Liver Diseases of Childhood

Incidence and Demographics

Cholestatic jaundice occurs in 1 in 2500 to 1 in 5000 newborns. Obstructive causes dominated by biliary atresia account for one-quarter to one-third of cases, whereas nonobstructive conditions account for approximately 60% to 70% of cases. An increasing number of specific disorders, especially those of bile acid synthesis and bilirubin metabolism, have been identified as causes of nonobstructive neonatal cholestasis (“neonatal hepatitis” on biopsy) in the past two decades; therefore although 65% of cases of neonatal hepatitis remained idiopathic in 1970, only about 15% remain undesignated at present. Various forms of inherited cholestasis account for 20% of all cases of neonatal hepatitis, alpha-1 antitrypsin deficiency for 10%, and other inborn errors of metabolism for 20%. Congenital infections, including those caused by the TORCH ( t oxoplasmosis, o ther agents, r ubella, c ytomegalovirus, h erpes simplex) agents, parvovirus, and various bacteria, account for about 5% of cases. However, as might be expected, there is considerable geographic variation in the incidence and prevalence of these conditions; thus although alpha-1 antitrypsin deficiency accounts for 25% of all cases of neonatal hepatitis in England, it is present in only 6% of children in Malaysia. Citrin deficiency is an important cause of neonatal cholestasis in Southeast Asia, whereas sepsis and hypoxia-ischemia account for most cases of neonatal cholestasis in Turkey. One-quarter of 101 infants with nonobstructive cholestasis in Brazil had infections, 25% had metabolic disease, and about 50% had idiopathic disease. Biliary atresia accounted for 25% of 101 infants with neonatal cholestasis in India, followed in frequency by sepsis/urinary tract infections and galactosemia.

Role of Liver Biopsy in Neonatal Cholestasis

Management of cholestatic jaundice consists first and foremost of excluding obstructive causes amenable to surgical correction. The most common cause of obstruction in infants is biliary atresia, which requires urgent portoenterostomy to allow bile flow and prevent rapidly progressive fibrosis and cirrhosis. Other conditions that require surgical intervention include choledochal cyst, choledocholithiasis, biliary strictures, and spontaneous idiopathic perforation of bile ducts. Although some of these can be diagnosed on imaging studies, the diagnosis of biliary atresia often requires a liver biopsy. Liver biopsy is also required to identify, whenever possible, the cause of nonobstructive cholestatic jaundice, often referred to as intrahepatic cholestasis to distinguish it from obstructive causes in the extra hepatic biliary tree. Thus evaluation of a liver biopsy specimen from a patient with neonatal cholestasis entails, first and foremost, the exclusion of biliary atresia, and, second, the identification of etiologic factors of nonobstructive jaundice that can be visualized on microscopic examination of a liver biopsy. Unfortunately, most injurious agents may simply produce a nonspecific histologic pattern referred to as neonatal hepatitis , a misleading term because a significant number of cases of “neonatal hepatitis” represent metabolic, chromosomal, and endocrine diseases, which are primarily neither infectious nor inflammatory in nature. However, although less than appropriate, the term benefits from familiarity of use in routine clinical practice and effectively describes a constellation of histologic findings that commonly occur in neonatal cholestatic diseases of various causes.

Incidence and Demographics

Biliary atresia occurs in 1 in 8000 to 1 in 18,000 newborns; it is the most common cause of cirrhosis, as well as the most common indication for liver transplantation in children. The incidence is greater in girls and in Asians and Africans. There is lack of concordance among twins, and familial predisposition has not been noted.

Clinical Manifestations

Children with biliary atresia are asymptomatic and almost always appear healthy and well nourished. In one study, low birth weight and prematurity were risk factors, but not in another study that reported normal birth weight and no association with prematurity. Patients are anicteric at birth and have jaundice in the first few weeks of life or may be continually jaundiced from birth, especially in the embryonal form. Jaundice is accompanied by dark urine and pale stools. The pathologic lesion evolves over time; thus the stools may be pigmented early in the course of the disease and become progressively paler as the obstruction worsens, ultimately acquiring the typical “clay-colored” appearance. Although pale stools may occur randomly in children, persistent pale-colored stools indicate biliary atresia or severe nonobstructive liver disease. Stool color cards are distributed to parents of newborns in Japan and Europe to facilitate identification of pale stools and therefore early diagnosis of biliary atresia. In 2014, an app that works across mobile platforms was made available in the United States with a similar goal. Congenital anomalies such as abnormalities of the abdominal veins, midline symmetrical liver, intestinal malrotation, situs anomalies, and polysplenia/asplenia may be present in children with the embryonal form of biliary atresia. Patients with biliary atresia usually have an enlarged, hard liver at presentation. Laboratory tests show elevation of bilirubin, alkaline phosphatase, and gamma glutamyltransferase (GGT).

Radiologic Features

An abdominal ultrasound is often performed for diagnosis, and specialized institutions have reported an accuracy rate of 98% for the diagnosis of biliary atresia on ultrasound findings. However, diagnostic accuracy is highly dependent on operator skill and experience. Ultrasonography helps to exclude other causes of obstruction such as choledochal cyst and stones and detect abnormalities of the gallbladder and bile ducts as well as other developmental anomalies associated with biliary atresia. The gallbladder is absent, atretic, or otherwise abnormal in biliary atresia; “gallbladder ghost triad,” composed of gallbladder length less than 19 mm, irregular wall, and indistinct mucosal lining, is considered to be characteristic. The triangular cord sign ( Fig. 5.1 ) is the presence of a triangular or tubular echogenic density above the bifurcation of the portal vein on ultrasonography, which corresponds to a cone of fibrous tissue at the porta hepatis. Reported sensitivity and specificity of the triangular cord sign for the diagnosis of biliary atresia are 70% to 85% and 90% to 100%, respectively. The positive predictive value of the triangular cord for biliary atresia is reported to be 100% when the gallbladder is abnormal and 88% when it is normal; an abnormal gallbladder without a triangular cord has a 25% positive predictive value.

The hepatobiliary iminodiacetic acid (HIDA) scan is based on excretion of an injected Tc-99m–labeled iminodiacetic isotope into the gut; nonexcretion into the gut in 24 hours indicates obstruction of the biliary tree ( Fig. 5.2 ). The accuracy is improved by administration of ursodeoxycholic acid (UDCA) or phenobarbital to facilitate bile flow. Although excretion into the gut definitively rules out biliary atresia, absence of excretion does not necessarily indicate obstruction because this may also occur in Alagille syndrome and other causes of severe nonobstructive cholestasis. A false-negative result may occur early in the course of biliary atresia when obstruction may be incomplete.

Endoscopic retrograde cholangiopancreatography is useful in the diagnosis of biliary atresia and may obviate the need for open laparotomy in doubtful cases. However, it is technically challenging and requires general anesthesia in young children, thus limiting its use to tertiary institutions with specialized pediatric gastroenterologists and skilled support staff.

Pathology

Macroscopic Pathology

The liver removed at transplantation in patients who have not had a prior Kasai procedure has a hard, dark green appearance and shows micronodular cirrhosis ( Fig. 5.3 ). When removed many years after a successful Kasai procedure, the cut surface, although also hard and dark green, shows large nodules in the perihilar region surrounded by cirrhotic micronodular liver at the periphery. It is postulated that the Kasai procedure achieves drainage in segments 4, 5, and 8, allowing regeneration of the perihilar parenchyma, which corresponds morphologically to the large perihilar nodules. Dilated intrahepatic bile ducts with inspissated bile may be found in both instances ( Fig. 5.4 ).

Microscopic Pathology

The microscopic features of biliary atresia in a liver biopsy reflect obstruction of the biliary tree. The portal tracts are expanded by edema, marked bile ductular reaction with numerous bile ductules, and minimal to mild inflammatory infiltrate ( Fig. 5.5 ). Sometimes, there may be a configuration resembling ductal plate remnants. Bile plugs are present in bile ducts and ductules. The interlobular bile ducts are present early in the course of the disease but may be absent late in the disease. The hepatic arterioles appear prominent and thick-walled, and they may be more numerous ( Fig. 5.6 and eSlide 5.1A ). There is almost always an increase in portal fibrous tissue, and depending on the age of the child and evolution of the lesion, varying degrees of fibrosis ranging from periportal to bridging fibrosis to frank cirrhosis are seen (see Fig. 5.5 ) ( eSlide 5.1B ). The lobular parenchyma shows cholestasis of varying, but usually marked, degree. Cholestasis is both cellular and canalicular and begins in the centrilobular regions. Enlarged, multinucleated hepatocytes (giant cells), extramedullary hemopoiesis, hepatocellular ballooning and damage, and apoptotic bodies ( Fig. 5.7 ) may be seen, rendering distinction from neonatal hepatitis difficult (further discussed in section on neonatal hepatitis). Periportal hepatocytes may stain for the biliary keratins, K7 and K19. ^∗

∗ Although the prefix CK is widely used in surgical pathology to designate human cytokeratins, consensus nomenclature recommends the replacement of “cytokeratin” with “keratin” and the prefix “CK” with “K.”

Distinctive portal changes of biliary atresia may not be present in the early stages of the disease, and a nondiagnostic biopsy performed in the first 6 weeks of life should be repeated if clinical suspicion for biliary atresia is strong. Furthermore, liver biopsy interpretation is dependent not only on the dynamics of the disease at the time of biopsy, but also on the experience of the pathologist. However, both overdiagnosis and underdiagnosis of biliary atresia may occur in a small number of cases, even in the most experienced hands.

The porta hepatis excised during a Kasai procedure shows varying degrees of destruction and obliteration of the major hepatic ducts. Surgeons often request rapid intraoperative evaluation of excised portions of the porta hepatic by frozen section to look for the presence and size of bile duct remnants; the presence of ducts larger than 150 or 200 μm has been associated with better bile flow after Kasai portoenterostomy. These size criteria are not universally used in current practice; for example, in our institution, although a larger-caliber duct is desirable, bile flow at the porta hepatis trumps duct diameter in evaluating the adequacy of surgical resection.

Duct remnants in the excised atretic biliary tree have been classified into three types. Type I is characterized by a completely atretic duct. Type II is characterized by a partially destroyed duct, which appears as a cleftlike lumen lined focally by cuboidal or low columnar cells. Smaller ductal structures with lumina measuring less than 50 μm in diameter and varying degrees of inflammation are usually present in the connective tissue surrounding the duct. Type III is characterized by a duct lined partially by columnar epithelium and smaller ductal structures with lumina less than 50 μm in diameter. An alternate classification that is based on the size of the duct lumen has also been published: type 1 consists of ducts with lumina 150 μm or greater in size; type 2 consists of ducts with lumina less than 150 μm in size; and type 3 does not demonstrate epithelial-lined structures. However, duct remnants of different types and variable luminal sizes may be present simultaneously at different levels of the biliary tract, reflecting varying stages of inflammatory destruction ( Fig. 5.8 and eSlide 5.2 ). Therefore it is not surprising that these classifications do not appear to have major clinical impact. The gallbladder appears atretic and underdeveloped with a small lumen lined by denuded mucosa ( Fig. 5.9 and eSlide 5.3 ).

The liver removed many years after a successful Kasai procedure shows large regenerative nodules in the perihilar region and small micronodules in the periphery. A ductular reaction is present at the edges of the nodules and is usually more prominent around the larger perihilar nodules ( Fig. 5.10A ). Intrahepatic bile ducts may be present in the perihilar regions but are usually absent in the peripheral liver where numerous arterioles unaccompanied by bile ducts are seen ( Fig. 5.10B and eSlide 5.4 ). The perihilar bile ducts may be dilated and contain biliary sludge or bile stones.

Differential Diagnosis

There are three considerations in the differential diagnosis of biliary atresia ( Table 5.2 ). The first is the distinction of biliary atresia from neonatal hepatitis. Although both neonatal hepatitis and biliary atresia show cholestasis, inflammation, giant cells, and ductular reaction, lobular changes predominate in neonatal hepatitis whereas portal tract changes are more prominent in biliary atresia. Ductular reaction, bile plugs, and portal fibrosis serve as the most important discriminatory features favoring biliary atresia.

The second consideration is the distinction of biliary atresia from nonobstructive conditions that may also show ductular reaction; these include alpha-1 antitrypsin deficiency ( Fig. 5.11 ), total parenteral nutrition ( Fig. 5.12 ), cystic fibrosis, progressive familial intrahepatic cholestasis-3, perinatal hypoxia, and Alagille syndrome ( Fig. 5.13 ). Differentiation among these conditions is based on a combination of clinical and laboratory findings (see Table 5.1 ). In this context, it is useful to remember that periodic acid–Schiff-positive globules may not be seen histologically in the very young infant with alpha-1 antitrypsin deficiency; in these patients the diagnosis should be established by serum enzyme levels and isoenzyme typing.

Finally, biliary atresia has to be distinguished from other obstructive conditions amenable to surgical resection such as choledochal cyst, bile duct strictures, choledocholithiasis, inspissated bile syndrome, neonatal sclerosing cholangitis, and idiopathic perforation of bile ducts. The differential diagnosis of these conditions cannot be resolved by histologic examination alone and requires correlation with clinical and imaging findings.

Treatment and Prognosis

Kasai portoenterostomy is the first line of treatment for biliary atresia. Best outcomes are associated with surgery performed by 60 days of age at a center experienced in this technique. Bile drainage is achieved in 60% of patients when surgery is performed at 60 to 70 days of life, in 40% when performed at 70 to 90 days of life, and in 25% when performed at 90 to 120 days of life. Age at surgery also influences the chances of long-term survival without transplantation. Better outcomes have been consistently linked to experience of the center performing the surgery. The presence of ducts 150 to 200 μm or greater in size and regeneration of perihilar liver parenchyma are associated with better outcomes. The prognostic significance of ductular reaction is not certain. Better results are reported in Japanese studies, especially over the long term, with one series reporting 44% survival with native liver 20 years after the Kasai procedure. Postoperative care in Japan routinely consists of long-term corticosteroids to prevent inflammatory destruction of intrahepatic bile ducts; intravenous antibiotics to prevent ascending cholangitis, a common complication that hastens bile duct damage and liver failure; long-term UDCA to facilitate bile flow; and herbal therapy to enhance liver function. A randomized clinical trial failed to demonstrate a statistically significant difference in bile flow at 6 months with use of post-Kasai high dose steroids; there was, however, an earlier onset of serious adverse events. The efficacy of the other measures routinely used post-Kasai in Japan has not been tested in randomized trials.

The Kasai procedure fails to restore bile flow in approximately one-third of infants, one-third require transplantation before 10 years of age, and the remaining third survive for more than 10 years with the portoenterostomy. Although most children eventually require transplantation, the Kasai procedure is an effective first line of management because it allows the child to grow and develop before transplantation. Despite clinical or biochemical evidence of liver disease, more than half of all children report a good quality of life after transplantation. A previous Kasai procedure does not render subsequent transplantation medically or technically difficult. Biliary atresia accounts for almost 50% of all liver transplants in children. In a study of 1976 patients who underwent liver transplantation for biliary atresia, the 1-, 5-, and 10-year survival rates were 90%, 87.2%, and 85.8%, respectively.