of the dysplastic process is our treatment of choice whenever appropriate.

Background/aims: Uncontrolled studies suggest that cyclosporine administered as an enema may be of benefit for left-sided ulcerative colitis and safer than intravenous or oral administration. The efficacy and safety of cyclosporine enemas for left-sided ulcerative colitis in a placebo-controlled trial was assessed., Methods: Steroid and mesalamine enemas were withdrawn before the study. Forty patients were assigned to 1 of 4 strata: no concomitant therapy, oral steroids, oral salicylates, or oral steroids and salicylates. After stratification, patients were randomized to nightly treatment with 350 mg cyclosporine (n = 20) or placebo (n = 20) enemas. Clinical response was determined at baseline and 4 weeks by endoscopy, physician assessment, and a patient diary of daily symptoms. Trough blood cyclosporine levels were measured by high-performance liquid chromatography., Results: At 4 weeks, 8 of 20 patients (40%) who received cyclosporine showed clinical improvement compared with 9 of 20 patients (45%) who received placebo. One patient receiving cyclosporine had reversible neutropenia attributable to sulfasalazine, and another patient receiving cyclosporine was unable to tolerate the enema vehicle. No other toxicity was noted during the trial. Blood cyclosporine levels were detectable in only two patients., Conclusions: Cyclosporine enemas administered in a dosage of 350 mg/day for 4 weeks are not efficacious in mildly to moderately active left-sided ulcerative colitis.

Objective: To develop a "Pouchitis Disease Activity Index" (PDAI) and to compare it with other diagnostic scoring systems for pouchitis., Design: We compared patients who had an optimal outcome with those who had a poor result attributable to recurrent pouchitis after ileal pouch-anal anastomosis (IPAA) for ulcerative colitis at the Mayo Clinic., Material and Methods: We evaluated the applicability of a PDAI that quantitated clinical findings and the endoscopic and histologic features of acute inflammation in four groups of patients: (1) 10 who underwent IPAA for ulcerative colitis and had symptoms compatible with a clinical diagnosis of pouchitis, (2) 5 who underwent IPAA for ulcerative colitis and did not have pouchitis, (3) 5 who underwent IPAA for familial adenomatous polyposis and had no symptoms of pouchitis, and (4) 5 who had a Brooke ileostomy for ulcerative colitis (control group)., Results: The PDAI was significantly greater in patients with the clinical features of pouchitis than it was for patients in the other three groups. All 10 patients with pouchitis fulfilled the PDAI criteria for a diagnosis of pouchitis; in contrast, only 1 of these 10 patients met the diagnostic criteria for pouchitis by application of previously established scoring systems. No asymptomatic patient qualified for a diagnosis of pouchitis by the PDAI criteria., Conclusion: The PDAI provides simple, objective, and quantitative criteria for pouch inflammation after IPAA and is more sensitive than prior scoring systems.

Background: Iron overload such as that in idiopathic hemochromatosis is a well-established, albeit rare, cause of non-insulin-dependent diabetes mellitus (NIDDM). Most patients with NIDDM have no recognized cause of their disease. Investigators have proposed that subclinical iron overload may cause diabetes mellitus in a substantial number of patients with NIDDM., Objective: The aim of the current study was to evaluate hepatic iron stores in autopsy specimens from a group of community residents with NIDDM., Methods: Fifteen patients with NIDDM and 17 age-matched control subjects were identified from a review of medical records of decreased residents of Olmsted County, Minnesota. Formalin-fixed liver tissue was analyzed for iron concentration by flameless atomic absorption spectrophotometry, and distribution of hepatic iron was determined histochemically., Results: No significant difference was found in either the distribution or the mean amount of hepatic iron between the diabetic and the control group (1,303 versus 1,349 micrograms Fe/g dry weight; P = 0.87). Thus, the mean difference was -46 micrograms Fe/g dry weight (confidence interval, -631 to 540)., Conclusion: Because hepatic iron quantification is the definitive means of assessing total body iron stores, our results suggest that NIDDM is typically not associated with a substantial level of iron overload.

Objective: This study evaluated the outcome of liver grafts from ABO incompatible donors, focusing on biliary complications, and compared the results to an ABO compatible control group. Also, the expression of donor ABH antigens in the liver graft was analyzed., Summary Background Data: The outcome of liver transplantation using an ABO incompatible graft is still debated. These blood group related (ABH) antigens are known to be expressed not only on the surface of the erythrocytes, but also on the epithelial cells of large bile ducts. Because the biliary epithelium of hepatic allografts may continue to express donor ABH antigens, it may be more susceptible to immunologic bile duct injury after transplantation across the ABO barrier., Methods: Eighteen ABO incompatible grafts were compared with 18 ABO compatible grafts in patients who were matched according to medical urgency, primary liver disease (PLD), and recipient age. After transplantation, the grafts were analyzed with cholangiography, Doppler ultrasound, or arteriography and liver histology according to protocol. Immunoperoxidase staining for ABH antigens was performed on hepatic tissue., Results: Biliary complications developed in 82% of the ABO incompatible donors, compared to 6% of the ABO matched controls. Hepatic artery thrombosis occurred in 24%. Cellular rejection was diagnosed in 65% versus only 28% in the control group. The 1-year actuarial graft survival rate was 44% versus 78% in the control group. ABH antigens of the donor were expressed on vascular endothelium and bile duct epithelial cells as long as 150 days after transplant., Conclusions: Using ABO incompatible allografts, a high incidence of biliary and hepatic artery complications and decreased graft survival in liver transplantation were found. An immunologic injury to the bile duct epithelium and/or to vascular endothelium is suspected.

Surreptitious abuse of laxatives is a common cause of severe chronic diarrhea. Standard laboratory screening studies of urine and stool specimens may identify phenolphthalein, diuretics, and magnesium-containing agents. An assay for bisacodyl, a commonly used over-the-counter laxative, however, is not included in routine screening tests. Herein we describe two patients with chronic watery diarrhea of large volume; analysis of stool and urine samples revealed that surreptitious use of bisacodyl was the cause. In one patient, nonspecific inflammatory changes of the colonic mucosa were noted on biopsy, and fecal leukocytes were detected in both patients. In a prospective study of eight patients who received bisacodyl as part of a preparation for colonoscopy, we analyzed serial urine samples for bisacodyl diphenol during a 48-hour period. This metabolite was found in seven of eight hydrolyzed urine samples obtained 12 hours after oral administration of bisacodyl but not in samples obtained 24 and 48 hours after ingestion of the laxative. We recommend that urinalysis and, in some cases, stool analysis for bisacodyl should be considered in the diagnostic assessment for surreptitious use of laxatives.

The diagnosis of homozygous hemochromatosis (HH) should be based on appropriate findings on liver biopsy specimens. In cases with equivocal morphologic features, quantitative tissue iron determination and calculation of the hepatic iron index generally enable one to distinguish HH from other types of hepatic iron overload. In this article, we describe a diagnostic algorithm that is designed to avoid diagnostic errors because of histologic misinterpretation or erroneous, usually false-negative, chemical iron studies. The algorithm also delineates a cost-effective method of using quantitative tissue iron analysis. Diagnostic biopsy features of uncomplicated HH include (1) hemosiderosis that involves primarily hepatocytes, with or without inactive cirrhosis, and (2) a tissue iron index of 1.9 or higher. Findings such as prominent fatty changes or lymphocytic piecemeal necrosis indicate the presence of HH in conjunction with another complicating condition or secondary iron overload in the absence of HH.

In this report, our objectives are to introduce the term "ischemic cholangitis" as an etiologic designation and to describe its manifestations. Herein we use the label "ischemic cholangitis" as a collective term for ischemic bile duct necrosis, cholangitis caused by ischemia but without necrosis, and biliary fibrosis as a manifestation of ischemic damage. The condition was observed in 12 allografts, either at the time of retransplantation (9 cases) or at autopsy (3 cases). Ischemic cholangitis involved primarily perihilar extrahepatic and intrahepatic bile ducts. The findings included duct necroses (eight cases), strictures (four cases), and cholangiectases (four cases); some of these features coexisted. In addition, complicating ascending cholangitis and cholangitic abscesses were noted in three cases. Ischemic cholangitis was caused by hepatic artery thrombosis (in nine patients) or stenosis (in one) or by occlusion of parabiliary arteries by fibrointimal proliferations, probably attributable to old thromboses (in two, in conjunction with associated foam cell arteriopathy in one). Biopsy specimens before retransplantation or autopsy were obtained in 11 patients, only 1 of whom had an infarct as direct evidence of ischemia. Nine patients had evidence of biliary obstruction or bile flow impairment; in two cases, specimens were normal or nondiagnostic relative to cholangitis. Features of cellular rejection associated with the manifestations of bile flow impairment and ischemia were noted in five cases. Thus, biopsy features that suggest biliary obstruction, with or without cellular rejection, may be a manifestation of ischemic cholangitis. We conclude that ischemic cholangitis is an important cause of cholestatic graft failure but that this type of cholangitis is difficult to diagnose because of its misleading biopsy manifestations.

A 62-year-old man with a 20-year history of chronic ulcerative colitis and a 9-year history of primary sclerosing cholangitis (PSC) underwent orthotopic liver transplantation because of symptoms related to PSC and cholangiographic features compatible with a biliary neoplasm. Study of the excised liver revealed papillary mucosal lesions in the common hepatic duct and the right and left hepatic ducts as well as cholangiectases and other features typically associated with PSC. The papillary lesions consisted of abundant fibrovascular stroma covered by biliary epithelium with low-grade and high-grade dysplasia. Some periductal glands were also dysplastic. These features distinguished papillary dysplasia from classic biliary papillomatosis. Only one focus of microinvasion was found; there were no metastases. Among 60 cases of PSC in whom the entire liver could be studied after orthotopic liver transplantation, this was the only instance of unequivocal dysplasia. However, in one specimen, papillary hyperplasia was found. Detailed macroscopic and microscopic rereview of 23 livers from our patients with the longest history of PSC (range, 5-24 years) failed to reveal any additional cases with dysplasia. It is concluded that (a) papillary mucosal lesions in PSC may represent papillary dysplasia without invasion; (b) these lesions may evolve from papillary hyperplasia; (c) the process may be largely, if not entirely, in situ; and (d) the prevalence of dysplasia and carcinoma of bile ducts may be less than the 7%-9% reported in the literature for malignancies associated with PSC.

The authors present a nonimmunocompromised patient who developed refractory fungal (Candida albicans) cholangitis and who subsequently underwent successful orthotopic liver transplantation for progressive destruction of intrahepatic and extrahepatic bile ducts. Liver biopsy specimens showed periductal abscesses, and cholangiograms showed rapidly progressive destruction of intrahepatic and extrahepatic bile ducts. Standard immunosuppressive therapy and perioperative amphotericin B were administered. At 6 months, the patient was clinically well with no evidence of recurrent biliary candidiasis. The authors suggest that orthotopic liver transplantation may be a reasonable therapeutic option for patients with refractory biliary candidiasis.

Endotheliitis (EN) is a feature of hepatic allograft rejection, characterized by the adherence of immunocytes to the endothelium of veins, often leading to endothelial damage, endophlebitis, and, sometimes, panphlebitis. We found EN at least once in 28 of 41 allografts (68%) that had survived 6 months or longer. In approximately half the affected cases, the condition recurred. The EN was mild in most instances; moderate or severe manifestations were found in only 13% of the cases. The histologic changes were present for about 1 week in approximately half the cases; a duration of more than 2 weeks was noted in 17%, and then EN usually persisted. After retransplantation, recurrence of EN was observed in all of nine cases. We were unable to establish specific clinicopathologic and laboratory correlations for EN. The immunocytes in EN consisted mainly of helper and suppressor/cytotoxic T cells as well as natural killer cells. Sometimes, the immunocytes assumed a blastlike appearance; in these instances, the condition was severe. Such blastlike changes may be specific for EN. The immunocytes were attached to the endothelium by pseudopodia, broad bases, or both; some also were interconnected by cytoplasmic bridges. Underlying endothelial cells often showed evidence of cytoplasmic damage. The pathogenesis of EN is not completely understood; the immunocytes probably attach themselves to antigenic epitopes. Their nature, however, has not been clearly identified; HLA-A, B, and C and HLA-DR were displayed in areas of EN, but the antigens also were found in vessels without EN.

Chronic liver disease associated with hepatitis B virus infection is both common and serious; no satisfactory treatment currently exists. Orthotopic liver transplantation is an option for patients with end-stage liver disease associated with hepatitis B virus infection despite the risk of allograft reinfection. Passive immunoprophylaxis has been attempted perioperatively to prevent graft infection but has not been beneficial. Some patients with chronic type B hepatitis have benefited clinically from antiviral therapy and, in particular, interferon, but its use has not previously been reported as an approach to prevent allograft infection. We administered recombinant leukocyte A interferon perioperatively to a patient who underwent liver transplantation for type B chronic active hepatitis and cirrhosis. Circulating hepatitis B virus DNA was found postoperatively while the patient was receiving interferon, and stainable viral antigen subsequently reappeared in the transplanted liver. Thus, the drug failed to prevent viral replication and allograft infection. Thus far, no evidence of progression of the chronic hepatitis has been noted.