Genetic and morphological findings in progressive familial intrahepatic cholestasis (Byler disease [PFIC-1] and Byler syndrome): evidence for heterogeneity.

Byler disease (ByD) is an autosomal recessive disorder in which cholestasis of onset in infancy leads to hepatic fibrosis and death. Children who have a clinically similar disorder, but are not members of the Amish kindred in which ByD was described, are said to have Byler syndrome (ByS). Controversy exists as to whether ByD and ByS (subtypes of progressive familial intrahepatic cholestasis [PFIC]) represent one clinicopathological entity. The gene for ByD has been mapped to a 19-cM region of 18q21-q22. PFIC caused by a lesion in this region, including ByD, can be designated PFIC-1. Examination of haplotypes in siblings with ByS in two unrelated non-Amish families showed that the gene(s) responsible for their disorder(s) did not lie in the PFIC-1 candidate region. On light microscopy and transmission electron microscopy (TEM), liver tissue differed between Amish children with PFIC-1, who had coarsely granular bile and at presentation had bland intracanalicular cholestasis, and the children with ByS in the two non-Amish families, who had amorphous or finely filamentous bile and at presentation had neonatal hepatitis. Bile acid composition of bile also differed: In the Amish children with PFIC-1 and in one ByS family, the proportional concentration of chenodeoxycholic acid (CDCA) in bile was low compared with normal bile; in the other ByS family, it was only slightly reduced. Genetic analysis and light microscopy and TEM of liver may help distinguish PFIC-1 from other forms of ByS.

Diabetes-like renal glomerular disease in Fanconi-Bickel syndrome.

The Fanconi-Bickel syndrome is a rare inherited disorder of metabolism characterized by hepatic glyconeogenesis, galactose intolerance, renal Fanconi syndrome with nephromegaly, and glycogen accumulation in proximal renal tubular cells. An 8-year-old patient with this disease and severe rickets due to medically resistant hypophosphatemia was found to have the previously unrecognized complication of renal glomerular hyperfiltration, microalbuminuria, and diffuse glomerular mesangial expansion. Similar to patients with glucose-6-phosphatase deficiency, the glomerular disease in this patient resembles incipient diabetic nephropathy. The Fanconi syndrome may be due to the defective transport of glucose at the proximal tubular basolateral membrane, which results in accumulation of glucose and secondarily glycogen within tubular cells. Since the metabolic defect, as evidenced by glycogen accumulation, selectively involves proximal renal tubular cells in the kidney of patients with Fanconi-Bickel syndrome and glucose-6-phosphatase deficiency, the abnormalities in renal glomerular hemodynamics and mesangial construct in these rare diseases are likely due to renal tubular factors, if the mechanism originates in the kidney. A delineation of these phenomena may further our understanding of the pathogenesis of diabetic nephropathy.

Severe perinatal liver disease and Down syndrome: an apparent relationship.

Down syndrome (DS) is not usually thought of in association with significant infantile liver disease. We present clinical and histopathologic data from 10 patients with DS who presented with severe liver disease at birth or within the first few weeks of life, and summarize the findings of eight previously reported cases. The liver disease was fatal in all but one case. Diffuse lobular fibrosis surrounding proliferating ductular elements and residual hepatocytes characterized the pathologic findings in the liver in all patients. A large number of megakaryocytes were present in the liver in nine of 12 patients. The phenotype of "perinatal hemochromatosis" was documented in eight of nine cases in which the presence of iron was investigated. Since only a fraction of the patients with this phenotype have DS, the patients we describe seem to represent a relatively well-defined subset of the perinatal hemochromatosis phenotype. The existence of such a subset suggests that the perinatal hemochromatosis phenotype does not represent a single etiopathogenetic disorder. The association between DS, megakaryocytic infiltrates in the liver, and fatal subacute/chronic liver disease gives rise to the speculation that fibrosis-promoting factors and/or metabolic abnormalities, such as those resulting from a gene dosage effect, may play a role in the genesis of the liver disease, perhaps due to a particular susceptibility of fetal liver.

Protracted mononucleosis-like illness associated with acquired cytomegalovirus infection in a previously healthy child: transient cellular immune defects and chronic hepatopathy.

Ordinarily, severe disease due to acquired cytomegalovirus (CMV) infection does not occur in immunocompetent children. We describe a previously healthy boy who acquired primary CMV infection at approximately 2 years of age and experienced a 2-year-long debilitating multisystem illness from which he ultimately recovered. Clinical features of this illness included fatigue, poor weight gain, pallor, unexplained fever, musculoskeletal complaints, drenching night sweats, lymphadenopathy, and massive hepatosplenomegaly. Laboratory abnormalities included elevated erythrocyte sedimentation rate, lymphocytosis, and elevated immune complex levels. Cellular immune function was impaired during the illness but was demonstrably normal during convalescence, and there was no other evidence for a known immunodeficiency state. Immunoblot analysis showed enhanced antibody response to a 66-kd infected cell protein after symptomatic recovery. Despite consistently normal indices of hepatic function, liver enlargement persisted after other symptoms had resolved. Liver biopsy demonstrated a mononuclear cell portal tract infiltrate with fibrosis, but CMV could not be demonstrated directly in this tissue. Primary CMV infection has not been reported previously to cause the persistent symptoms seen in this child.

Synergism between hepatic injuries and a nonhepatotropic reovirus in mice. Enhanced hepatic infection and death.

Reovirus type 1, after intravenous inoculation in the adult mouse, is secreted via bile into the intestine in an infectious form. Although reovirus type 1 is rapidly removed from systemic circulation by the liver and the lung, very few hepatocytes express reovirus antigen during infection. In intestinal cells, reovirus replicates selectively in the crypts. This site preference may be due to active cell proliferation in the crypts. We hypothesized that the state of the cell may affect virus replication and tested this hypothesis by using chemical and surgical means to increase hepatic mitotic activity. Adult mice were treated with carbon tetrachloride or surgical trauma, inoculated with reovirus type 1 intravenously, and subsequently killed. Virus antigen was identified using a highly specific immunohistochemical technique. Liver sections were stained using immunoperoxidase with specific rabbit antireovirus antibody. Hepatotoxin and surgical trauma increase reovirus antigen detection in both Kupffer cells and hepatocytes. Only the sequential administration of CCl4 and virus caused mortality at doses sublethal for each alone. These data demonstrate a synergism between hepatic injury and reovirus which results in a significant increase in the magnitude of viral infection and contributes to mortality. Such synergism may be important in idiopathic liver disease.

Site of reovirus replication in liver is determined by the type of hepatocellular insult.

Reovirus type 1 strain Lang is restricted from replicating in adult murine livers. In noninjured livers, approximately 1% of hepatocytes express reovirus antigen during infection. However, hepatocytes can be induced to express reovirus antigen if challenged with either toxins or trauma. We used selective hepatotoxins or surgical trauma to demonstrate that reovirus antigen localization in liver is determined by the site of hepatocellular insult and the timing of the virus inoculum.

Reovirus-induced liver disease in severe combined immunodeficient (SCID) mice. A model for the study of viral infection, pathogenesis, and clearance.

Adult severe combined immunodeficient (SCID) mice can be infected by the oral route with reovirus, and a systemic infection can be established. Infectious virus is recovered from all internal organs, and the mice die in 4-6 wk. Chronic, discrete inflammatory lesions appear in the liver of infected mice, and are associated with hepatocytes containing demonstrable levels of viral antigen. The adoptive transfer of Peyer's patch (PP) cells from congenic mice before infection protects the SCID mice against disease and death. Immune donor PP cells can be distinguished from nonimmune cells by their ability to contain and resolve infection by 1 wk after challenge.

Perinatal hemochromatosis: entity or end result?

A clinicopathologic picture of severe neonatal liver disease associated with heavy stainable iron stores in a "hemochromatotic" distribution has been designated as neonatal or perinatal iron storage disease or hemochromatosis. In an attempt to determine the specificity of these findings, we determined the amount and distribution of stainable iron in four groups of autopsied infants: those with severe subacute/chronic liver disease of undetermined etiology, those with subacute/chronic liver disease of known etiology, those with severe acute liver disease, and those with no liver disease. Fourteen of 15 infants with severe subacute/chronic liver disease of unknown etiology showed heavy iron accumulation in a hemochromatotic distribution in more than one organ, in contrast to only one of 37 cases in the other three groups. The findings are compatible with the possibility that the combination of morphologic findings defines a single disease entity (of currently unknown etiology), and also with the possibility that this disease entity is specific only to the extent of establishing a relationship between severe early (in utero) liver injury and a hemochromatotic distribution of iron in multiple organs.

HIV as a cause of giant cell hepatitis.

The 9-month-old daughter of human immunodeficiency virus (HIV)-seropositive parents presented with cholestasis and was found on liver biopsy to have giant cell hepatitis. No viral inclusions or particles were seen by light or electron microscopy. Ultrastructural studies of the liver biopsy demonstrated tubuloreticular structures in the endothelium and cylindrical confronting cisternae in inflammatory cells in the portal tracts. Serologic studies for hepatitis B, hepatitis A, and Epstein-Barr viruses were negative. Cytomegalovirus (CMV) was cultured from the urine, but buffy coat, nasopharyngeal, and liver cultures were negative and CMV antibody titer was low. The hepatitis responded dramatically to prednisone therapy. A repeat biopsy several months later revealed similar morphologic findings. AIDS was suspected on clinical and immunologic grounds, and was confirmed by the demonstration of HIV-specific IgG and IgM in serum. Five months after initial presentation, the infant developed Pneumocystis pneumonia, disseminated CMV infection, and died. This appears to be the first reported association of infantile giant cell hepatitis with HIV infection.

Bile canalicular morphometry in arteriohepatic dysplasia.

The diagnosis of arteriohepatic dysplasia may be difficult, particularly in very young patients with no family history, and the pathogenesis of the disorder remains obscure. It has been reported that the typical ultrastructural changes of cholestasis are scant in this condition and proposed that there is a failure of the hepatocytes to secrete bile into the canaliculi. In an attempt to expand on these observations and to test the value of ultrastructure as a differential diagnostic test for this condition, we chose two parameters felt to be associated with cholestasis that could be reliably and relatively simply determined by ultrastructural morphometry: canalicular dilatation and loss of canalicular microvilli. The values for these parameters were compared in arteriohepatic dysplasia with those in other infantile cholestatic conditions and with those in noncholestatic controls. The results of these morphometric studies corroborated the previous observation that canalicular dilatation in arteriohepatic dysplasia is minimal. There was, however, only a marginal difference in canalicular area between patients with arteriohepatic dysplasia and those with extrahepatic biliary atresia. There was significant overlap of values for canalicular area among all groups. Surprisingly, the studies showed no real loss of canalicular microvilli in any of the cholestatic conditions. The findings suggest that, at least for these two parameters, caution should be exercised in using qualitative electron microscopic evaluation as a diagnostic test for arteriohepatic dysplasia.

Transport of infectious reovirus into bile: class II major histocompatibility antigen-bearing cells determine reovirus transport.

We have previously demonstrated that mammalian reovirus type 1 enters the bile and gut lumen after systemic administration. In the present study, we showed that Kupffer cell uptake is essential for the transport of reovirus into the bile. Furthermore, class II major histocompatibility antigen (I-A)-bearing cells are a major determinant for the transit of reovirus from the hepatic environment, as well as from the intestine, during the course of systemic infection. These findings may provide an approach to the control of viral pathogens that cause systemic disease by selective utilization or modification of I-A-bearing cells.

Manganese-bilirubin cholestasis. Further studies in pathogenesis.

To further delineate the mechanism of manganese-bilirubin cholestasis, studies were done to determine: the structural status of the canalicular tight junctions in the model, and the effect of the lack of bilirubin conjugation. The results indicate that canalicular tight junction alterations characteristic of many types of cholestasis develop in this model and that bilirubin conjugation is not a requisite for the cholestatic effect of bilirubin to occur. It is unclear whether the tight junction alterations are the cause of the cholestasis or either secondary effects of the cholestasis or associated noncausal occurrences. The fact that bilirubin does not need to be conjugated in order to effect cholestasis confirms that the cholestasis is due to intrahepatocytic (as opposed to intracanalicular) effects, and may be useful in determining the precise locus and nature of these effects.

Sphingomyelin lipidosis variant with cirrhosis in the pediatric age group.

Two children with a variant of sphingomyelin lipidosis had otherwise unexplained cirrhosis that was histologically inactive and appeared to run an indolent course. The primary clinical problems involved the central nervous system, with vertical supranuclear ophthalmoplegia being the most distinctive feature. Biochemical analysis of cultured skin fibroblasts obtained from one of the children revealed that sphingomyelinase activity was 42% of control values. The typical inconspicuous hepatic storage and cirrhosis, coupled with the important morphologic finding of sea-blue histiocytes in the marrow, suggested that in cases of unexplained infantile or childhood cirrhosis the marrow should be closely examined for such histiocytes. Likewise, in cases of sea-blue histiocytes without evident etiology, with or without cirrhosis, this disease should be considered.

Medical management of bilateral renal malakoplakia.

Renal malakoplakia is reported in an 11 year old girl with myelomeningocele and associated neuropathic bladder. She is the first reported child to have survived bilateral renal malakoplakia with full recovery of renal function after medical management.

Electron microscopy of glandular Schwannoma.

This report is the first description of the ultrastructure of the glandular elements in a so-called glandular Schwannoma (malignant peripheral nerve sheath tumor with glandular elements). The appearance of this uncommon tumor was typical by light microscopy. Electron microscopy of the sarcomatous component revealed features consistent with Schwann cell origin. Ultrastructure of the glands revealed a well-differentiated epithelium with basement membrane, desmosomes and the presence of microvilli with core rootlets, glycocalyx, and intracytoplasmic R-bodies. No cilia or blepharoplasts were present, arguing against ependymal origin of these glands. The microvilli with core rootlets, the glycocalyx and the R-bodies indicate the glandular epithelium was of the "intestinal type." The basis for the presence of such epithelium in Schwann cell neoplasms is uncertain, but pathologic induction-interaction between neural crest and mesenchyme may be the most plausible explanation.

Antenatal diagnosis and early surgery for choledochal cyst.

The pathogenesis and optimal treatment of choledochal cyst have long remained questions of considerable speculation and dispute. Because the pregnancy of a 37-year-old woman was felt to be at risk, five antenatal ultrasound examinations were made. The fourth examination at 31.5 weeks of gestation demonstrated a choledochal cyst. Following birth, the child was studied with repeat ultrasound examinations and scintigraphy. The results of these studies plus the findings at surgery contributed new evidence regarding the possible pathogenesis and optimal treatment of choledochal cysts. Although the time at which the cyst originated is compatible with the concept of reflux of pancreatic juice into the common duct, no abnormal junction of the pancreatic and common duct was identified. No evidence of obstruction as part of the pathogenesis could be demonstrated. Progressive changes in the choledochal cyst in the first ten days of life suggested that delay in diagnosis and treatment of a choledochal cyst may contribute to early complications such as cholangitis. Early excision of the cyst in the newborn is considered to be the optimal treatment and may pose less risk to the patient than delayed surgical exploration.