Acute Viral Hepatitis

Hepatocellular damage

Changes seen under the light microscope range from minor degrees of cell swelling to cell death. They are accompanied by the inflammatory infiltration described below, reflecting the important role of cellular immunity in the pathogenesis of most forms of hepatitis. Both hepatocellular damage and inflammation are usually most severe in perivenular areas, giving rise to a characteristic histological pattern ( Fig. 6.1 ). A periportal pattern of necrosis and inflammation, sometimes seen in hepatitis A, is less common.

The mildest change is cell swelling, and this is probably reversible. The cytoplasm of affected cells is rarified, granular and sometimes finely vacuolated. The more severe degrees of cell swelling are called ballooning degeneration ( Fig. 6.2 ). This differs from the feathery degeneration of cholestasis, in which the cytoplasm has a reticular pattern ( see Fig. 5.3 ), and from the ballooning in steatohepatitis where the cytoplasm is less granular and more oedematous and ‘clarified’ ( see Fig. 7.10C ). Other hepatocytes undergo apoptosis, which is an important method of cell death in hepatitis. Shrinkage and increased staining of the cytoplasm, sometimes called acidophilic change or degeneration, is probably a precursor of apoptosis, in which the hepatocytes shrink further, become very dense and undergo fragmentation. The apoptotic bodies seen lying free in the sinusoids represent the largest fragments or entire unfragmented apoptotic cells ( Fig. 6.2 ). They are also called acidophil bodies or Councilman bodies, Councilman having first described them in yellow fever ( Fig. 6.3 ). Apoptotic bodies sometimes contain pyknotic nuclear remnants and often appear to bulge beyond the plane of the section. Another form of hepatocellular damage in acute hepatitis is focal (spotty) necrosis, in which liver-cell plates are disrupted or replaced by small groups of lymphocytes and macrophages. Whether these mark a site of necrosis or of apoptosis is not clear; the damage to hepatocytes is deduced from their absence rather than seen. Whatever its mechanism, loss of hepatocytes or liver-cell drop-out, coupled with focal regeneration, leads to a characteristic irregularity of the liver-cell plates, which usually allows acute hepatitis to be distinguished from hepatocellular damage secondary to cholestasis. The loss of hepatocytes also leads to condensation of the extracellular matrix, best seen in reticulin preparations ( Fig. 6.4 ).

Hepatocyte nuclei show prominent nucleoli and increased variation in size and may be multiple. When syncytial giant hepatocytes are very prominent, the term giant-cell hepatitis is appropriate. This is only rarely of proven viral origin and is also more characteristic of acute hepatitis in neonates. In adults, autoimmune hepatitis and hepatitis C virus with or without human immunodeficiency virus co-infection are important associations.

Cholestasis in the form of bile thrombi in canaliculi is common in acute hepatitis but rare in chronic hepatitis, which is diagnostically helpful. It is a result of damage to the bile-secretory apparatus of the hepatocytes, but may also result from interference with bile flow at the level of the portal tracts. The term cholestatic hepatitis is best kept as a clinical description of patients with a prolonged cholestatic course. Mild hepatocellular siderosis or steatosis is occasionally seen.

The inflammatory infiltrate

Unlike classic acute inflammation, viral hepatitis is characterised by a mainly lymphocytic infiltrate within the parenchyma and portal tracts. In acute hepatitis, the most conspicuous inflammation is usually perivenular. The extent of portal inflammation is very variable, and portal tracts may be either normal in size or expanded. The larger conducting tracts are often spared. The edges of small portal tracts may be well defined or blurred by outward extension of the infiltrate. This so-called spillover resembles the interface hepatitis of chronic hepatitis ( Ch. 9 ) and may be difficult to distinguish from it. The parenchymal changes, clinical history and virological findings usually make the correct diagnosis clear.

While most of the infiltrating cells in acute hepatitis are small T lymphocytes, plasma cells may also be prominent, and there are often a few neutrophils and eosinophils. The plasma cells do not necessarily indicate autoimmune hepatitis, nor do a few eosinophils prove a diagnosis of drug injury. Kupffer cells and other macrophages accumulate and enlarge, many of them forming discrete clumps together with lymphocytes. They may contain tan-brown ceroid pigment, staining with periodic acid–Schiff (PAS) agent after diastase digestion ( Fig. 6.5 ). They may also contain stainable iron ( Fig. 6.6 ), but this is less common.

Sinusoidal and venular endothelial cells also take part in the hepatitic process. Sinusoidal endothelial cells become swollen and may contain dense iron-positive granules ( Fig. 6.6 ). Terminal hepatic venules may show disruption of the endothelium and lymphocytic infiltration.

Portal changes

In contrast to chronic hepatitis, the parenchymal changes dominate the picture, but there is always some portal inflammation, affecting most or all of the small portal tracts ( Fig. 6.7 ). The density of the infiltrate varies. Interlobular bile ducts may show abnormalities, including irregularity, crowding and stratification of the epithelium, cytoplasmic vacuolation and infiltration by lymphocytes ( Fig. 6.8 ). These changes, together with formation of dense lymphoid structures (aggregates and follicles), are most often seen in hepatitis C. Bile-duct loss (ductopenia) is very rare.

Histological variants

The histological changes in acute hepatitis are infinitely variable, but a few patterns deserve special mention. These are confluent necrosis, bridging necrosis, necrosis of entire lobules and periportal necrosis.

Confluent necrosis signifies death of a substantial area of the parenchyma. Focal as opposed to zonal areas of confluent necrosis haphazardly distributed in relation to lobular zones are more likely to be due to causes other than acute viral hepatitis; possibilities to be considered include opportunistic infections with herpes simplex or zoster viruses and lymphoma. Bridging necrosis ( Figs 6.9 and 6.10 , and see Fig. 4.8 ) is the term given to confluent necrosis linking terminal venules to portal tracts. A possible explanation for this location is that it represents the entire zone 3 of an acinus, a view supported by the curved shape of many bridges. Bridging necrosis is a manifestation of severe acute hepatitis, but its distribution even within a single biopsy may be irregular. Necrosis and inflammation linking adjacent portal tracts without involvement of terminal venules should not strictly be called bridging because it almost certainly has different pathogenetic significance; it results from widening of portal tracts, with or without periportal necrosis.

Bridges of confluent necrosis with subsequent collapse may be mistaken for the septa of chronic liver disease. In making the important distinction between them, the pathologist is often helped by stains for elastic tissue. Unlike stains for collagens, these normally give negative results in the parenchyma, but elastic tissue accumulates as septa age. Recent collapse is therefore negative ( Fig. 6.11 ), whereas old septa are positive. Substantial amounts of elastic tissue take months or years to accumulate, but small amounts can be detected by sensitive methods such as Victoria blue as early as 1 or 2 months after onset of hepatitis.

In a minority of patients with acute viral hepatitis, confluent necrosis extends throughout entire lobules or acini ( panlobular or panacinar necrosis ) or several adjacent ones ( multilobular or multiacinar necrosis ). This is a common feature in patients with fulminant hepatitis. The term ‘massive necrosis’ is also sometimes used, but can be misleading in so far as a needle biopsy specimen may not be representative of the liver as a whole and can lead to over- or underestimation of the true extent of liver damage. This throws doubt on the usefulness of liver biopsy as a means of assessing prognosis in severe acute hepatitis. Sometimes multilobular necrosis involves only the subcapsular zone, and a small needle specimen may then give a falsely pessimistic picture ( see Fig. 1.3 ). In multilobular necrosis the parenchyma is replaced by collapsed stroma, inflammatory cells and activated macrophages ( Fig. 6.12 ). Around the surviving portal tracts there are prominent duct-like structures, some of which probably represent proliferation of pluripotential progenitor cells ( see Fig. 4.13D ). ‘Late-onset hepatic failure’ is a term used for patients developing encephalopathy between 8 and 24 weeks after onset of symptoms. Study of liver biopsies and explanted livers from these patients has shown a consistent pattern of map-like necrosis together with areas of nodular regeneration.

Periportal necrosis rather than the more usual perivenular necrosis is a feature in some patients with hepatitis A (discussed later).